[Met_help] [rt.rap.ucar.edu #95747] History for MET V9.0 Point2Grid (UNCLASSIFIED)

John Halley Gotway via RT met_help at ucar.edu
Mon Jul 20 17:16:08 MDT 2020 

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

CLASSIFICATION: UNCLASSIFIED

The optional argument for smoothing or interpolation shown in the User's
Guide is Gaussian, but are other methods available? If so, can you provide
the names and how to specify them as an optional argument?

Thanks.

Stay safe.

T/
John

Mr. John W. Raby
U.S. Army Research Laboratory
White Sands Missile Range, NM 88002
Office:(575) 678-2004 DSN 258-2004
Teleworking: Contact Admin Specialist at (410) 278-6974
FAX (575) 678-1230 DSN 258-1230
Email: john.w.raby2.civ at mail.mil

CLASSIFICATION: UNCLASSIFIED


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Fri Jun 26 23:06:50 2020

John,

I see you have a question about interpolation options in the
point2grid
tool. This XML file contains the calls to the point2grid tool that are
run
for our unit tests:
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml

While most of them DO NOT use the "-method" command line option, a few
at
the end do use "-method MAX" to select the maximum value falling
within
each grid box.

Inspecting the code in this file:
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
I see that the supported options for -method are MIN, MAX, MEDIAN, and
UW_MEAN.

Seems like we should add that info to the usage statement! I wrote up
this
GitHub issue to make sure we remember to do so before the 9.1 release:
https://github.com/NCAR/MET/issues/1396

Hope that helps clarify.

Thanks,
John

On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> Transaction: Ticket created by john.w.raby2.civ at mail.mil
>        Queue: met_help
>      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.w.raby2.civ at mail.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
>
> CLASSIFICATION: UNCLASSIFIED
>
> The optional argument for smoothing or interpolation shown in the
User's
> Guide is Gaussian, but are other methods available? If so, can you
provide
> the names and how to specify them as an optional argument?
>
> Thanks.
>
> Stay safe.
>
> T/
> John
>
> Mr. John W. Raby
> U.S. Army Research Laboratory
> White Sands Missile Range, NM 88002
> Office:(575) 678-2004 DSN 258-2004
> Teleworking: Contact Admin Specialist at (410) 278-6974
> FAX (575) 678-1230 DSN 258-1230
> Email: john.w.raby2.civ at mail.mil
>
> CLASSIFICATION: UNCLASSIFIED
>
>

------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Sat Jun 27 07:59:33 2020

John -

Thanks for your quick response which is helpful. Appreciate your
dedication so late into the evening! Can I add new variables in the
config file? Any info on the science behind the interpolation methods
would be appreciated, but that can wait until Monday.

Take care and stay safe.

R/
John
________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Friday, June 26, 2020 11:06 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

I see you have a question about interpolation options in the
point2grid
tool. This XML file contains the calls to the point2grid tool that are
run
for our unit tests:
Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml

While most of them DO NOT use the "-method" command line option, a few
at
the end do use "-method MAX" to select the maximum value falling
within
each grid box.

Inspecting the code in this file:
Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
I see that the supported options for -method are MIN, MAX, MEDIAN, and
UW_MEAN.

Seems like we should add that info to the usage statement! I wrote up
this
GitHub issue to make sure we remember to do so before the 9.1 release:
Caution-https://github.com/NCAR/MET/issues/1396

Hope that helps clarify.

Thanks,
John

On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> Transaction: Ticket created by john.w.raby2.civ at mail.mil
>        Queue: met_help
>      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.w.raby2.civ at mail.mil
>       Status: new
>  Ticket <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
>
> CLASSIFICATION: UNCLASSIFIED
>
> The optional argument for smoothing or interpolation shown in the
User's
> Guide is Gaussian, but are other methods available? If so, can you
provide
> the names and how to specify them as an optional argument?
>
> Thanks.
>
> Stay safe.
>
> T/
> John
>
> Mr. John W. Raby
> U.S. Army Research Laboratory
> White Sands Missile Range, NM 88002
> Office:(575) 678-2004 DSN 258-2004
> Teleworking: Contact Admin Specialist at (410) 278-6974
> FAX (575) 678-1230 DSN 258-1230
> Email: john.w.raby2.civ at mail.mil
>
> CLASSIFICATION: UNCLASSIFIED
>
>



------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Sat Jun 27 20:48:28 2020

Another question I have is how is -method used in this tool? In the
User's Guide, -method provides options for regridding from one grid to
another such as is used in regrid_data_plane. In the case of
Point2Grid, the process involved is not regridding in the way that I
understand it. If I understand it correctly, it takes point
observations and distributes them on to a user provided grid using
some interpolation method. The next step I think is the smoothing of
the field to even out irregularities which are caused by errors,
noise, etc resulting in the fitting of the observed values to some
function which appear to be a Gaussian function. I guess I also need
to understand what happens when the tool is used without any of the
optional arguments.

Thanks.

R/
John
________________________________________
From: Raby, John W CIV USARMY CCDC ARL (USA)
Sent: Saturday, June 27, 2020 7:59 AM
To: met_help at ucar.edu
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

John -

Thanks for your quick response which is helpful. Appreciate your
dedication so late into the evening! Can I add new variables in the
config file? Any info on the science behind the interpolation methods
would be appreciated, but that can wait until Monday.

Take care and stay safe.

R/
John
________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Friday, June 26, 2020 11:06 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

I see you have a question about interpolation options in the
point2grid
tool. This XML file contains the calls to the point2grid tool that are
run
for our unit tests:
Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml

While most of them DO NOT use the "-method" command line option, a few
at
the end do use "-method MAX" to select the maximum value falling
within
each grid box.

Inspecting the code in this file:
Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
I see that the supported options for -method are MIN, MAX, MEDIAN, and
UW_MEAN.

Seems like we should add that info to the usage statement! I wrote up
this
GitHub issue to make sure we remember to do so before the 9.1 release:
Caution-https://github.com/NCAR/MET/issues/1396

Hope that helps clarify.

Thanks,
John

On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> Transaction: Ticket created by john.w.raby2.civ at mail.mil
>        Queue: met_help
>      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.w.raby2.civ at mail.mil
>       Status: new
>  Ticket <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
>
> CLASSIFICATION: UNCLASSIFIED
>
> The optional argument for smoothing or interpolation shown in the
User's
> Guide is Gaussian, but are other methods available? If so, can you
provide
> the names and how to specify them as an optional argument?
>
> Thanks.
>
> Stay safe.
>
> T/
> John
>
> Mr. John W. Raby
> U.S. Army Research Laboratory
> White Sands Missile Range, NM 88002
> Office:(575) 678-2004 DSN 258-2004
> Teleworking: Contact Admin Specialist at (410) 278-6974
> FAX (575) 678-1230 DSN 258-1230
> Email: john.w.raby2.civ at mail.mil
>
> CLASSIFICATION: UNCLASSIFIED
>
>



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jun 29 09:41:16 2020

John,

While the "-field" command line argument indicates that it can be used
one
or more times, in this first version of the tool, it can only be used
once.

If you specify "-field" more that once you'll get this error message:
ERROR  :
ERROR  : process_point_file() -> Support only 1 variable. "2"
variables
were requested.
ERROR  :

We do plan to support the processing of multiple fields but it has not
yet
been implemented.

As for the science behind the interpolation methods, it's very simple.
Keep
track of the point observations falling inside each of the target grid
boxes. Write to the output the min, max, unweighted mean, or median of
those values based on the method chosen. If you only have 1 point obs
falling inside each grid box, then all those methods will result in
the
same output value. Whereas if you have many, as in the case of GOES
input
files, then the choice of the derivation method would matter.

Hope that helps clarify.

Thanks,
John


On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Another question I have is how is -method used in this tool? In the
User's
> Guide, -method provides options for regridding from one grid to
another
> such as is used in regrid_data_plane. In the case of Point2Grid, the
> process involved is not regridding in the way that I understand it.
If I
> understand it correctly, it takes point observations and distributes
them
> on to a user provided grid using some interpolation method. The next
step I
> think is the smoothing of the field to even out irregularities which
are
> caused by errors, noise, etc resulting in the fitting of the
observed
> values to some function which appear to be a Gaussian function. I
guess I
> also need to understand what happens when the tool is used without
any of
> the optional arguments.
>
> Thanks.
>
> R/
> John
> ________________________________________
> From: Raby, John W CIV USARMY CCDC ARL (USA)
> Sent: Saturday, June 27, 2020 7:59 AM
> To: met_help at ucar.edu
> Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> John -
>
> Thanks for your quick response which is helpful. Appreciate your
> dedication so late into the evening! Can I add new variables in the
config
> file? Any info on the science behind the interpolation methods would
be
> appreciated, but that can wait until Monday.
>
> Take care and stay safe.
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Friday, June 26, 2020 11:06 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> I see you have a question about interpolation options in the
point2grid
> tool. This XML file contains the calls to the point2grid tool that
are run
> for our unit tests:
> Caution-
>
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
>
> While most of them DO NOT use the "-method" command line option, a
few at
> the end do use "-method MAX" to select the maximum value falling
within
> each grid box.
>
> Inspecting the code in this file:
> Caution-
>
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> I see that the supported options for -method are MIN, MAX, MEDIAN,
and
> UW_MEAN.
>
> Seems like we should add that info to the usage statement! I wrote
up this
> GitHub issue to make sure we remember to do so before the 9.1
release:
> Caution-https://github.com/NCAR/MET/issues/1396
>
> Hope that helps clarify.
>
> Thanks,
> John
>
> On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> >        Queue: met_help
> >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> >        Owner: Nobody
> >   Requestors: john.w.raby2.civ at mail.mil
> >       Status: new
> >  Ticket <Caution-url: Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> > The optional argument for smoothing or interpolation shown in the
User's
> > Guide is Gaussian, but are other methods available? If so, can you
> provide
> > the names and how to specify them as an optional argument?
> >
> > Thanks.
> >
> > Stay safe.
> >
> > T/
> > John
> >
> > Mr. John W. Raby
> > U.S. Army Research Laboratory
> > White Sands Missile Range, NM 88002
> > Office:(575) 678-2004 DSN 258-2004
> > Teleworking: Contact Admin Specialist at (410) 278-6974
> > FAX (575) 678-1230 DSN 258-1230
> > Email: john.w.raby2.civ at mail.mil
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> >
>
>
>
>

------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jun 29 10:16:52 2020

John -

Thanks for the clarification which is helping me to understand more
about this tool. That link to the github is really helpful to for
seeing some example usages. I will be interested in this tool since it
promises new possibilities for a source of gridded obs at smaller grid
spacing for spatial verification. For the smoothing, is Gaussian the
only interpolation method available?

R/
John

From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Monday, June 29, 2020 9:41 AM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

While the "-field" command line argument indicates that it can be used
one
or more times, in this first version of the tool, it can only be used
once.

If you specify "-field" more that once you'll get this error message:
ERROR  :
ERROR  : process_point_file() -> Support only 1 variable. "2"
variables
were requested.
ERROR  :

We do plan to support the processing of multiple fields but it has not
yet
been implemented.

As for the science behind the interpolation methods, it's very simple.
Keep
track of the point observations falling inside each of the target grid
boxes. Write to the output the min, max, unweighted mean, or median of
those values based on the method chosen. If you only have 1 point obs
falling inside each grid box, then all those methods will result in
the
same output value. Whereas if you have many, as in the case of GOES
input
files, then the choice of the derivation method would matter.

Hope that helps clarify.

Thanks,
John


On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Another question I have is how is -method used in this tool? In the
User's
> Guide, -method provides options for regridding from one grid to
another
> such as is used in regrid_data_plane. In the case of Point2Grid, the
> process involved is not regridding in the way that I understand it.
If I
> understand it correctly, it takes point observations and distributes
them
> on to a user provided grid using some interpolation method. The next
step I
> think is the smoothing of the field to even out irregularities which
are
> caused by errors, noise, etc resulting in the fitting of the
observed
> values to some function which appear to be a Gaussian function. I
guess I
> also need to understand what happens when the tool is used without
any of
> the optional arguments.
>
> Thanks.
>
> R/
> John
> ________________________________________
> From: Raby, John W CIV USARMY CCDC ARL (USA)
> Sent: Saturday, June 27, 2020 7:59 AM
> To: met_help at ucar.edu
> Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> John -
>
> Thanks for your quick response which is helpful. Appreciate your
> dedication so late into the evening! Can I add new variables in the
config
> file? Any info on the science behind the interpolation methods would
be
> appreciated, but that can wait until Monday.
>
> Take care and stay safe.
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Friday, June 26, 2020 11:06 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> I see you have a question about interpolation options in the
point2grid
> tool. This XML file contains the calls to the point2grid tool that
are run
> for our unit tests:
> Caution-
> Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
>
> While most of them DO NOT use the "-method" command line option, a
few at
> the end do use "-method MAX" to select the maximum value falling
within
> each grid box.
>
> Inspecting the code in this file:
> Caution-
> Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> I see that the supported options for -method are MIN, MAX, MEDIAN,
and
> UW_MEAN.
>
> Seems like we should add that info to the usage statement! I wrote
up this
> GitHub issue to make sure we remember to do so before the 9.1
release:
> Caution-Caution-https://github.com/NCAR/MET/issues/1396
>
> Hope that helps clarify.
>
> Thanks,
> John
>
> On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> >        Queue: met_help
> >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> >        Owner: Nobody
> >   Requestors: john.w.raby2.civ at mail.mil
> >       Status: new
> >  Ticket <Caution-Caution-url: Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> > The optional argument for smoothing or interpolation shown in the
User's
> > Guide is Gaussian, but are other methods available? If so, can you
> provide
> > the names and how to specify them as an optional argument?
> >
> > Thanks.
> >
> > Stay safe.
> >
> > T/
> > John
> >
> > Mr. John W. Raby
> > U.S. Army Research Laboratory
> > White Sands Missile Range, NM 88002
> > Office:(575) 678-2004 DSN 258-2004
> > Teleworking: Contact Admin Specialist at (410) 278-6974
> > FAX (575) 678-1230 DSN 258-1230
> > Email: john.w.raby2.civ at mail.mil
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> >
>
>
>
>



------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jun 29 10:57:48 2020

John -

I ran it to see the usage and there are details in the usage dialog
and also in the XML file you pointed me to which helped to answer my
own question (I think).
There is an additional interpolation method called GAUSSIAN which
applies only when using the probability variable (-prob_cat_thresh
280.0). So, it looks like the two optional smoothing arguments
-gaussian_dx and -gaussian_radius are only used when the method is
GAUSSIAN. IS this correct? This means that there is no smoothing
except when using probability variables. Is this correct?

R/
John
________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Monday, June 29, 2020 9:41 AM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

While the "-field" command line argument indicates that it can be used
one
or more times, in this first version of the tool, it can only be used
once.

If you specify "-field" more that once you'll get this error message:
ERROR  :
ERROR  : process_point_file() -> Support only 1 variable. "2"
variables
were requested.
ERROR  :

We do plan to support the processing of multiple fields but it has not
yet
been implemented.

As for the science behind the interpolation methods, it's very simple.
Keep
track of the point observations falling inside each of the target grid
boxes. Write to the output the min, max, unweighted mean, or median of
those values based on the method chosen. If you only have 1 point obs
falling inside each grid box, then all those methods will result in
the
same output value. Whereas if you have many, as in the case of GOES
input
files, then the choice of the derivation method would matter.

Hope that helps clarify.

Thanks,
John


On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Another question I have is how is -method used in this tool? In the
User's
> Guide, -method provides options for regridding from one grid to
another
> such as is used in regrid_data_plane. In the case of Point2Grid, the
> process involved is not regridding in the way that I understand it.
If I
> understand it correctly, it takes point observations and distributes
them
> on to a user provided grid using some interpolation method. The next
step I
> think is the smoothing of the field to even out irregularities which
are
> caused by errors, noise, etc resulting in the fitting of the
observed
> values to some function which appear to be a Gaussian function. I
guess I
> also need to understand what happens when the tool is used without
any of
> the optional arguments.
>
> Thanks.
>
> R/
> John
> ________________________________________
> From: Raby, John W CIV USARMY CCDC ARL (USA)
> Sent: Saturday, June 27, 2020 7:59 AM
> To: met_help at ucar.edu
> Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> John -
>
> Thanks for your quick response which is helpful. Appreciate your
> dedication so late into the evening! Can I add new variables in the
config
> file? Any info on the science behind the interpolation methods would
be
> appreciated, but that can wait until Monday.
>
> Take care and stay safe.
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Friday, June 26, 2020 11:06 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> I see you have a question about interpolation options in the
point2grid
> tool. This XML file contains the calls to the point2grid tool that
are run
> for our unit tests:
> Caution-
> Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
>
> While most of them DO NOT use the "-method" command line option, a
few at
> the end do use "-method MAX" to select the maximum value falling
within
> each grid box.
>
> Inspecting the code in this file:
> Caution-
> Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> I see that the supported options for -method are MIN, MAX, MEDIAN,
and
> UW_MEAN.
>
> Seems like we should add that info to the usage statement! I wrote
up this
> GitHub issue to make sure we remember to do so before the 9.1
release:
> Caution-Caution-https://github.com/NCAR/MET/issues/1396
>
> Hope that helps clarify.
>
> Thanks,
> John
>
> On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> >        Queue: met_help
> >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> >        Owner: Nobody
> >   Requestors: john.w.raby2.civ at mail.mil
> >       Status: new
> >  Ticket <Caution-Caution-url: Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> > The optional argument for smoothing or interpolation shown in the
User's
> > Guide is Gaussian, but are other methods available? If so, can you
> provide
> > the names and how to specify them as an optional argument?
> >
> > Thanks.
> >
> > Stay safe.
> >
> > T/
> > John
> >
> > Mr. John W. Raby
> > U.S. Army Research Laboratory
> > White Sands Missile Range, NM 88002
> > Office:(575) 678-2004 DSN 258-2004
> > Teleworking: Contact Admin Specialist at (410) 278-6974
> > FAX (575) 678-1230 DSN 258-1230
> > Email: john.w.raby2.civ at mail.mil
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> >
>
>
>
>



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jun 29 11:31:23 2020

John,

I'm reassigning this ticket to Howard Soh who worked on the
development of
the point2grid tool. He should be able to better answer these
questions.

Thanks,
John

On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> John -
>
> I ran it to see the usage and there are details in the usage dialog
and
> also in the XML file you pointed me to which helped to answer my own
> question (I think).
> There is an additional interpolation method called GAUSSIAN which
applies
> only when using the probability variable (-prob_cat_thresh 280.0).
So, it
> looks like the two optional smoothing arguments -gaussian_dx and
> -gaussian_radius are only used when the method is GAUSSIAN. IS this
> correct? This means that there is no smoothing except when using
> probability variables. Is this correct?
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, June 29, 2020 9:41 AM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> While the "-field" command line argument indicates that it can be
used one
> or more times, in this first version of the tool, it can only be
used once.
>
> If you specify "-field" more that once you'll get this error
message:
> ERROR  :
> ERROR  : process_point_file() -> Support only 1 variable. "2"
variables
> were requested.
> ERROR  :
>
> We do plan to support the processing of multiple fields but it has
not yet
> been implemented.
>
> As for the science behind the interpolation methods, it's very
simple. Keep
> track of the point observations falling inside each of the target
grid
> boxes. Write to the output the min, max, unweighted mean, or median
of
> those values based on the method chosen. If you only have 1 point
obs
> falling inside each grid box, then all those methods will result in
the
> same output value. Whereas if you have many, as in the case of GOES
input
> files, then the choice of the derivation method would matter.
>
> Hope that helps clarify.
>
> Thanks,
> John
>
>
> On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-url: Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > Another question I have is how is -method used in this tool? In
the
> User's
> > Guide, -method provides options for regridding from one grid to
another
> > such as is used in regrid_data_plane. In the case of Point2Grid,
the
> > process involved is not regridding in the way that I understand
it. If I
> > understand it correctly, it takes point observations and
distributes them
> > on to a user provided grid using some interpolation method. The
next
> step I
> > think is the smoothing of the field to even out irregularities
which are
> > caused by errors, noise, etc resulting in the fitting of the
observed
> > values to some function which appear to be a Gaussian function. I
guess I
> > also need to understand what happens when the tool is used without
any of
> > the optional arguments.
> >
> > Thanks.
> >
> > R/
> > John
> > ________________________________________
> > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > Sent: Saturday, June 27, 2020 7:59 AM
> > To: met_help at ucar.edu
> > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > John -
> >
> > Thanks for your quick response which is helpful. Appreciate your
> > dedication so late into the evening! Can I add new variables in
the
> config
> > file? Any info on the science behind the interpolation methods
would be
> > appreciated, but that can wait until Monday.
> >
> > Take care and stay safe.
> >
> > R/
> > John
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Friday, June 26, 2020 11:06 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > I see you have a question about interpolation options in the
point2grid
> > tool. This XML file contains the calls to the point2grid tool that
are
> run
> > for our unit tests:
> > Caution-
> > Caution-
>
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
> >
> > While most of them DO NOT use the "-method" command line option, a
few at
> > the end do use "-method MAX" to select the maximum value falling
within
> > each grid box.
> >
> > Inspecting the code in this file:
> > Caution-
> > Caution-
>
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> > I see that the supported options for -method are MIN, MAX, MEDIAN,
and
> > UW_MEAN.
> >
> > Seems like we should add that info to the usage statement! I wrote
up
> this
> > GitHub issue to make sure we remember to do so before the 9.1
release:
> > Caution-Caution-https://github.com/NCAR/MET/issues/1396
> >
> > Hope that helps clarify.
> >
> > Thanks,
> > John
> >
> > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > >        Queue: met_help
> > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > >        Owner: Nobody
> > >   Requestors: john.w.raby2.civ at mail.mil
> > >       Status: new
> > >  Ticket <Caution-Caution-url: Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > The optional argument for smoothing or interpolation shown in
the
> User's
> > > Guide is Gaussian, but are other methods available? If so, can
you
> > provide
> > > the names and how to specify them as an optional argument?
> > >
> > > Thanks.
> > >
> > > Stay safe.
> > >
> > > T/
> > > John
> > >
> > > Mr. John W. Raby
> > > U.S. Army Research Laboratory
> > > White Sands Missile Range, NM 88002
> > > Office:(575) 678-2004 DSN 258-2004
> > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > FAX (575) 678-1230 DSN 258-1230
> > > Email: john.w.raby2.civ at mail.mil
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> >
> >
> >
> >
>
>
>
>

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jun 29 12:50:05 2020

Two layers for the interpolation with point2grid:
- layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
- layer 2: Optional, the GAUSSIAN smoothing for the probability
variable. (the existence of valid values if the probability threshold
is not given)

+ "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which is a
default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
+ "-method MAXGAUSS" is processed as the "-method MAX" for the layer 1
and "-method GAUSSIAN" for the layer 2.
+ Override the interpolation method of layer 1 with "-method <METHOD>"

1. The Gaussian smoothing at point2grid is activated and applied into
the probability variable only if -method GAUSSIAN or -method MAXGAUSS
is given as the command line argument. It's not applied into the
variable at "-field" argument. If you want, You can do this by using
regrid_data_plane (passing the output of point2grid as the input of
regrid_data_plane with GAUSSIAN method).

2. -gaussian_dx and -gaussian_radius are ignored if -method GAUSSIAN
or -method MAXGAUSS is NOT given. The default values are applied if
not given.

3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
MAXGAUSS argument.

4 Without -prob_cat_thresh argument, it will be the status of data
existence. With -prob_cat_thresh argument, it will be meeting the
condition or not. So the values of the probability are 0 or 1.

5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
(between 0. and 1.0 with GAUSSIAN smoothing).

Cheers,
Howard

On Mon Jun 29 11:31:23 2020, johnhg wrote:
> John,
>
> I'm reassigning this ticket to Howard Soh who worked on the
> development of
> the point2grid tool. He should be able to better answer these
> questions.

>
> Thanks,
> John
>
> On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > John -
> >
> > I ran it to see the usage and there are details in the usage
dialog
> > and
> > also in the XML file you pointed me to which helped to answer my
own
> > question (I think).
> > There is an additional interpolation method called GAUSSIAN which
> > applies
> > only when using the probability variable (-prob_cat_thresh 280.0).
> > So, it
> > looks like the two optional smoothing arguments -gaussian_dx and
> > -gaussian_radius are only used when the method is GAUSSIAN. IS
this
> > correct? This means that there is no smoothing except when using
> > probability variables. Is this correct?
> >
> > R/
> > John
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Monday, June 29, 2020 9:41 AM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > While the "-field" command line argument indicates that it can be
> > used one
> > or more times, in this first version of the tool, it can only be
used
> > once.
> >
> > If you specify "-field" more that once you'll get this error
message:
> > ERROR  :
> > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > variables
> > were requested.
> > ERROR  :
> >
> > We do plan to support the processing of multiple fields but it has
> > not yet
> > been implemented.
> >
> > As for the science behind the interpolation methods, it's very
> > simple. Keep
> > track of the point observations falling inside each of the target
> > grid
> > boxes. Write to the output the min, max, unweighted mean, or
median
> > of
> > those values based on the method chosen. If you only have 1 point
obs
> > falling inside each grid box, then all those methods will result
in
> > the
> > same output value. Whereas if you have many, as in the case of
GOES
> > input
> > files, then the choice of the derivation method would matter.
> >
> > Hope that helps clarify.
> >
> > Thanks,
> > John
> >
> >
> > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-url: Caution-
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > Another question I have is how is -method used in this tool? In
the
> > User's
> > > Guide, -method provides options for regridding from one grid to
> > > another
> > > such as is used in regrid_data_plane. In the case of Point2Grid,
> > > the
> > > process involved is not regridding in the way that I understand
it.
> > > If I
> > > understand it correctly, it takes point observations and
> > > distributes them
> > > on to a user provided grid using some interpolation method. The
> > > next
> > step I
> > > think is the smoothing of the field to even out irregularities
> > > which are
> > > caused by errors, noise, etc resulting in the fitting of the
> > > observed
> > > values to some function which appear to be a Gaussian function.
I
> > > guess I
> > > also need to understand what happens when the tool is used
without
> > > any of
> > > the optional arguments.
> > >
> > > Thanks.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Sent: Saturday, June 27, 2020 7:59 AM
> > > To: met_help at ucar.edu
> > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > John -
> > >
> > > Thanks for your quick response which is helpful. Appreciate your
> > > dedication so late into the evening! Can I add new variables in
the
> > config
> > > file? Any info on the science behind the interpolation methods
> > > would be
> > > appreciated, but that can wait until Monday.
> > >
> > > Take care and stay safe.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > Sent: Friday, June 26, 2020 11:06 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > the
> > > identity of the sender, and confirm the authenticity of all
links
> > contained
> > > within the message prior to copying and pasting the address to a
> > > Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > John,
> > >
> > > I see you have a question about interpolation options in the
> > > point2grid
> > > tool. This XML file contains the calls to the point2grid tool
that
> > > are
> > run
> > > for our unit tests:
> > > Caution-
> > > Caution-
> >
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
> > >
> > > While most of them DO NOT use the "-method" command line option,
a
> > > few at
> > > the end do use "-method MAX" to select the maximum value falling
> > > within
> > > each grid box.
> > >
> > > Inspecting the code in this file:
> > > Caution-
> > > Caution-
> >
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> > > I see that the supported options for -method are MIN, MAX,
MEDIAN,
> > > and
> > > UW_MEAN.
> > >
> > > Seems like we should add that info to the usage statement! I
wrote
> > > up
> > this
> > > GitHub issue to make sure we remember to do so before the 9.1
> > > release:
> > > Caution-Caution-https://github.com/NCAR/MET/issues/1396
> > >
> > > Hope that helps clarify.
> > >
> > > Thanks,
> > > John
> > >
> > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > >        Queue: met_help
> > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > >        Owner: Nobody
> > > >   Requestors: john.w.raby2.civ at mail.mil
> > > >       Status: new
> > > >  Ticket <Caution-Caution-url: Caution-
> > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747
>
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > The optional argument for smoothing or interpolation shown in
the
> > User's
> > > > Guide is Gaussian, but are other methods available? If so, can
> > > > you
> > > provide
> > > > the names and how to specify them as an optional argument?
> > > >
> > > > Thanks.
> > > >
> > > > Stay safe.
> > > >
> > > > T/
> > > > John
> > > >
> > > > Mr. John W. Raby
> > > > U.S. Army Research Laboratory
> > > > White Sands Missile Range, NM 88002
> > > > Office:(575) 678-2004 DSN 258-2004
> > > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > > FAX (575) 678-1230 DSN 258-1230
> > > > Email: john.w.raby2.civ at mail.mil
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >



------------------------------------------------
Subject: RE: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jun 29 14:13:44 2020

Howard -

Thanks for your quick response with details in answer to my question.
Right now my interest is with non-probability fields, so your point in
1. about the possibility of using Gaussian smoothing for -field
argument is interesting.

R/
John
________________________________________
From: Howard Soh via RT [met_help at ucar.edu]
Sent: Monday, June 29, 2020 12:50 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

Two layers for the interpolation with point2grid:
- layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
- layer 2: Optional, the GAUSSIAN smoothing for the probability
variable. (the existence of valid values if the probability threshold
is not given)

+ "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which is a
default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
+ "-method MAXGAUSS" is processed as the "-method MAX" for the layer 1
and "-method GAUSSIAN" for the layer 2.
+ Override the interpolation method of layer 1 with "-method <METHOD>"

1. The Gaussian smoothing at point2grid is activated and applied into
the probability variable only if -method GAUSSIAN or -method MAXGAUSS
is given as the command line argument. It's not applied into the
variable at "-field" argument. If you want, You can do this by using
regrid_data_plane (passing the output of point2grid as the input of
regrid_data_plane with GAUSSIAN method).

2. -gaussian_dx and -gaussian_radius are ignored if -method GAUSSIAN
or -method MAXGAUSS is NOT given. The default values are applied if
not given.

3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
MAXGAUSS argument.

4 Without -prob_cat_thresh argument, it will be the status of data
existence. With -prob_cat_thresh argument, it will be meeting the
condition or not. So the values of the probability are 0 or 1.

5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
(between 0. and 1.0 with GAUSSIAN smoothing).

Cheers,
Howard

On Mon Jun 29 11:31:23 2020, johnhg wrote:
> John,
>
> I'm reassigning this ticket to Howard Soh who worked on the
> development of
> the point2grid tool. He should be able to better answer these
> questions.

>
> Thanks,
> John
>
> On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > John -
> >
> > I ran it to see the usage and there are details in the usage
dialog
> > and
> > also in the XML file you pointed me to which helped to answer my
own
> > question (I think).
> > There is an additional interpolation method called GAUSSIAN which
> > applies
> > only when using the probability variable (-prob_cat_thresh 280.0).
> > So, it
> > looks like the two optional smoothing arguments -gaussian_dx and
> > -gaussian_radius are only used when the method is GAUSSIAN. IS
this
> > correct? This means that there is no smoothing except when using
> > probability variables. Is this correct?
> >
> > R/
> > John
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Monday, June 29, 2020 9:41 AM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > While the "-field" command line argument indicates that it can be
> > used one
> > or more times, in this first version of the tool, it can only be
used
> > once.
> >
> > If you specify "-field" more that once you'll get this error
message:
> > ERROR  :
> > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > variables
> > were requested.
> > ERROR  :
> >
> > We do plan to support the processing of multiple fields but it has
> > not yet
> > been implemented.
> >
> > As for the science behind the interpolation methods, it's very
> > simple. Keep
> > track of the point observations falling inside each of the target
> > grid
> > boxes. Write to the output the min, max, unweighted mean, or
median
> > of
> > those values based on the method chosen. If you only have 1 point
obs
> > falling inside each grid box, then all those methods will result
in
> > the
> > same output value. Whereas if you have many, as in the case of
GOES
> > input
> > files, then the choice of the derivation method would matter.
> >
> > Hope that helps clarify.
> >
> > Thanks,
> > John
> >
> >
> > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-Caution-url: Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > Another question I have is how is -method used in this tool? In
the
> > User's
> > > Guide, -method provides options for regridding from one grid to
> > > another
> > > such as is used in regrid_data_plane. In the case of Point2Grid,
> > > the
> > > process involved is not regridding in the way that I understand
it.
> > > If I
> > > understand it correctly, it takes point observations and
> > > distributes them
> > > on to a user provided grid using some interpolation method. The
> > > next
> > step I
> > > think is the smoothing of the field to even out irregularities
> > > which are
> > > caused by errors, noise, etc resulting in the fitting of the
> > > observed
> > > values to some function which appear to be a Gaussian function.
I
> > > guess I
> > > also need to understand what happens when the tool is used
without
> > > any of
> > > the optional arguments.
> > >
> > > Thanks.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Sent: Saturday, June 27, 2020 7:59 AM
> > > To: met_help at ucar.edu
> > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > John -
> > >
> > > Thanks for your quick response which is helpful. Appreciate your
> > > dedication so late into the evening! Can I add new variables in
the
> > config
> > > file? Any info on the science behind the interpolation methods
> > > would be
> > > appreciated, but that can wait until Monday.
> > >
> > > Take care and stay safe.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > Sent: Friday, June 26, 2020 11:06 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > the
> > > identity of the sender, and confirm the authenticity of all
links
> > contained
> > > within the message prior to copying and pasting the address to a
> > > Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > John,
> > >
> > > I see you have a question about interpolation options in the
> > > point2grid
> > > tool. This XML file contains the calls to the point2grid tool
that
> > > are
> > run
> > > for our unit tests:
> > > Caution-
> > > Caution-
> > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
> > >
> > > While most of them DO NOT use the "-method" command line option,
a
> > > few at
> > > the end do use "-method MAX" to select the maximum value falling
> > > within
> > > each grid box.
> > >
> > > Inspecting the code in this file:
> > > Caution-
> > > Caution-
> > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> > > I see that the supported options for -method are MIN, MAX,
MEDIAN,
> > > and
> > > UW_MEAN.
> > >
> > > Seems like we should add that info to the usage statement! I
wrote
> > > up
> > this
> > > GitHub issue to make sure we remember to do so before the 9.1
> > > release:
> > > Caution-Caution-Caution-https://github.com/NCAR/MET/issues/1396
> > >
> > > Hope that helps clarify.
> > >
> > > Thanks,
> > > John
> > >
> > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > >        Queue: met_help
> > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > >        Owner: Nobody
> > > >   Requestors: john.w.raby2.civ at mail.mil
> > > >       Status: new
> > > >  Ticket <Caution-Caution-Caution-url: Caution-
> > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > The optional argument for smoothing or interpolation shown in
the
> > User's
> > > > Guide is Gaussian, but are other methods available? If so, can
> > > > you
> > > provide
> > > > the names and how to specify them as an optional argument?
> > > >
> > > > Thanks.
> > > >
> > > > Stay safe.
> > > >
> > > > T/
> > > > John
> > > >
> > > > Mr. John W. Raby
> > > > U.S. Army Research Laboratory
> > > > White Sands Missile Range, NM 88002
> > > > Office:(575) 678-2004 DSN 258-2004
> > > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > > FAX (575) 678-1230 DSN 258-1230
> > > > Email: john.w.raby2.civ at mail.mil
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >





------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jun 29 14:50:03 2020

The Gaussian smoothing is available with regrid_data_plane, too. In
this case, we are interested in Gaussian smoothing without regriding.
The input and the target grid should be the same, which is the output
of point2grid.

Here is an example (assuming the point data file is generated by
pb2nc).

./point2grid ndas.20120409.t12z.prepbufr.nc \
      G212 pb2nc_TMP.nc \
      -field 'name="TMP"; level="*";' \
      -v 1

./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
      -field 'name="TMP"; level="(*,*)";' \
      -method GAUSSIAN -v 1

Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
necessary

Cheers,
Howard

On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> Howard -
>
> Thanks for your quick response with details in answer to my
question.
> Right now my interest is with non-probability fields, so your point
in
> 1. about the possibility of using Gaussian smoothing for -field
> argument is interesting.
>
> R/
> John
> ________________________________________
> From: Howard Soh via RT [met_help at ucar.edu]
> Sent: Monday, June 29, 2020 12:50 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Two layers for the interpolation with point2grid:
> - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> - layer 2: Optional, the GAUSSIAN smoothing for the probability
> variable. (the existence of valid values if the probability
threshold
> is not given)
>
> + "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which is
a
> default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> + "-method MAXGAUSS" is processed as the "-method MAX" for the layer
1
> and "-method GAUSSIAN" for the layer 2.
> + Override the interpolation method of layer 1 with "-method
<METHOD>"
>
> 1. The Gaussian smoothing at point2grid is activated and applied
into
> the probability variable only if -method GAUSSIAN or -method
MAXGAUSS
> is given as the command line argument. It's not applied into the
> variable at "-field" argument. If you want, You can do this by using
> regrid_data_plane (passing the output of point2grid as the input of
> regrid_data_plane with GAUSSIAN method).
>
> 2. -gaussian_dx and -gaussian_radius are ignored if -method GAUSSIAN
> or -method MAXGAUSS is NOT given. The default values are applied if
> not given.
>
> 3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
> MAXGAUSS argument.
>
> 4 Without -prob_cat_thresh argument, it will be the status of data
> existence. With -prob_cat_thresh argument, it will be meeting the
> condition or not. So the values of the probability are 0 or 1.
>
> 5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
> (between 0. and 1.0 with GAUSSIAN smoothing).
>
> Cheers,
> Howard
>
> On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > John,
> >
> > I'm reassigning this ticket to Howard Soh who worked on the
> > development of
> > the point2grid tool. He should be able to better answer these
> > questions.
>
> >
> > Thanks,
> > John
> >
> > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-url: Caution-
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > John -
> > >
> > > I ran it to see the usage and there are details in the usage
dialog
> > > and
> > > also in the XML file you pointed me to which helped to answer my
> > > own
> > > question (I think).
> > > There is an additional interpolation method called GAUSSIAN
which
> > > applies
> > > only when using the probability variable (-prob_cat_thresh
280.0).
> > > So, it
> > > looks like the two optional smoothing arguments -gaussian_dx and
> > > -gaussian_radius are only used when the method is GAUSSIAN. IS
this
> > > correct? This means that there is no smoothing except when using
> > > probability variables. Is this correct?
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > Sent: Monday, June 29, 2020 9:41 AM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the
> > > identity of the sender, and confirm the authenticity of all
links
> > > contained
> > > within the message prior to copying and pasting the address to a
> > > Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > John,
> > >
> > > While the "-field" command line argument indicates that it can
be
> > > used one
> > > or more times, in this first version of the tool, it can only be
> > > used
> > > once.
> > >
> > > If you specify "-field" more that once you'll get this error
> > > message:
> > > ERROR  :
> > > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > > variables
> > > were requested.
> > > ERROR  :
> > >
> > > We do plan to support the processing of multiple fields but it
has
> > > not yet
> > > been implemented.
> > >
> > > As for the science behind the interpolation methods, it's very
> > > simple. Keep
> > > track of the point observations falling inside each of the
target
> > > grid
> > > boxes. Write to the output the min, max, unweighted mean, or
median
> > > of
> > > those values based on the method chosen. If you only have 1
point
> > > obs
> > > falling inside each grid box, then all those methods will result
in
> > > the
> > > same output value. Whereas if you have many, as in the case of
GOES
> > > input
> > > files, then the choice of the derivation method would matter.
> > >
> > > Hope that helps clarify.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-url: Caution-
> > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747
>
> > > >
> > > > Another question I have is how is -method used in this tool?
In
> > > > the
> > > User's
> > > > Guide, -method provides options for regridding from one grid
to
> > > > another
> > > > such as is used in regrid_data_plane. In the case of
Point2Grid,
> > > > the
> > > > process involved is not regridding in the way that I
understand
> > > > it.
> > > > If I
> > > > understand it correctly, it takes point observations and
> > > > distributes them
> > > > on to a user provided grid using some interpolation method.
The
> > > > next
> > > step I
> > > > think is the smoothing of the field to even out irregularities
> > > > which are
> > > > caused by errors, noise, etc resulting in the fitting of the
> > > > observed
> > > > values to some function which appear to be a Gaussian
function. I
> > > > guess I
> > > > also need to understand what happens when the tool is used
> > > > without
> > > > any of
> > > > the optional arguments.
> > > >
> > > > Thanks.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > To: met_help at ucar.edu
> > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > V9.0
> > > > Point2Grid (UNCLASSIFIED)
> > > >
> > > > John -
> > > >
> > > > Thanks for your quick response which is helpful. Appreciate
your
> > > > dedication so late into the evening! Can I add new variables
in
> > > > the
> > > config
> > > > file? Any info on the science behind the interpolation methods
> > > > would be
> > > > appreciated, but that can wait until Monday.
> > > >
> > > > Take care and stay safe.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > > Point2Grid (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify
> > > the
> > > > identity of the sender, and confirm the authenticity of all
links
> > > contained
> > > > within the message prior to copying and pasting the address to
a
> > > > Web
> > > > browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > John,
> > > >
> > > > I see you have a question about interpolation options in the
> > > > point2grid
> > > > tool. This XML file contains the calls to the point2grid tool
> > > > that
> > > > are
> > > run
> > > > for our unit tests:
> > > > Caution-
> > > > Caution-
> > > Caution-
> > >
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit_point2grid.xml
> > > >
> > > > While most of them DO NOT use the "-method" command line
option,
> > > > a
> > > > few at
> > > > the end do use "-method MAX" to select the maximum value
falling
> > > > within
> > > > each grid box.
> > > >
> > > > Inspecting the code in this file:
> > > > Caution-
> > > > Caution-
> > > Caution-
> > >
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools/other/point2grid/point2grid.cc
> > > > I see that the supported options for -method are MIN, MAX,
> > > > MEDIAN,
> > > > and
> > > > UW_MEAN.
> > > >
> > > > Seems like we should add that info to the usage statement! I
> > > > wrote
> > > > up
> > > this
> > > > GitHub issue to make sure we remember to do so before the 9.1
> > > > release:
> > > > Caution-Caution-Caution-
https://github.com/NCAR/MET/issues/1396
> > > >
> > > > Hope that helps clarify.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > >        Queue: met_help
> > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > >        Owner: Nobody
> > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > >       Status: new
> > > > >  Ticket <Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > The optional argument for smoothing or interpolation shown
in
> > > > > the
> > > User's
> > > > > Guide is Gaussian, but are other methods available? If so,
can
> > > > > you
> > > > provide
> > > > > the names and how to specify them as an optional argument?
> > > > >
> > > > > Thanks.
> > > > >
> > > > > Stay safe.
> > > > >
> > > > > T/
> > > > > John
> > > > >
> > > > > Mr. John W. Raby
> > > > > U.S. Army Research Laboratory
> > > > > White Sands Missile Range, NM 88002
> > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > > > FAX (575) 678-1230 DSN 258-1230
> > > > > Email: john.w.raby2.civ at mail.mil
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Tue Jun 30 06:58:02 2020

CLASSIFICATION: UNCLASSIFIED

Howard -

Thanks for tailoring the command for my particular interests. Just
curious
about your regrid_data_plane command. You provided the input file from
point2grid (pb2nc_TMP.nc ) as the input_filename to regrid_data_plane
and then
you provide the name of the file to be output from regrid_data_plane
(pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't provide the
to_grid. Is
this because it knows that the grids of the input and output files are
the
same and thus ignores the missing to_grid argument?

R/
John

-----Original Message-----
From: Howard Soh via RT <met_help at ucar.edu>
Sent: Monday, June 29, 2020 2:50 PM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

The Gaussian smoothing is available with regrid_data_plane, too. In
this case,
we are interested in Gaussian smoothing without regriding. The input
and the
target grid should be the same, which is the output of point2grid.

Here is an example (assuming the point data file is generated by
pb2nc).

./point2grid ndas.20120409.t12z.prepbufr.nc \
      G212 pb2nc_TMP.nc \
      -field 'name="TMP"; level="*";' \
      -v 1

./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
      -field 'name="TMP"; level="(*,*)";' \
      -method GAUSSIAN -v 1

Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
necessary

Cheers,
Howard

On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> Howard -
>
> Thanks for your quick response with details in answer to my
question.
> Right now my interest is with non-probability fields, so your point
in
> 1. about the possibility of using Gaussian smoothing for -field
> argument is interesting.
>
> R/
> John
> ________________________________________
> From: Howard Soh via RT [met_help at ucar.edu]
> Sent: Monday, June 29, 2020 12:50 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Two layers for the interpolation with point2grid:
> - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> - layer 2: Optional, the GAUSSIAN smoothing for the probability
> variable. (the existence of valid values if the probability
threshold
> is not given)
>
> + "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which is
a
> default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> + "-method MAXGAUSS" is processed as the "-method MAX" for the layer
1
> and "-method GAUSSIAN" for the layer 2.
> + Override the interpolation method of layer 1 with "-method
<METHOD>"
>
> 1. The Gaussian smoothing at point2grid is activated and applied
into
> the probability variable only if -method GAUSSIAN or -method
MAXGAUSS
> is given as the command line argument. It's not applied into the
> variable at "-field" argument. If you want, You can do this by using
> regrid_data_plane (passing the output of point2grid as the input of
> regrid_data_plane with GAUSSIAN method).
>
> 2. -gaussian_dx and -gaussian_radius are ignored if -method GAUSSIAN
> or -method MAXGAUSS is NOT given. The default values are applied if
> not given.
>
> 3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
> MAXGAUSS argument.
>
> 4 Without -prob_cat_thresh argument, it will be the status of data
> existence. With -prob_cat_thresh argument, it will be meeting the
> condition or not. So the values of the probability are 0 or 1.
>
> 5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
> (between 0. and 1.0 with GAUSSIAN smoothing).
>
> Cheers,
> Howard
>
> On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > John,
> >
> > I'm reassigning this ticket to Howard Soh who worked on the
> > development of the point2grid tool. He should be able to better
> > answer these questions.
>
> >
> > Thanks,
> > John
> >
> > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-Caution-url: Caution-
> > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747
>
> > >
> > > John -
> > >
> > > I ran it to see the usage and there are details in the usage
> > > dialog and also in the XML file you pointed me to which helped
to
> > > answer my own question (I think).
> > > There is an additional interpolation method called GAUSSIAN
which
> > > applies only when using the probability variable (-
prob_cat_thresh
> > > 280.0).
> > > So, it
> > > looks like the two optional smoothing arguments -gaussian_dx and
> > > -gaussian_radius are only used when the method is GAUSSIAN. IS
> > > this correct? This means that there is no smoothing except when
> > > using probability variables. Is this correct?
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > Sent: Monday, June 29, 2020 9:41 AM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > V9.0 Point2Grid (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
> > > pasting the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > John,
> > >
> > > While the "-field" command line argument indicates that it can
be
> > > used one or more times, in this first version of the tool, it
can
> > > only be used once.
> > >
> > > If you specify "-field" more that once you'll get this error
> > > message:
> > > ERROR  :
> > > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > > variables
> > > were requested.
> > > ERROR  :
> > >
> > > We do plan to support the processing of multiple fields but it
has
> > > not yet been implemented.
> > >
> > > As for the science behind the interpolation methods, it's very
> > > simple. Keep track of the point observations falling inside each
> > > of the target grid boxes. Write to the output the min, max,
> > > unweighted mean, or median of those values based on the method
> > > chosen. If you only have 1 point obs falling inside each grid
box,
> > > then all those methods will result in the same output value.
> > > Whereas if you have many, as in the case of GOES input files,
then
> > > the choice of the derivation method would matter.
> > >
> > > Hope that helps clarify.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-url: Caution-
> > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > 95747 >
> > > >
> > > > Another question I have is how is -method used in this tool?
In
> > > > the
> > > User's
> > > > Guide, -method provides options for regridding from one grid
to
> > > > another such as is used in regrid_data_plane. In the case of
> > > > Point2Grid, the process involved is not regridding in the way
> > > > that I understand it.
> > > > If I
> > > > understand it correctly, it takes point observations and
> > > > distributes them on to a user provided grid using some
> > > > interpolation method. The next
> > > step I
> > > > think is the smoothing of the field to even out irregularities
> > > > which are caused by errors, noise, etc resulting in the
fitting
> > > > of the observed values to some function which appear to be a
> > > > Gaussian function. I guess I also need to understand what
> > > > happens when the tool is used without any of the optional
> > > > arguments.
> > > >
> > > > Thanks.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > To: met_help at ucar.edu
> > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > V9.0
> > > > Point2Grid (UNCLASSIFIED)
> > > >
> > > > John -
> > > >
> > > > Thanks for your quick response which is helpful. Appreciate
your
> > > > dedication so late into the evening! Can I add new variables
in
> > > > the
> > > config
> > > > file? Any info on the science behind the interpolation methods
> > > > would be appreciated, but that can wait until Monday.
> > > >
> > > > Take care and stay safe.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > > Point2Grid (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify
> > > the
> > > > identity of the sender, and confirm the authenticity of all
> > > > links
> > > contained
> > > > within the message prior to copying and pasting the address to
a
> > > > Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > John,
> > > >
> > > > I see you have a question about interpolation options in the
> > > > point2grid tool. This XML file contains the calls to the
> > > > point2grid tool that are
> > > run
> > > > for our unit tests:
> > > > Caution-
> > > > Caution-
> > > Caution-
> > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit
> > > _point2grid.xml
> > > >
> > > > While most of them DO NOT use the "-method" command line
option,
> > > > a few at the end do use "-method MAX" to select the maximum
> > > > value falling within each grid box.
> > > >
> > > > Inspecting the code in this file:
> > > > Caution-
> > > > Caution-
> > > Caution-
> > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools
> > > /other/point2grid/point2grid.cc
> > > > I see that the supported options for -method are MIN, MAX,
> > > > MEDIAN, and UW_MEAN.
> > > >
> > > > Seems like we should add that info to the usage statement! I
> > > > wrote up
> > > this
> > > > GitHub issue to make sure we remember to do so before the 9.1
> > > > release:
> > > > Caution-Caution-Caution-Caution-
https://github.com/NCAR/MET/issu
> > > > es/1396
> > > >
> > > > Hope that helps clarify.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > >        Queue: met_help
> > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > >        Owner: Nobody
> > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > >       Status: new
> > > > >  Ticket <Caution-Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
> > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > The optional argument for smoothing or interpolation shown
in
> > > > > the
> > > User's
> > > > > Guide is Gaussian, but are other methods available? If so,
can
> > > > > you
> > > > provide
> > > > > the names and how to specify them as an optional argument?
> > > > >
> > > > > Thanks.
> > > > >
> > > > > Stay safe.
> > > > >
> > > > > T/
> > > > > John
> > > > >
> > > > > Mr. John W. Raby
> > > > > U.S. Army Research Laboratory
> > > > > White Sands Missile Range, NM 88002
> > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > Teleworking: Contact Admin Specialist at (410) 278-6974 FAX
> > > > > (575) 678-1230 DSN 258-1230
> > > > > Email: john.w.raby2.civ at mail.mil
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >



CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jul 06 23:38:06 2020

Sorry for the late response.

The input and to_grid must be the same for Gaussian smoothing and
regrid_data_plane shows the usage message if the to_grid is missing.
There are two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.

./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
-field 'name="TMP"; level="(*,*)";' \
-method GAUSSIAN -v 1

Cheers,
Howard



On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> Thanks for tailoring the command for my particular interests. Just
> curious
> about your regrid_data_plane command. You provided the input file
from
> point2grid (pb2nc_TMP.nc ) as the input_filename to
regrid_data_plane
> and then
> you provide the name of the file to be output from regrid_data_plane
> (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't provide the
> to_grid. Is
> this because it knows that the grids of the input and output files
are
> the
> same and thus ignores the missing to_grid argument?
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, June 29, 2020 2:50 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> The Gaussian smoothing is available with regrid_data_plane, too. In
> this case,
> we are interested in Gaussian smoothing without regriding. The input
> and the
> target grid should be the same, which is the output of point2grid.
>
> Here is an example (assuming the point data file is generated by
> pb2nc).
>
> ./point2grid ndas.20120409.t12z.prepbufr.nc \
>       G212 pb2nc_TMP.nc \
>       -field 'name="TMP"; level="*";' \
>       -v 1
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
>       -field 'name="TMP"; level="(*,*)";' \
>       -method GAUSSIAN -v 1
>
> Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> necessary
>
> Cheers,
> Howard
>
> On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > Howard -
> >
> > Thanks for your quick response with details in answer to my
question.
> > Right now my interest is with non-probability fields, so your
point
> > in
> > 1. about the possibility of using Gaussian smoothing for -field
> > argument is interesting.
> >
> > R/
> > John
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Monday, June 29, 2020 12:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Two layers for the interpolation with point2grid:
> > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > variable. (the existence of valid values if the probability
threshold
> > is not given)
> >
> > + "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which
is
> > a
> > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > + "-method MAXGAUSS" is processed as the "-method MAX" for the
layer
> > 1
> > and "-method GAUSSIAN" for the layer 2.
> > + Override the interpolation method of layer 1 with "-method
> > <METHOD>"
> >
> > 1. The Gaussian smoothing at point2grid is activated and applied
into
> > the probability variable only if -method GAUSSIAN or -method
MAXGAUSS
> > is given as the command line argument. It's not applied into the
> > variable at "-field" argument. If you want, You can do this by
using
> > regrid_data_plane (passing the output of point2grid as the input
of
> > regrid_data_plane with GAUSSIAN method).
> >
> > 2. -gaussian_dx and -gaussian_radius are ignored if -method
GAUSSIAN
> > or -method MAXGAUSS is NOT given. The default values are applied
if
> > not given.
> >
> > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
> > MAXGAUSS argument.
> >
> > 4 Without -prob_cat_thresh argument, it will be the status of data
> > existence. With -prob_cat_thresh argument, it will be meeting the
> > condition or not. So the values of the probability are 0 or 1.
> >
> > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
> > (between 0. and 1.0 with GAUSSIAN smoothing).
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > John,
> > >
> > > I'm reassigning this ticket to Howard Soh who worked on the
> > > development of the point2grid tool. He should be able to better
> > > answer these questions.
> >
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-url: Caution-
> > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > >
> > > > John -
> > > >
> > > > I ran it to see the usage and there are details in the usage
> > > > dialog and also in the XML file you pointed me to which helped
to
> > > > answer my own question (I think).
> > > > There is an additional interpolation method called GAUSSIAN
which
> > > > applies only when using the probability variable (-
> > > > prob_cat_thresh
> > > > 280.0).
> > > > So, it
> > > > looks like the two optional smoothing arguments -gaussian_dx
and
> > > > -gaussian_radius are only used when the method is GAUSSIAN. IS
> > > > this correct? This means that there is no smoothing except
when
> > > > using probability variables. Is this correct?
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > John,
> > > >
> > > > While the "-field" command line argument indicates that it can
be
> > > > used one or more times, in this first version of the tool, it
can
> > > > only be used once.
> > > >
> > > > If you specify "-field" more that once you'll get this error
> > > > message:
> > > > ERROR  :
> > > > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > > > variables
> > > > were requested.
> > > > ERROR  :
> > > >
> > > > We do plan to support the processing of multiple fields but it
> > > > has
> > > > not yet been implemented.
> > > >
> > > > As for the science behind the interpolation methods, it's very
> > > > simple. Keep track of the point observations falling inside
each
> > > > of the target grid boxes. Write to the output the min, max,
> > > > unweighted mean, or median of those values based on the method
> > > > chosen. If you only have 1 point obs falling inside each grid
> > > > box,
> > > > then all those methods will result in the same output value.
> > > > Whereas if you have many, as in the case of GOES input files,
> > > > then
> > > > the choice of the derivation method would matter.
> > > >
> > > > Hope that helps clarify.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > > 95747 >
> > > > >
> > > > > Another question I have is how is -method used in this tool?
In
> > > > > the
> > > > User's
> > > > > Guide, -method provides options for regridding from one grid
to
> > > > > another such as is used in regrid_data_plane. In the case of
> > > > > Point2Grid, the process involved is not regridding in the
way
> > > > > that I understand it.
> > > > > If I
> > > > > understand it correctly, it takes point observations and
> > > > > distributes them on to a user provided grid using some
> > > > > interpolation method. The next
> > > > step I
> > > > > think is the smoothing of the field to even out
irregularities
> > > > > which are caused by errors, noise, etc resulting in the
fitting
> > > > > of the observed values to some function which appear to be a
> > > > > Gaussian function. I guess I also need to understand what
> > > > > happens when the tool is used without any of the optional
> > > > > arguments.
> > > > >
> > > > > Thanks.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > To: met_help at ucar.edu
> > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0
> > > > > Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > John -
> > > > >
> > > > > Thanks for your quick response which is helpful. Appreciate
> > > > > your
> > > > > dedication so late into the evening! Can I add new variables
in
> > > > > the
> > > > config
> > > > > file? Any info on the science behind the interpolation
methods
> > > > > would be appreciated, but that can wait until Monday.
> > > > >
> > > > > Take care and stay safe.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify
> > > > the
> > > > > identity of the sender, and confirm the authenticity of all
> > > > > links
> > > > contained
> > > > > within the message prior to copying and pasting the address
to
> > > > > a
> > > > > Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > I see you have a question about interpolation options in the
> > > > > point2grid tool. This XML file contains the calls to the
> > > > > point2grid tool that are
> > > > run
> > > > > for our unit tests:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/unit
> > > > _point2grid.xml
> > > > >
> > > > > While most of them DO NOT use the "-method" command line
> > > > > option,
> > > > > a few at the end do use "-method MAX" to select the maximum
> > > > > value falling within each grid box.
> > > > >
> > > > > Inspecting the code in this file:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/tools
> > > > /other/point2grid/point2grid.cc
> > > > > I see that the supported options for -method are MIN, MAX,
> > > > > MEDIAN, and UW_MEAN.
> > > > >
> > > > > Seems like we should add that info to the usage statement! I
> > > > > wrote up
> > > > this
> > > > > GitHub issue to make sure we remember to do so before the
9.1
> > > > > release:
> > > > > Caution-Caution-Caution-Caution-
> > > > > https://github.com/NCAR/MET/issu
> > > > > es/1396
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > >        Queue: met_help
> > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >        Owner: Nobody
> > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > >       Status: new
> > > > > >  Ticket <Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > The optional argument for smoothing or interpolation shown
in
> > > > > > the
> > > > User's
> > > > > > Guide is Gaussian, but are other methods available? If so,
> > > > > > can
> > > > > > you
> > > > > provide
> > > > > > the names and how to specify them as an optional argument?
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > Stay safe.
> > > > > >
> > > > > > T/
> > > > > > John
> > > > > >
> > > > > > Mr. John W. Raby
> > > > > > U.S. Army Research Laboratory
> > > > > > White Sands Missile Range, NM 88002
> > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
FAX
> > > > > > (575) 678-1230 DSN 258-1230
> > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
>
>
>
> CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Tue Jul 07 06:48:32 2020

CLASSIFICATION: UNCLASSIFIED

Howard -

No problem. You are correct. I see now where your original command did
have
the to_grid specified. My oversight. I guess my eyes were stressing
some to
separate out that third ".nc" in the command.

Thanks.

R/
John

-----Original Message-----
From: Howard Soh via RT <met_help at ucar.edu>
Sent: Monday, July 6, 2020 11:38 PM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

Sorry for the late response.

The input and to_grid must be the same for Gaussian smoothing and
regrid_data_plane shows the usage message if the to_grid is missing.
There are
two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.

./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
-field
'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1

Cheers,
Howard



On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> Thanks for tailoring the command for my particular interests. Just
> curious about your regrid_data_plane command. You provided the input
> file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> regrid_data_plane and then you provide the name of the file to be
> output from regrid_data_plane (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc),
> but you didn't provide the to_grid. Is this because it knows that
the
> grids of the input and output files are the same and thus ignores
the
> missing to_grid argument?
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, June 29, 2020 2:50 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> The Gaussian smoothing is available with regrid_data_plane, too. In
> this case, we are interested in Gaussian smoothing without
regriding.
> The input and the target grid should be the same, which is the
output
> of point2grid.
>
> Here is an example (assuming the point data file is generated by
> pb2nc).
>
> ./point2grid ndas.20120409.t12z.prepbufr.nc \
>       G212 pb2nc_TMP.nc \
>       -field 'name="TMP"; level="*";' \
>       -v 1
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
>       -field 'name="TMP"; level="(*,*)";' \
>       -method GAUSSIAN -v 1
>
> Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> necessary
>
> Cheers,
> Howard
>
> On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > Howard -
> >
> > Thanks for your quick response with details in answer to my
question.
> > Right now my interest is with non-probability fields, so your
point
> > in 1. about the possibility of using Gaussian smoothing for -field
> > argument is interesting.
> >
> > R/
> > John
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Monday, June 29, 2020 12:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the identity of the sender, and confirm the authenticity of
> > all links contained within the message prior to copying and
pasting
> > the address to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Two layers for the interpolation with point2grid:
> > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > variable. (the existence of valid values if the probability
> > threshold is not given)
> >
> > + "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which
is
> > a
> > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > + "-method MAXGAUSS" is processed as the "-method MAX" for the
layer
> > 1
> > and "-method GAUSSIAN" for the layer 2.
> > + Override the interpolation method of layer 1 with "-method
> > <METHOD>"
> >
> > 1. The Gaussian smoothing at point2grid is activated and applied
> > into the probability variable only if -method GAUSSIAN or -method
> > MAXGAUSS is given as the command line argument. It's not applied
> > into the variable at "-field" argument. If you want, You can do
this
> > by using regrid_data_plane (passing the output of point2grid as
the
> > input of regrid_data_plane with GAUSSIAN method).
> >
> > 2. -gaussian_dx and -gaussian_radius are ignored if -method
GAUSSIAN
> > or -method MAXGAUSS is NOT given. The default values are applied
if
> > not given.
> >
> > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
> > MAXGAUSS argument.
> >
> > 4 Without -prob_cat_thresh argument, it will be the status of data
> > existence. With -prob_cat_thresh argument, it will be meeting the
> > condition or not. So the values of the probability are 0 or 1.
> >
> > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
> > (between 0. and 1.0 with GAUSSIAN smoothing).
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > John,
> > >
> > > I'm reassigning this ticket to Howard Soh who worked on the
> > > development of the point2grid tool. He should be able to better
> > > answer these questions.
> >
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > d=95747 >
> > > >
> > > > John -
> > > >
> > > > I ran it to see the usage and there are details in the usage
> > > > dialog and also in the XML file you pointed me to which helped
> > > > to answer my own question (I think).
> > > > There is an additional interpolation method called GAUSSIAN
> > > > which applies only when using the probability variable (-
> > > > prob_cat_thresh 280.0).
> > > > So, it
> > > > looks like the two optional smoothing arguments -gaussian_dx
and
> > > > -gaussian_radius are only used when the method is GAUSSIAN. IS
> > > > this correct? This means that there is no smoothing except
when
> > > > using probability variables. Is this correct?
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > John,
> > > >
> > > > While the "-field" command line argument indicates that it can
> > > > be used one or more times, in this first version of the tool,
it
> > > > can only be used once.
> > > >
> > > > If you specify "-field" more that once you'll get this error
> > > > message:
> > > > ERROR  :
> > > > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > > > variables
> > > > were requested.
> > > > ERROR  :
> > > >
> > > > We do plan to support the processing of multiple fields but it
> > > > has not yet been implemented.
> > > >
> > > > As for the science behind the interpolation methods, it's very
> > > > simple. Keep track of the point observations falling inside
each
> > > > of the target grid boxes. Write to the output the min, max,
> > > > unweighted mean, or median of those values based on the method
> > > > chosen. If you only have 1 point obs falling inside each grid
> > > > box, then all those methods will result in the same output
> > > > value.
> > > > Whereas if you have many, as in the case of GOES input files,
> > > > then the choice of the derivation method would matter.
> > > >
> > > > Hope that helps clarify.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
> > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > > 95747 >
> > > > >
> > > > > Another question I have is how is -method used in this tool?
> > > > > In the
> > > > User's
> > > > > Guide, -method provides options for regridding from one grid
> > > > > to another such as is used in regrid_data_plane. In the case
> > > > > of Point2Grid, the process involved is not regridding in the
> > > > > way that I understand it.
> > > > > If I
> > > > > understand it correctly, it takes point observations and
> > > > > distributes them on to a user provided grid using some
> > > > > interpolation method. The next
> > > > step I
> > > > > think is the smoothing of the field to even out
irregularities
> > > > > which are caused by errors, noise, etc resulting in the
> > > > > fitting of the observed values to some function which appear
> > > > > to be a Gaussian function. I guess I also need to understand
> > > > > what happens when the tool is used without any of the
optional
> > > > > arguments.
> > > > >
> > > > > Thanks.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > To: met_help at ucar.edu
> > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0
> > > > > Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > John -
> > > > >
> > > > > Thanks for your quick response which is helpful. Appreciate
> > > > > your dedication so late into the evening! Can I add new
> > > > > variables in the
> > > > config
> > > > > file? Any info on the science behind the interpolation
methods
> > > > > would be appreciated, but that can wait until Monday.
> > > > >
> > > > > Take care and stay safe.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
> > > > > Please verify
> > > > the
> > > > > identity of the sender, and confirm the authenticity of all
> > > > > links
> > > > contained
> > > > > within the message prior to copying and pasting the address
to
> > > > > a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > I see you have a question about interpolation options in the
> > > > > point2grid tool. This XML file contains the calls to the
> > > > > point2grid tool that are
> > > > run
> > > > > for our unit tests:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/un
> > > > it
> > > > _point2grid.xml
> > > > >
> > > > > While most of them DO NOT use the "-method" command line
> > > > > option, a few at the end do use "-method MAX" to select the
> > > > > maximum value falling within each grid box.
> > > > >
> > > > > Inspecting the code in this file:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/too
> > > > ls
> > > > /other/point2grid/point2grid.cc
> > > > > I see that the supported options for -method are MIN, MAX,
> > > > > MEDIAN, and UW_MEAN.
> > > > >
> > > > > Seems like we should add that info to the usage statement! I
> > > > > wrote up
> > > > this
> > > > > GitHub issue to make sure we remember to do so before the
9.1
> > > > > release:
> > > > > Caution-Caution-Caution-Caution-
> > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > es/1396
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > >        Queue: met_help
> > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >        Owner: Nobody
> > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > >       Status: new
> > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-url:
> > > > > > Caution-
> > > > > Caution-Caution-
> > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > ?id=95747
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > The optional argument for smoothing or interpolation shown
> > > > > > in the
> > > > User's
> > > > > > Guide is Gaussian, but are other methods available? If so,
> > > > > > can you
> > > > > provide
> > > > > > the names and how to specify them as an optional argument?
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > Stay safe.
> > > > > >
> > > > > > T/
> > > > > > John
> > > > > >
> > > > > > Mr. John W. Raby
> > > > > > U.S. Army Research Laboratory White Sands Missile Range,
NM
> > > > > > 88002
> > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
FAX
> > > > > > (575) 678-1230 DSN 258-1230
> > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
>
>
>
> CLASSIFICATION: UNCLASSIFIED



CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 13 12:55:31 2020

CLASSIFICATION: UNCLASSIFIED

Howard -

I tried running MET V9.0 regrid_data_plane following your example
below, but
I'm getting and error:
ERROR  :
ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
regridding, only
interpolating.

The following is the command I used to run it:
[~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
 -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gauss_test
-v 9

Any suggestions on how to perform Gaussian smoothing on the output TMP
field
from point2grid?

Thanks.

R/
John

-----Original Message-----
From: Howard Soh via RT <met_help at ucar.edu>
Sent: Monday, July 6, 2020 11:38 PM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

Sorry for the late response.

The input and to_grid must be the same for Gaussian smoothing and
regrid_data_plane shows the usage message if the to_grid is missing.
There are
two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.

./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
-field
'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1

Cheers,
Howard



On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> Thanks for tailoring the command for my particular interests. Just
> curious about your regrid_data_plane command. You provided the input
> file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> regrid_data_plane and then you provide the name of the file to be
> output from regrid_data_plane (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc),
> but you didn't provide the to_grid. Is this because it knows that
the
> grids of the input and output files are the same and thus ignores
the
> missing to_grid argument?
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, June 29, 2020 2:50 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> The Gaussian smoothing is available with regrid_data_plane, too. In
> this case, we are interested in Gaussian smoothing without
regriding.
> The input and the target grid should be the same, which is the
output
> of point2grid.
>
> Here is an example (assuming the point data file is generated by
> pb2nc).
>
> ./point2grid ndas.20120409.t12z.prepbufr.nc \
>       G212 pb2nc_TMP.nc \
>       -field 'name="TMP"; level="*";' \
>       -v 1
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
>       -field 'name="TMP"; level="(*,*)";' \
>       -method GAUSSIAN -v 1
>
> Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> necessary
>
> Cheers,
> Howard
>
> On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > Howard -
> >
> > Thanks for your quick response with details in answer to my
question.
> > Right now my interest is with non-probability fields, so your
point
> > in 1. about the possibility of using Gaussian smoothing for -field
> > argument is interesting.
> >
> > R/
> > John
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Monday, June 29, 2020 12:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the identity of the sender, and confirm the authenticity of
> > all links contained within the message prior to copying and
pasting
> > the address to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Two layers for the interpolation with point2grid:
> > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > variable. (the existence of valid values if the probability
> > threshold is not given)
> >
> > + "-method GAUSSIAN" is processed as the "-method UW_MEAN" (which
is
> > a
> > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > + "-method MAXGAUSS" is processed as the "-method MAX" for the
layer
> > 1
> > and "-method GAUSSIAN" for the layer 2.
> > + Override the interpolation method of layer 1 with "-method
> > <METHOD>"
> >
> > 1. The Gaussian smoothing at point2grid is activated and applied
> > into the probability variable only if -method GAUSSIAN or -method
> > MAXGAUSS is given as the command line argument. It's not applied
> > into the variable at "-field" argument. If you want, You can do
this
> > by using regrid_data_plane (passing the output of point2grid as
the
> > input of regrid_data_plane with GAUSSIAN method).
> >
> > 2. -gaussian_dx and -gaussian_radius are ignored if -method
GAUSSIAN
> > or -method MAXGAUSS is NOT given. The default values are applied
if
> > not given.
> >
> > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or -method
> > MAXGAUSS argument.
> >
> > 4 Without -prob_cat_thresh argument, it will be the status of data
> > existence. With -prob_cat_thresh argument, it will be meeting the
> > condition or not. So the values of the probability are 0 or 1.
> >
> > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is requested
> > (between 0. and 1.0 with GAUSSIAN smoothing).
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > John,
> > >
> > > I'm reassigning this ticket to Howard Soh who worked on the
> > > development of the point2grid tool. He should be able to better
> > > answer these questions.
> >
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > d=95747 >
> > > >
> > > > John -
> > > >
> > > > I ran it to see the usage and there are details in the usage
> > > > dialog and also in the XML file you pointed me to which helped
> > > > to answer my own question (I think).
> > > > There is an additional interpolation method called GAUSSIAN
> > > > which applies only when using the probability variable (-
> > > > prob_cat_thresh 280.0).
> > > > So, it
> > > > looks like the two optional smoothing arguments -gaussian_dx
and
> > > > -gaussian_radius are only used when the method is GAUSSIAN. IS
> > > > this correct? This means that there is no smoothing except
when
> > > > using probability variables. Is this correct?
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > John,
> > > >
> > > > While the "-field" command line argument indicates that it can
> > > > be used one or more times, in this first version of the tool,
it
> > > > can only be used once.
> > > >
> > > > If you specify "-field" more that once you'll get this error
> > > > message:
> > > > ERROR  :
> > > > ERROR  : process_point_file() -> Support only 1 variable. "2"
> > > > variables
> > > > were requested.
> > > > ERROR  :
> > > >
> > > > We do plan to support the processing of multiple fields but it
> > > > has not yet been implemented.
> > > >
> > > > As for the science behind the interpolation methods, it's very
> > > > simple. Keep track of the point observations falling inside
each
> > > > of the target grid boxes. Write to the output the min, max,
> > > > unweighted mean, or median of those values based on the method
> > > > chosen. If you only have 1 point obs falling inside each grid
> > > > box, then all those methods will result in the same output
> > > > value.
> > > > Whereas if you have many, as in the case of GOES input files,
> > > > then the choice of the derivation method would matter.
> > > >
> > > > Hope that helps clarify.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > Caution-Caution-
> > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > > 95747 >
> > > > >
> > > > > Another question I have is how is -method used in this tool?
> > > > > In the
> > > > User's
> > > > > Guide, -method provides options for regridding from one grid
> > > > > to another such as is used in regrid_data_plane. In the case
> > > > > of Point2Grid, the process involved is not regridding in the
> > > > > way that I understand it.
> > > > > If I
> > > > > understand it correctly, it takes point observations and
> > > > > distributes them on to a user provided grid using some
> > > > > interpolation method. The next
> > > > step I
> > > > > think is the smoothing of the field to even out
irregularities
> > > > > which are caused by errors, noise, etc resulting in the
> > > > > fitting of the observed values to some function which appear
> > > > > to be a Gaussian function. I guess I also need to understand
> > > > > what happens when the tool is used without any of the
optional
> > > > > arguments.
> > > > >
> > > > > Thanks.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > To: met_help at ucar.edu
> > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0
> > > > > Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > John -
> > > > >
> > > > > Thanks for your quick response which is helpful. Appreciate
> > > > > your dedication so late into the evening! Can I add new
> > > > > variables in the
> > > > config
> > > > > file? Any info on the science behind the interpolation
methods
> > > > > would be appreciated, but that can wait until Monday.
> > > > >
> > > > > Take care and stay safe.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
> > > > > Please verify
> > > > the
> > > > > identity of the sender, and confirm the authenticity of all
> > > > > links
> > > > contained
> > > > > within the message prior to copying and pasting the address
to
> > > > > a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > I see you have a question about interpolation options in the
> > > > > point2grid tool. This XML file contains the calls to the
> > > > > point2grid tool that are
> > > > run
> > > > > for our unit tests:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/un
> > > > it
> > > > _point2grid.xml
> > > > >
> > > > > While most of them DO NOT use the "-method" command line
> > > > > option, a few at the end do use "-method MAX" to select the
> > > > > maximum value falling within each grid box.
> > > > >
> > > > > Inspecting the code in this file:
> > > > > Caution-
> > > > > Caution-
> > > > Caution-
> > > > Caution-
> > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/too
> > > > ls
> > > > /other/point2grid/point2grid.cc
> > > > > I see that the supported options for -method are MIN, MAX,
> > > > > MEDIAN, and UW_MEAN.
> > > > >
> > > > > Seems like we should add that info to the usage statement! I
> > > > > wrote up
> > > > this
> > > > > GitHub issue to make sure we remember to do so before the
9.1
> > > > > release:
> > > > > Caution-Caution-Caution-Caution-
> > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > es/1396
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > >        Queue: met_help
> > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >        Owner: Nobody
> > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > >       Status: new
> > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-url:
> > > > > > Caution-
> > > > > Caution-Caution-
> > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > ?id=95747
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > The optional argument for smoothing or interpolation shown
> > > > > > in the
> > > > User's
> > > > > > Guide is Gaussian, but are other methods available? If so,
> > > > > > can you
> > > > > provide
> > > > > > the names and how to specify them as an optional argument?
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > Stay safe.
> > > > > >
> > > > > > T/
> > > > > > John
> > > > > >
> > > > > > Mr. John W. Raby
> > > > > > U.S. Army Research Laboratory White Sands Missile Range,
NM
> > > > > > 88002
> > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
FAX
> > > > > > (575) 678-1230 DSN 258-1230
> > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
>
>
>
> CLASSIFICATION: UNCLASSIFIED



CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jul 13 13:13:39 2020

I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS and
GAUSSIAN method is only used for the interpolation. Please set "-
method MAXGAUSE". The from_grid and to_grid are the same, there is no
difference with additional MAX interpolation.

Cheers,
Howard


On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> I tried running MET V9.0 regrid_data_plane following your example
> below, but
> I'm getting and error:
> ERROR  :
> ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> regridding, only
> interpolating.
>
> The following is the command I used to run it:
>  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
>   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gauss_test
> -v 9
>
> Any suggestions on how to perform Gaussian smoothing on the output
TMP
> field
> from point2grid?
>
> Thanks.
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, July 6, 2020 11:38 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> Sorry for the late response.
>
> The input and to_grid must be the same for Gaussian smoothing and
> regrid_data_plane shows the usage message if the to_grid is missing.
> There are
> two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
> -field
> 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
>
> Cheers,
> Howard
>
>
>
> On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > Thanks for tailoring the command for my particular interests. Just
> > curious about your regrid_data_plane command. You provided the
input
> > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > regrid_data_plane and then you provide the name of the file to be
> > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > but you didn't provide the to_grid. Is this because it knows that
the
> > grids of the input and output files are the same and thus ignores
the
> > missing to_grid argument?
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, June 29, 2020 2:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > this case, we are interested in Gaussian smoothing without
regriding.
> > The input and the target grid should be the same, which is the
output
> > of point2grid.
> >
> > Here is an example (assuming the point data file is generated by
> > pb2nc).
> >
> > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> >       G212 pb2nc_TMP.nc \
> >       -field 'name="TMP"; level="*";' \
> >       -v 1
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> >       -field 'name="TMP"; level="(*,*)";' \
> >       -method GAUSSIAN -v 1
> >
> > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> > necessary
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > Howard -
> > >
> > > Thanks for your quick response with details in answer to my
> > > question.
> > > Right now my interest is with non-probability fields, so your
point
> > > in 1. about the possibility of using Gaussian smoothing for
-field
> > > argument is interesting.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Monday, June 29, 2020 12:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
pasting
> > > the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Two layers for the interpolation with point2grid:
> > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > > variable. (the existence of valid values if the probability
> > > threshold is not given)
> > >
> > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > is
> > > a
> > > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > layer
> > > 1
> > > and "-method GAUSSIAN" for the layer 2.
> > > + Override the interpolation method of layer 1 with "-method
> > > <METHOD>"
> > >
> > > 1. The Gaussian smoothing at point2grid is activated and applied
> > > into the probability variable only if -method GAUSSIAN or
-method
> > > MAXGAUSS is given as the command line argument. It's not applied
> > > into the variable at "-field" argument. If you want, You can do
> > > this
> > > by using regrid_data_plane (passing the output of point2grid as
the
> > > input of regrid_data_plane with GAUSSIAN method).
> > >
> > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > GAUSSIAN
> > > or -method MAXGAUSS is NOT given. The default values are applied
if
> > > not given.
> > >
> > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > MAXGAUSS argument.
> > >
> > > 4 Without -prob_cat_thresh argument, it will be the status of
data
> > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > condition or not. So the values of the probability are 0 or 1.
> > >
> > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
requested
> > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > John,
> > > >
> > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > development of the point2grid tool. He should be able to
better
> > > > answer these questions.
> > >
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > d=95747 >
> > > > >
> > > > > John -
> > > > >
> > > > > I ran it to see the usage and there are details in the usage
> > > > > dialog and also in the XML file you pointed me to which
helped
> > > > > to answer my own question (I think).
> > > > > There is an additional interpolation method called GAUSSIAN
> > > > > which applies only when using the probability variable (-
> > > > > prob_cat_thresh 280.0).
> > > > > So, it
> > > > > looks like the two optional smoothing arguments -gaussian_dx
> > > > > and
> > > > > -gaussian_radius are only used when the method is GAUSSIAN.
IS
> > > > > this correct? This means that there is no smoothing except
when
> > > > > using probability variables. Is this correct?
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of all links contained within the message prior to copying
and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > While the "-field" command line argument indicates that it
can
> > > > > be used one or more times, in this first version of the
tool,
> > > > > it
> > > > > can only be used once.
> > > > >
> > > > > If you specify "-field" more that once you'll get this error
> > > > > message:
> > > > > ERROR  :
> > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > variables
> > > > > were requested.
> > > > > ERROR  :
> > > > >
> > > > > We do plan to support the processing of multiple fields but
it
> > > > > has not yet been implemented.
> > > > >
> > > > > As for the science behind the interpolation methods, it's
very
> > > > > simple. Keep track of the point observations falling inside
> > > > > each
> > > > > of the target grid boxes. Write to the output the min, max,
> > > > > unweighted mean, or median of those values based on the
method
> > > > > chosen. If you only have 1 point obs falling inside each
grid
> > > > > box, then all those methods will result in the same output
> > > > > value.
> > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > then the choice of the derivation method would matter.
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > >
> > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > > > 95747 >
> > > > > >
> > > > > > Another question I have is how is -method used in this
tool?
> > > > > > In the
> > > > > User's
> > > > > > Guide, -method provides options for regridding from one
grid
> > > > > > to another such as is used in regrid_data_plane. In the
case
> > > > > > of Point2Grid, the process involved is not regridding in
the
> > > > > > way that I understand it.
> > > > > > If I
> > > > > > understand it correctly, it takes point observations and
> > > > > > distributes them on to a user provided grid using some
> > > > > > interpolation method. The next
> > > > > step I
> > > > > > think is the smoothing of the field to even out
> > > > > > irregularities
> > > > > > which are caused by errors, noise, etc resulting in the
> > > > > > fitting of the observed values to some function which
appear
> > > > > > to be a Gaussian function. I guess I also need to
understand
> > > > > > what happens when the tool is used without any of the
> > > > > > optional
> > > > > > arguments.
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > To: met_help at ucar.edu
> > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0
> > > > > > Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > your dedication so late into the evening! Can I add new
> > > > > > variables in the
> > > > > config
> > > > > > file? Any info on the science behind the interpolation
> > > > > > methods
> > > > > > would be appreciated, but that can wait until Monday.
> > > > > >
> > > > > > Take care and stay safe.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify
> > > > > the
> > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > links
> > > > > contained
> > > > > > within the message prior to copying and pasting the
address
> > > > > > to
> > > > > > a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > I see you have a question about interpolation options in
the
> > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > point2grid tool that are
> > > > > run
> > > > > > for our unit tests:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/un
> > > > > it
> > > > > _point2grid.xml
> > > > > >
> > > > > > While most of them DO NOT use the "-method" command line
> > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > maximum value falling within each grid box.
> > > > > >
> > > > > > Inspecting the code in this file:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/too
> > > > > ls
> > > > > /other/point2grid/point2grid.cc
> > > > > > I see that the supported options for -method are MIN, MAX,
> > > > > > MEDIAN, and UW_MEAN.
> > > > > >
> > > > > > Seems like we should add that info to the usage statement!
I
> > > > > > wrote up
> > > > > this
> > > > > > GitHub issue to make sure we remember to do so before the
9.1
> > > > > > release:
> > > > > > Caution-Caution-Caution-Caution-
> > > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > > es/1396
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > > >        Queue: met_help
> > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >        Owner: Nobody
> > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > >       Status: new
> > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
> > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > ?id=95747
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > The optional argument for smoothing or interpolation
shown
> > > > > > > in the
> > > > > User's
> > > > > > > Guide is Gaussian, but are other methods available? If
so,
> > > > > > > can you
> > > > > > provide
> > > > > > > the names and how to specify them as an optional
argument?
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > Stay safe.
> > > > > > >
> > > > > > > T/
> > > > > > > John
> > > > > > >
> > > > > > > Mr. John W. Raby
> > > > > > > U.S. Army Research Laboratory White Sands Missile Range,
NM
> > > > > > > 88002
> > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
FAX
> > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
>
>
>
> CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jul 13 13:14:50 2020

typo: MAXGAUSS, not MAXGAUSE

Cheers,
Howard

On Mon Jul 13 13:13:39 2020, hsoh wrote:
> I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS
and
> GAUSSIAN method is only used for the interpolation. Please set "-
> method MAXGAUSE". The from_grid and to_grid are the same, there is
no
> difference with additional MAX interpolation.
>
> Cheers,
>  Howard
>
>
> On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > I tried running MET V9.0 regrid_data_plane following your example
> > below, but
> > I'm getting and error:
> > ERROR  :
> > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > regridding, only
> > interpolating.
> >
> > The following is the command I used to run it:
> >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NOV2016.nc
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gauss_test
> > -v 9
> >
> > Any suggestions on how to perform Gaussian smoothing on the output
> > TMP
> > field
> > from point2grid?
> >
> > Thanks.
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, July 6, 2020 11:38 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > Sorry for the late response.
> >
> > The input and to_grid must be the same for Gaussian smoothing and
> > regrid_data_plane shows the usage message if the to_grid is
missing.
> > There are
> > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > -field
> > 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> >
> > Cheers,
> > Howard
> >
> >
> >
> > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > Thanks for tailoring the command for my particular interests.
Just
> > > curious about your regrid_data_plane command. You provided the
> > > input
> > > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > > regrid_data_plane and then you provide the name of the file to
be
> > > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > > but you didn't provide the to_grid. Is this because it knows
that
> > > the
> > > grids of the input and output files are the same and thus
ignores
> > > the
> > > missing to_grid argument?
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, June 29, 2020 2:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > > the identity of the sender, and confirm the authenticity of all
> > > links
> > > contained within the message prior to copying and pasting the
> > > address
> > > to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > > this case, we are interested in Gaussian smoothing without
> > > regriding.
> > > The input and the target grid should be the same, which is the
> > > output
> > > of point2grid.
> > >
> > > Here is an example (assuming the point data file is generated by
> > > pb2nc).
> > >
> > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > >       G212 pb2nc_TMP.nc \
> > >       -field 'name="TMP"; level="*";' \
> > >       -v 1
> > >
> > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > >       -field 'name="TMP"; level="(*,*)";' \
> > >       -method GAUSSIAN -v 1
> > >
> > > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0)
if
> > > necessary
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Howard -
> > > >
> > > > Thanks for your quick response with details in answer to my
> > > > question.
> > > > Right now my interest is with non-probability fields, so your
> > > > point
> > > > in 1. about the possibility of using Gaussian smoothing for
> > > > -field
> > > > argument is interesting.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting
> > > > the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Two layers for the interpolation with point2grid:
> > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw
data
> > > > - layer 2: Optional, the GAUSSIAN smoothing for the
probability
> > > > variable. (the existence of valid values if the probability
> > > > threshold is not given)
> > > >
> > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > > is
> > > > a
> > > > default) for the layer 1 and "-method GAUSSIAN" for the layer
2.
> > > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > > layer
> > > > 1
> > > > and "-method GAUSSIAN" for the layer 2.
> > > > + Override the interpolation method of layer 1 with "-method
> > > > <METHOD>"
> > > >
> > > > 1. The Gaussian smoothing at point2grid is activated and
applied
> > > > into the probability variable only if -method GAUSSIAN or
-method
> > > > MAXGAUSS is given as the command line argument. It's not
applied
> > > > into the variable at "-field" argument. If you want, You can
do
> > > > this
> > > > by using regrid_data_plane (passing the output of point2grid
as
> > > > the
> > > > input of regrid_data_plane with GAUSSIAN method).
> > > >
> > > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > > GAUSSIAN
> > > > or -method MAXGAUSS is NOT given. The default values are
applied
> > > > if
> > > > not given.
> > > >
> > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > > MAXGAUSS argument.
> > > >
> > > > 4 Without -prob_cat_thresh argument, it will be the status of
> > > > data
> > > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > > condition or not. So the values of the probability are 0 or 1.
> > > >
> > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > requested
> > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > John,
> > > > >
> > > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > > development of the point2grid tool. He should be able to
better
> > > > > answer these questions.
> > > >
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-url: Caution-
> > > > > > Caution-Caution-
> > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > d=95747 >
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > I ran it to see the usage and there are details in the
usage
> > > > > > dialog and also in the XML file you pointed me to which
> > > > > > helped
> > > > > > to answer my own question (I think).
> > > > > > There is an additional interpolation method called
GAUSSIAN
> > > > > > which applies only when using the probability variable (-
> > > > > > prob_cat_thresh 280.0).
> > > > > > So, it
> > > > > > looks like the two optional smoothing arguments
-gaussian_dx
> > > > > > and
> > > > > > -gaussian_radius are only used when the method is
GAUSSIAN.
> > > > > > IS
> > > > > > this correct? This means that there is no smoothing except
> > > > > > when
> > > > > > using probability variables. Is this correct?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please
> > > > > > verify the identity of the sender, and confirm the
> > > > > > authenticity
> > > > > > of all links contained within the message prior to copying
> > > > > > and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > While the "-field" command line argument indicates that it
> > > > > > can
> > > > > > be used one or more times, in this first version of the
tool,
> > > > > > it
> > > > > > can only be used once.
> > > > > >
> > > > > > If you specify "-field" more that once you'll get this
error
> > > > > > message:
> > > > > > ERROR  :
> > > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > > variables
> > > > > > were requested.
> > > > > > ERROR  :
> > > > > >
> > > > > > We do plan to support the processing of multiple fields
but
> > > > > > it
> > > > > > has not yet been implemented.
> > > > > >
> > > > > > As for the science behind the interpolation methods, it's
> > > > > > very
> > > > > > simple. Keep track of the point observations falling
inside
> > > > > > each
> > > > > > of the target grid boxes. Write to the output the min,
max,
> > > > > > unweighted mean, or median of those values based on the
> > > > > > method
> > > > > > chosen. If you only have 1 point obs falling inside each
grid
> > > > > > box, then all those methods will result in the same output
> > > > > > value.
> > > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > > then the choice of the derivation method would matter.
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > >
> > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via
RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=
> > > > > > 95747 >
> > > > > > >
> > > > > > > Another question I have is how is -method used in this
> > > > > > > tool?
> > > > > > > In the
> > > > > > User's
> > > > > > > Guide, -method provides options for regridding from one
> > > > > > > grid
> > > > > > > to another such as is used in regrid_data_plane. In the
> > > > > > > case
> > > > > > > of Point2Grid, the process involved is not regridding in
> > > > > > > the
> > > > > > > way that I understand it.
> > > > > > > If I
> > > > > > > understand it correctly, it takes point observations and
> > > > > > > distributes them on to a user provided grid using some
> > > > > > > interpolation method. The next
> > > > > > step I
> > > > > > > think is the smoothing of the field to even out
> > > > > > > irregularities
> > > > > > > which are caused by errors, noise, etc resulting in the
> > > > > > > fitting of the observed values to some function which
> > > > > > > appear
> > > > > > > to be a Gaussian function. I guess I also need to
> > > > > > > understand
> > > > > > > what happens when the tool is used without any of the
> > > > > > > optional
> > > > > > > arguments.
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > To: met_help at ucar.edu
> > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > MET
> > > > > > > V9.0
> > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > John -
> > > > > > >
> > > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > > your dedication so late into the evening! Can I add new
> > > > > > > variables in the
> > > > > > config
> > > > > > > file? Any info on the science behind the interpolation
> > > > > > > methods
> > > > > > > would be appreciated, but that can wait until Monday.
> > > > > > >
> > > > > > > Take care and stay safe.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify
> > > > > > the
> > > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > > links
> > > > > > contained
> > > > > > > within the message prior to copying and pasting the
address
> > > > > > > to
> > > > > > > a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > John,
> > > > > > >
> > > > > > > I see you have a question about interpolation options in
> > > > > > > the
> > > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > > point2grid tool that are
> > > > > > run
> > > > > > > for our unit tests:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/un
> > > > > > it
> > > > > > _point2grid.xml
> > > > > > >
> > > > > > > While most of them DO NOT use the "-method" command line
> > > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > > maximum value falling within each grid box.
> > > > > > >
> > > > > > > Inspecting the code in this file:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/too
> > > > > > ls
> > > > > > /other/point2grid/point2grid.cc
> > > > > > > I see that the supported options for -method are MIN,
MAX,
> > > > > > > MEDIAN, and UW_MEAN.
> > > > > > >
> > > > > > > Seems like we should add that info to the usage
statement!
> > > > > > > I
> > > > > > > wrote up
> > > > > > this
> > > > > > > GitHub issue to make sure we remember to do so before
the
> > > > > > > 9.1
> > > > > > > release:
> > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > > > es/1396
> > > > > > >
> > > > > > > Hope that helps clarify.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT
> > > > > > > <
> > > > > > > met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
upon.
> > > > > > > > Transaction: Ticket created by
john.w.raby2.civ at mail.mil
> > > > > > > >        Queue: met_help
> > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >        Owner: Nobody
> > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > >       Status: new
> > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > ?id=95747
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > The optional argument for smoothing or interpolation
> > > > > > > > shown
> > > > > > > > in the
> > > > > > User's
> > > > > > > > Guide is Gaussian, but are other methods available? If
> > > > > > > > so,
> > > > > > > > can you
> > > > > > > provide
> > > > > > > > the names and how to specify them as an optional
> > > > > > > > argument?
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > Stay safe.
> > > > > > > >
> > > > > > > > T/
> > > > > > > > John
> > > > > > > >
> > > > > > > > Mr. John W. Raby
> > > > > > > > U.S. Army Research Laboratory White Sands Missile
Range,
> > > > > > > > NM
> > > > > > > > 88002
> > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > Teleworking: Contact Admin Specialist at (410) 278-
6974
> > > > > > > > FAX
> > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 13 13:19:02 2020

CLASSIFICATION: UNCLASSIFIED

Howard -

Thanks for your fast response. No problem. I'll use MAXGAUSS vice
GUASSIAN.

R/
John

-----Original Message-----
From: Howard Soh via RT <met_help at ucar.edu>
Sent: Monday, July 13, 2020 1:14 PM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS and
GAUSSIAN
method is only used for the interpolation. Please set "-method
MAXGAUSE". The
from_grid and to_grid are the same, there is no difference with
additional MAX
interpolation.

Cheers,
Howard


On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> I tried running MET V9.0 regrid_data_plane following your example
> below, but I'm getting and error:
> ERROR  :
> ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> regridding, only interpolating.
>
> The following is the command I used to run it:
>  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> V2016.nc
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> V2016.nc
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
>   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
>
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> ss_test
> -v 9
>
> Any suggestions on how to perform Gaussian smoothing on the output
TMP
> field from point2grid?
>
> Thanks.
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, July 6, 2020 11:38 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Sorry for the late response.
>
> The input and to_grid must be the same for Gaussian smoothing and
> regrid_data_plane shows the usage message if the to_grid is missing.
> There are
> two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
> -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
>
> Cheers,
> Howard
>
>
>
> On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > Thanks for tailoring the command for my particular interests. Just
> > curious about your regrid_data_plane command. You provided the
input
> > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > regrid_data_plane and then you provide the name of the file to be
> > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > but you didn't provide the to_grid. Is this because it knows that
> > the grids of the input and output files are the same and thus
> > ignores the missing to_grid argument?
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, June 29, 2020 2:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the identity of the sender, and confirm the authenticity of
> > all links contained within the message prior to copying and
pasting
> > the address to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > this case, we are interested in Gaussian smoothing without
regriding.
> > The input and the target grid should be the same, which is the
> > output of point2grid.
> >
> > Here is an example (assuming the point data file is generated by
> > pb2nc).
> >
> > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> >       G212 pb2nc_TMP.nc \
> >       -field 'name="TMP"; level="*";' \
> >       -v 1
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> >       -field 'name="TMP"; level="(*,*)";' \
> >       -method GAUSSIAN -v 1
> >
> > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> > necessary
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > Howard -
> > >
> > > Thanks for your quick response with details in answer to my
> > > question.
> > > Right now my interest is with non-probability fields, so your
> > > point in 1. about the possibility of using Gaussian smoothing
for
> > > -field argument is interesting.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Monday, June 29, 2020 12:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
> > > pasting the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Two layers for the interpolation with point2grid:
> > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > > variable. (the existence of valid values if the probability
> > > threshold is not given)
> > >
> > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > is
> > > a
> > > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > layer
> > > 1
> > > and "-method GAUSSIAN" for the layer 2.
> > > + Override the interpolation method of layer 1 with "-method
> > > <METHOD>"
> > >
> > > 1. The Gaussian smoothing at point2grid is activated and applied
> > > into the probability variable only if -method GAUSSIAN or
-method
> > > MAXGAUSS is given as the command line argument. It's not applied
> > > into the variable at "-field" argument. If you want, You can do
> > > this by using regrid_data_plane (passing the output of
point2grid
> > > as the input of regrid_data_plane with GAUSSIAN method).
> > >
> > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > GAUSSIAN or -method MAXGAUSS is NOT given. The default values
are
> > > applied if not given.
> > >
> > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > MAXGAUSS argument.
> > >
> > > 4 Without -prob_cat_thresh argument, it will be the status of
data
> > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > condition or not. So the values of the probability are 0 or 1.
> > >
> > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
requested
> > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > John,
> > > >
> > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > development of the point2grid tool. He should be able to
better
> > > > answer these questions.
> > >
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > d=95747 >
> > > > >
> > > > > John -
> > > > >
> > > > > I ran it to see the usage and there are details in the usage
> > > > > dialog and also in the XML file you pointed me to which
helped
> > > > > to answer my own question (I think).
> > > > > There is an additional interpolation method called GAUSSIAN
> > > > > which applies only when using the probability variable (-
> > > > > prob_cat_thresh 280.0).
> > > > > So, it
> > > > > looks like the two optional smoothing arguments -gaussian_dx
> > > > > and -gaussian_radius are only used when the method is
> > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > smoothing except when using probability variables. Is this
> > > > > correct?
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
> > > > > Please verify the identity of the sender, and confirm the
> > > > > authenticity of all links contained within the message prior
> > > > > to copying and pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > While the "-field" command line argument indicates that it
can
> > > > > be used one or more times, in this first version of the
tool,
> > > > > it can only be used once.
> > > > >
> > > > > If you specify "-field" more that once you'll get this error
> > > > > message:
> > > > > ERROR  :
> > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > variables
> > > > > were requested.
> > > > > ERROR  :
> > > > >
> > > > > We do plan to support the processing of multiple fields but
it
> > > > > has not yet been implemented.
> > > > >
> > > > > As for the science behind the interpolation methods, it's
very
> > > > > simple. Keep track of the point observations falling inside
> > > > > each of the target grid boxes. Write to the output the min,
> > > > > max, unweighted mean, or median of those values based on the
> > > > > method chosen. If you only have 1 point obs falling inside
> > > > > each grid box, then all those methods will result in the
same
> > > > > output value.
> > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > then the choice of the derivation method would matter.
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > >
> > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > ?id=
> > > > > 95747 >
> > > > > >
> > > > > > Another question I have is how is -method used in this
tool?
> > > > > > In the
> > > > > User's
> > > > > > Guide, -method provides options for regridding from one
grid
> > > > > > to another such as is used in regrid_data_plane. In the
case
> > > > > > of Point2Grid, the process involved is not regridding in
the
> > > > > > way that I understand it.
> > > > > > If I
> > > > > > understand it correctly, it takes point observations and
> > > > > > distributes them on to a user provided grid using some
> > > > > > interpolation method. The next
> > > > > step I
> > > > > > think is the smoothing of the field to even out
> > > > > > irregularities which are caused by errors, noise, etc
> > > > > > resulting in the fitting of the observed values to some
> > > > > > function which appear to be a Gaussian function. I guess I
> > > > > > also need to understand what happens when the tool is used
> > > > > > without any of the optional arguments.
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > To: met_help at ucar.edu
> > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0
> > > > > > Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > your dedication so late into the evening! Can I add new
> > > > > > variables in the
> > > > > config
> > > > > > file? Any info on the science behind the interpolation
> > > > > > methods would be appreciated, but that can wait until
> > > > > > Monday.
> > > > > >
> > > > > > Take care and stay safe.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify
> > > > > the
> > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > links
> > > > > contained
> > > > > > within the message prior to copying and pasting the
address
> > > > > > to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > I see you have a question about interpolation options in
the
> > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > point2grid tool that are
> > > > > run
> > > > > > for our unit tests:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > un
> > > > > it
> > > > > _point2grid.xml
> > > > > >
> > > > > > While most of them DO NOT use the "-method" command line
> > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > maximum value falling within each grid box.
> > > > > >
> > > > > > Inspecting the code in this file:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > oo
> > > > > ls
> > > > > /other/point2grid/point2grid.cc
> > > > > > I see that the supported options for -method are MIN, MAX,
> > > > > > MEDIAN, and UW_MEAN.
> > > > > >
> > > > > > Seems like we should add that info to the usage statement!
I
> > > > > > wrote up
> > > > > this
> > > > > > GitHub issue to make sure we remember to do so before the
> > > > > > 9.1
> > > > > > release:
> > > > > > Caution-Caution-Caution-Caution-
> > > > > > Caution-Caution-https://github.com/NCAR/MET/issu
> > > > > > es/1396
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT
> > > > > > < met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > > >        Queue: met_help
> > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >        Owner: Nobody
> > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > >       Status: new
> > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > ?id=95747
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > The optional argument for smoothing or interpolation
shown
> > > > > > > in the
> > > > > User's
> > > > > > > Guide is Gaussian, but are other methods available? If
so,
> > > > > > > can you
> > > > > > provide
> > > > > > > the names and how to specify them as an optional
argument?
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > Stay safe.
> > > > > > >
> > > > > > > T/
> > > > > > > John
> > > > > > >
> > > > > > > Mr. John W. Raby
> > > > > > > U.S. Army Research Laboratory White Sands Missile Range,
> > > > > > > NM
> > > > > > > 88002
> > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > > > > > FAX
> > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
>
>
>
> CLASSIFICATION: UNCLASSIFIED



CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 13 13:58:03 2020

CLASSIFICATION: UNCLASSIFIED

Howard -

The next run with MAXGAUSS was successful, although I have to plot it
to see
what it looks like. I've attached the log file. Does it indicate any
issues
with how it ran?

R/
John

-----Original Message-----
From: Howard Soh via RT <met_help at ucar.edu>
Sent: Monday, July 13, 2020 1:14 PM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS and
GAUSSIAN
method is only used for the interpolation. Please set "-method
MAXGAUSE". The
from_grid and to_grid are the same, there is no difference with
additional MAX
interpolation.

Cheers,
Howard


On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> I tried running MET V9.0 regrid_data_plane following your example
> below, but I'm getting and error:
> ERROR  :
> ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> regridding, only interpolating.
>
> The following is the command I used to run it:
>  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> V2016.nc
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> V2016.nc
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
>   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
>
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> ss_test
> -v 9
>
> Any suggestions on how to perform Gaussian smoothing on the output
TMP
> field from point2grid?
>
> Thanks.
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, July 6, 2020 11:38 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Sorry for the late response.
>
> The input and to_grid must be the same for Gaussian smoothing and
> regrid_data_plane shows the usage message if the to_grid is missing.
> There are
> two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
>
> ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc pb2nc_gaussian_TMP.nc
> -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
>
> Cheers,
> Howard
>
>
>
> On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > Thanks for tailoring the command for my particular interests. Just
> > curious about your regrid_data_plane command. You provided the
input
> > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > regrid_data_plane and then you provide the name of the file to be
> > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > but you didn't provide the to_grid. Is this because it knows that
> > the grids of the input and output files are the same and thus
> > ignores the missing to_grid argument?
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, June 29, 2020 2:50 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the identity of the sender, and confirm the authenticity of
> > all links contained within the message prior to copying and
pasting
> > the address to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > this case, we are interested in Gaussian smoothing without
regriding.
> > The input and the target grid should be the same, which is the
> > output of point2grid.
> >
> > Here is an example (assuming the point data file is generated by
> > pb2nc).
> >
> > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> >       G212 pb2nc_TMP.nc \
> >       -field 'name="TMP"; level="*";' \
> >       -v 1
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> >       -field 'name="TMP"; level="(*,*)";' \
> >       -method GAUSSIAN -v 1
> >
> > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0) if
> > necessary
> >
> > Cheers,
> > Howard
> >
> > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > Howard -
> > >
> > > Thanks for your quick response with details in answer to my
> > > question.
> > > Right now my interest is with non-probability fields, so your
> > > point in 1. about the possibility of using Gaussian smoothing
for
> > > -field argument is interesting.
> > >
> > > R/
> > > John
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Monday, June 29, 2020 12:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
> > > pasting the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Two layers for the interpolation with point2grid:
> > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw data
> > > - layer 2: Optional, the GAUSSIAN smoothing for the probability
> > > variable. (the existence of valid values if the probability
> > > threshold is not given)
> > >
> > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > is
> > > a
> > > default) for the layer 1 and "-method GAUSSIAN" for the layer 2.
> > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > layer
> > > 1
> > > and "-method GAUSSIAN" for the layer 2.
> > > + Override the interpolation method of layer 1 with "-method
> > > <METHOD>"
> > >
> > > 1. The Gaussian smoothing at point2grid is activated and applied
> > > into the probability variable only if -method GAUSSIAN or
-method
> > > MAXGAUSS is given as the command line argument. It's not applied
> > > into the variable at "-field" argument. If you want, You can do
> > > this by using regrid_data_plane (passing the output of
point2grid
> > > as the input of regrid_data_plane with GAUSSIAN method).
> > >
> > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > GAUSSIAN or -method MAXGAUSS is NOT given. The default values
are
> > > applied if not given.
> > >
> > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > MAXGAUSS argument.
> > >
> > > 4 Without -prob_cat_thresh argument, it will be the status of
data
> > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > condition or not. So the values of the probability are 0 or 1.
> > >
> > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
requested
> > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > John,
> > > >
> > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > development of the point2grid tool. He should be able to
better
> > > > answer these questions.
> > >
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > d=95747 >
> > > > >
> > > > > John -
> > > > >
> > > > > I ran it to see the usage and there are details in the usage
> > > > > dialog and also in the XML file you pointed me to which
helped
> > > > > to answer my own question (I think).
> > > > > There is an additional interpolation method called GAUSSIAN
> > > > > which applies only when using the probability variable (-
> > > > > prob_cat_thresh 280.0).
> > > > > So, it
> > > > > looks like the two optional smoothing arguments -gaussian_dx
> > > > > and -gaussian_radius are only used when the method is
> > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > smoothing except when using probability variables. Is this
> > > > > correct?
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
> > > > > Please verify the identity of the sender, and confirm the
> > > > > authenticity of all links contained within the message prior
> > > > > to copying and pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > John,
> > > > >
> > > > > While the "-field" command line argument indicates that it
can
> > > > > be used one or more times, in this first version of the
tool,
> > > > > it can only be used once.
> > > > >
> > > > > If you specify "-field" more that once you'll get this error
> > > > > message:
> > > > > ERROR  :
> > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > variables
> > > > > were requested.
> > > > > ERROR  :
> > > > >
> > > > > We do plan to support the processing of multiple fields but
it
> > > > > has not yet been implemented.
> > > > >
> > > > > As for the science behind the interpolation methods, it's
very
> > > > > simple. Keep track of the point observations falling inside
> > > > > each of the target grid boxes. Write to the output the min,
> > > > > max, unweighted mean, or median of those values based on the
> > > > > method chosen. If you only have 1 point obs falling inside
> > > > > each grid box, then all those methods will result in the
same
> > > > > output value.
> > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > then the choice of the derivation method would matter.
> > > > >
> > > > > Hope that helps clarify.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > >
> > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-Caution-Caution-url: Caution-
> > > > > Caution-Caution-
> > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > ?id=
> > > > > 95747 >
> > > > > >
> > > > > > Another question I have is how is -method used in this
tool?
> > > > > > In the
> > > > > User's
> > > > > > Guide, -method provides options for regridding from one
grid
> > > > > > to another such as is used in regrid_data_plane. In the
case
> > > > > > of Point2Grid, the process involved is not regridding in
the
> > > > > > way that I understand it.
> > > > > > If I
> > > > > > understand it correctly, it takes point observations and
> > > > > > distributes them on to a user provided grid using some
> > > > > > interpolation method. The next
> > > > > step I
> > > > > > think is the smoothing of the field to even out
> > > > > > irregularities which are caused by errors, noise, etc
> > > > > > resulting in the fitting of the observed values to some
> > > > > > function which appear to be a Gaussian function. I guess I
> > > > > > also need to understand what happens when the tool is used
> > > > > > without any of the optional arguments.
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > To: met_help at ucar.edu
> > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0
> > > > > > Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > your dedication so late into the evening! Can I add new
> > > > > > variables in the
> > > > > config
> > > > > > file? Any info on the science behind the interpolation
> > > > > > methods would be appreciated, but that can wait until
> > > > > > Monday.
> > > > > >
> > > > > > Take care and stay safe.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify
> > > > > the
> > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > links
> > > > > contained
> > > > > > within the message prior to copying and pasting the
address
> > > > > > to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > I see you have a question about interpolation options in
the
> > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > point2grid tool that are
> > > > > run
> > > > > > for our unit tests:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > un
> > > > > it
> > > > > _point2grid.xml
> > > > > >
> > > > > > While most of them DO NOT use the "-method" command line
> > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > maximum value falling within each grid box.
> > > > > >
> > > > > > Inspecting the code in this file:
> > > > > > Caution-
> > > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
> > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > oo
> > > > > ls
> > > > > /other/point2grid/point2grid.cc
> > > > > > I see that the supported options for -method are MIN, MAX,
> > > > > > MEDIAN, and UW_MEAN.
> > > > > >
> > > > > > Seems like we should add that info to the usage statement!
I
> > > > > > wrote up
> > > > > this
> > > > > > GitHub issue to make sure we remember to do so before the
> > > > > > 9.1
> > > > > > release:
> > > > > > Caution-Caution-Caution-Caution-
> > > > > > Caution-Caution-https://github.com/NCAR/MET/issu
> > > > > > es/1396
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT
> > > > > > < met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted upon.
> > > > > > > Transaction: Ticket created by john.w.raby2.civ at mail.mil
> > > > > > >        Queue: met_help
> > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >        Owner: Nobody
> > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > >       Status: new
> > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > ?id=95747
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > The optional argument for smoothing or interpolation
shown
> > > > > > > in the
> > > > > User's
> > > > > > > Guide is Gaussian, but are other methods available? If
so,
> > > > > > > can you
> > > > > > provide
> > > > > > > the names and how to specify them as an optional
argument?
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > Stay safe.
> > > > > > >
> > > > > > > T/
> > > > > > > John
> > > > > > >
> > > > > > > Mr. John W. Raby
> > > > > > > U.S. Army Research Laboratory White Sands Missile Range,
> > > > > > > NM
> > > > > > > 88002
> > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > Teleworking: Contact Admin Specialist at (410) 278-6974
> > > > > > > FAX
> > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
>
>
>
> CLASSIFICATION: UNCLASSIFIED



CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Mon Jul 13 15:33:21 2020

Something is not right. I will take a look

DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is 246.351
to 284.817.
DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is -9999
to -9999.

The input is OK (between 246.351 and 284.817), but the regrided output
is all missing value (-9999).

Cheers,
Howard

The input file has
On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> The next run with MAXGAUSS was successful, although I have to plot
it
> to see
> what it looks like. I've attached the log file. Does it indicate any
> issues
> with how it ran?
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, July 13, 2020 1:14 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS
and
> GAUSSIAN
> method is only used for the interpolation. Please set "-method
> MAXGAUSE". The
> from_grid and to_grid are the same, there is no difference with
> additional MAX
> interpolation.
>
> Cheers,
> Howard
>
>
> On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > I tried running MET V9.0 regrid_data_plane following your example
> > below, but I'm getting and error:
> > ERROR  :
> > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > regridding, only interpolating.
> >
> > The following is the command I used to run it:
> >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > V2016.nc
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > V2016.nc
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > ss_test
> > -v 9
> >
> > Any suggestions on how to perform Gaussian smoothing on the output
> > TMP
> > field from point2grid?
> >
> > Thanks.
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, July 6, 2020 11:38 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Sorry for the late response.
> >
> > The input and to_grid must be the same for Gaussian smoothing and
> > regrid_data_plane shows the usage message if the to_grid is
missing.
> > There are
> > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> >
> > Cheers,
> > Howard
> >
> >
> >
> > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > Thanks for tailoring the command for my particular interests.
Just
> > > curious about your regrid_data_plane command. You provided the
> > > input
> > > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > > regrid_data_plane and then you provide the name of the file to
be
> > > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > > but you didn't provide the to_grid. Is this because it knows
that
> > > the grids of the input and output files are the same and thus
> > > ignores the missing to_grid argument?
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, June 29, 2020 2:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
pasting
> > > the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > > this case, we are interested in Gaussian smoothing without
> > > regriding.
> > > The input and the target grid should be the same, which is the
> > > output of point2grid.
> > >
> > > Here is an example (assuming the point data file is generated by
> > > pb2nc).
> > >
> > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > >       G212 pb2nc_TMP.nc \
> > >       -field 'name="TMP"; level="*";' \
> > >       -v 1
> > >
> > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > >       -field 'name="TMP"; level="(*,*)";' \
> > >       -method GAUSSIAN -v 1
> > >
> > > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0)
if
> > > necessary
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Howard -
> > > >
> > > > Thanks for your quick response with details in answer to my
> > > > question.
> > > > Right now my interest is with non-probability fields, so your
> > > > point in 1. about the possibility of using Gaussian smoothing
for
> > > > -field argument is interesting.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Two layers for the interpolation with point2grid:
> > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw
data
> > > > - layer 2: Optional, the GAUSSIAN smoothing for the
probability
> > > > variable. (the existence of valid values if the probability
> > > > threshold is not given)
> > > >
> > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > > is
> > > > a
> > > > default) for the layer 1 and "-method GAUSSIAN" for the layer
2.
> > > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > > layer
> > > > 1
> > > > and "-method GAUSSIAN" for the layer 2.
> > > > + Override the interpolation method of layer 1 with "-method
> > > > <METHOD>"
> > > >
> > > > 1. The Gaussian smoothing at point2grid is activated and
applied
> > > > into the probability variable only if -method GAUSSIAN or
-method
> > > > MAXGAUSS is given as the command line argument. It's not
applied
> > > > into the variable at "-field" argument. If you want, You can
do
> > > > this by using regrid_data_plane (passing the output of
point2grid
> > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > >
> > > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default values
are
> > > > applied if not given.
> > > >
> > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > > MAXGAUSS argument.
> > > >
> > > > 4 Without -prob_cat_thresh argument, it will be the status of
> > > > data
> > > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > > condition or not. So the values of the probability are 0 or 1.
> > > >
> > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > requested
> > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > John,
> > > > >
> > > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > > development of the point2grid tool. He should be able to
better
> > > > > answer these questions.
> > > >
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-Caution-url: Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > d=95747 >
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > I ran it to see the usage and there are details in the
usage
> > > > > > dialog and also in the XML file you pointed me to which
> > > > > > helped
> > > > > > to answer my own question (I think).
> > > > > > There is an additional interpolation method called
GAUSSIAN
> > > > > > which applies only when using the probability variable (-
> > > > > > prob_cat_thresh 280.0).
> > > > > > So, it
> > > > > > looks like the two optional smoothing arguments
-gaussian_dx
> > > > > > and -gaussian_radius are only used when the method is
> > > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > > smoothing except when using probability variables. Is this
> > > > > > correct?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify the identity of the sender, and confirm the
> > > > > > authenticity of all links contained within the message
prior
> > > > > > to copying and pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > While the "-field" command line argument indicates that it
> > > > > > can
> > > > > > be used one or more times, in this first version of the
tool,
> > > > > > it can only be used once.
> > > > > >
> > > > > > If you specify "-field" more that once you'll get this
error
> > > > > > message:
> > > > > > ERROR  :
> > > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > > variables
> > > > > > were requested.
> > > > > > ERROR  :
> > > > > >
> > > > > > We do plan to support the processing of multiple fields
but
> > > > > > it
> > > > > > has not yet been implemented.
> > > > > >
> > > > > > As for the science behind the interpolation methods, it's
> > > > > > very
> > > > > > simple. Keep track of the point observations falling
inside
> > > > > > each of the target grid boxes. Write to the output the
min,
> > > > > > max, unweighted mean, or median of those values based on
the
> > > > > > method chosen. If you only have 1 point obs falling inside
> > > > > > each grid box, then all those methods will result in the
same
> > > > > > output value.
> > > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > > then the choice of the derivation method would matter.
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > >
> > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via
RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <Caution-Caution-Caution-Caution-Caution-url: Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
> > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > ?id=
> > > > > > 95747 >
> > > > > > >
> > > > > > > Another question I have is how is -method used in this
> > > > > > > tool?
> > > > > > > In the
> > > > > > User's
> > > > > > > Guide, -method provides options for regridding from one
> > > > > > > grid
> > > > > > > to another such as is used in regrid_data_plane. In the
> > > > > > > case
> > > > > > > of Point2Grid, the process involved is not regridding in
> > > > > > > the
> > > > > > > way that I understand it.
> > > > > > > If I
> > > > > > > understand it correctly, it takes point observations and
> > > > > > > distributes them on to a user provided grid using some
> > > > > > > interpolation method. The next
> > > > > > step I
> > > > > > > think is the smoothing of the field to even out
> > > > > > > irregularities which are caused by errors, noise, etc
> > > > > > > resulting in the fitting of the observed values to some
> > > > > > > function which appear to be a Gaussian function. I guess
I
> > > > > > > also need to understand what happens when the tool is
used
> > > > > > > without any of the optional arguments.
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > To: met_help at ucar.edu
> > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > MET
> > > > > > > V9.0
> > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > John -
> > > > > > >
> > > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > > your dedication so late into the evening! Can I add new
> > > > > > > variables in the
> > > > > > config
> > > > > > > file? Any info on the science behind the interpolation
> > > > > > > methods would be appreciated, but that can wait until
> > > > > > > Monday.
> > > > > > >
> > > > > > > Take care and stay safe.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify
> > > > > > the
> > > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > > links
> > > > > > contained
> > > > > > > within the message prior to copying and pasting the
address
> > > > > > > to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > John,
> > > > > > >
> > > > > > > I see you have a question about interpolation options in
> > > > > > > the
> > > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > > point2grid tool that are
> > > > > > run
> > > > > > > for our unit tests:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > un
> > > > > > it
> > > > > > _point2grid.xml
> > > > > > >
> > > > > > > While most of them DO NOT use the "-method" command line
> > > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > > maximum value falling within each grid box.
> > > > > > >
> > > > > > > Inspecting the code in this file:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > oo
> > > > > > ls
> > > > > > /other/point2grid/point2grid.cc
> > > > > > > I see that the supported options for -method are MIN,
MAX,
> > > > > > > MEDIAN, and UW_MEAN.
> > > > > > >
> > > > > > > Seems like we should add that info to the usage
statement!
> > > > > > > I
> > > > > > > wrote up
> > > > > > this
> > > > > > > GitHub issue to make sure we remember to do so before
the
> > > > > > > 9.1
> > > > > > > release:
> > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > Caution-Caution-https://github.com/NCAR/MET/issu
> > > > > > > es/1396
> > > > > > >
> > > > > > > Hope that helps clarify.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT
> > > > > > > < met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
upon.
> > > > > > > > Transaction: Ticket created by
john.w.raby2.civ at mail.mil
> > > > > > > >        Queue: met_help
> > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >        Owner: Nobody
> > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > >       Status: new
> > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > url:
> > > > > > > > Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > ?id=95747
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > The optional argument for smoothing or interpolation
> > > > > > > > shown
> > > > > > > > in the
> > > > > > User's
> > > > > > > > Guide is Gaussian, but are other methods available? If
> > > > > > > > so,
> > > > > > > > can you
> > > > > > > provide
> > > > > > > > the names and how to specify them as an optional
> > > > > > > > argument?
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > Stay safe.
> > > > > > > >
> > > > > > > > T/
> > > > > > > > John
> > > > > > > >
> > > > > > > > Mr. John W. Raby
> > > > > > > > U.S. Army Research Laboratory White Sands Missile
Range,
> > > > > > > > NM
> > > > > > > > 88002
> > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > Teleworking: Contact Admin Specialist at (410) 278-
6974
> > > > > > > > FAX
> > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
>
>
>
> CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: RE: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Wed Jul 15 10:30:53 2020

Howard -

Today I successfully ran point2grid but used -method MAXGAUSS and it
ran without error and produced output. I'm uploading the file list,
output, commands used, log file, and plots to the DoD SAFE so you can
take a look. I am also uploading the results of the previous run where
you suggested that I use "11" vice "TMP" which fixed the problem and
ran successfully.

The plots are not what I was expecting somehow. The plotted data look
like just points where the original obs were located rather than a
regularly spaced grid of TMP values. What are your thoughts on this?

Thanks
________________________________________
From: Howard Soh via RT [met_help at ucar.edu]
Sent: Monday, July 13, 2020 3:33 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

Something is not right. I will take a look

DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is 246.351
to 284.817.
DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is -9999
to -9999.

The input is OK (between 246.351 and 284.817), but the regrided output
is all missing value (-9999).

Cheers,
Howard

The input file has
On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Howard -
>
> The next run with MAXGAUSS was successful, although I have to plot
it
> to see
> what it looks like. I've attached the log file. Does it indicate any
> issues
> with how it ran?
>
> R/
> John
>
> -----Original Message-----
> From: Howard Soh via RT <met_help at ucar.edu>
> Sent: Monday, July 13, 2020 1:14 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
>  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS
and
> GAUSSIAN
> method is only used for the interpolation. Please set "-method
> MAXGAUSE". The
> from_grid and to_grid are the same, there is no difference with
> additional MAX
> interpolation.
>
> Cheers,
> Howard
>
>
> On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > I tried running MET V9.0 regrid_data_plane following your example
> > below, but I'm getting and error:
> > ERROR  :
> > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > regridding, only interpolating.
> >
> > The following is the command I used to run it:
> >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > V2016.nc
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > V2016.nc
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > ss_test
> > -v 9
> >
> > Any suggestions on how to perform Gaussian smoothing on the output
> > TMP
> > field from point2grid?
> >
> > Thanks.
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, July 6, 2020 11:38 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Sorry for the late response.
> >
> > The input and to_grid must be the same for Gaussian smoothing and
> > regrid_data_plane shows the usage message if the to_grid is
missing.
> > There are
> > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> >
> > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> >
> > Cheers,
> > Howard
> >
> >
> >
> > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > Thanks for tailoring the command for my particular interests.
Just
> > > curious about your regrid_data_plane command. You provided the
> > > input
> > > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > > regrid_data_plane and then you provide the name of the file to
be
> > > output from regrid_data_plane (pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc),
> > > but you didn't provide the to_grid. Is this because it knows
that
> > > the grids of the input and output files are the same and thus
> > > ignores the missing to_grid argument?
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, June 29, 2020 2:50 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
pasting
> > > the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > The Gaussian smoothing is available with regrid_data_plane, too.
In
> > > this case, we are interested in Gaussian smoothing without
> > > regriding.
> > > The input and the target grid should be the same, which is the
> > > output of point2grid.
> > >
> > > Here is an example (assuming the point data file is generated by
> > > pb2nc).
> > >
> > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > >       G212 pb2nc_TMP.nc \
> > >       -field 'name="TMP"; level="*";' \
> > >       -v 1
> > >
> > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > >       -field 'name="TMP"; level="(*,*)";' \
> > >       -method GAUSSIAN -v 1
> > >
> > > Note: override -gaussian_dx (81.271) & -gaussian_radius (120.0)
if
> > > necessary
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Howard -
> > > >
> > > > Thanks for your quick response with details in answer to my
> > > > question.
> > > > Right now my interest is with non-probability fields, so your
> > > > point in 1. about the possibility of using Gaussian smoothing
for
> > > > -field argument is interesting.
> > > >
> > > > R/
> > > > John
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Two layers for the interpolation with point2grid:
> > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw
data
> > > > - layer 2: Optional, the GAUSSIAN smoothing for the
probability
> > > > variable. (the existence of valid values if the probability
> > > > threshold is not given)
> > > >
> > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
(which
> > > > is
> > > > a
> > > > default) for the layer 1 and "-method GAUSSIAN" for the layer
2.
> > > > + "-method MAXGAUSS" is processed as the "-method MAX" for the
> > > > layer
> > > > 1
> > > > and "-method GAUSSIAN" for the layer 2.
> > > > + Override the interpolation method of layer 1 with "-method
> > > > <METHOD>"
> > > >
> > > > 1. The Gaussian smoothing at point2grid is activated and
applied
> > > > into the probability variable only if -method GAUSSIAN or
-method
> > > > MAXGAUSS is given as the command line argument. It's not
applied
> > > > into the variable at "-field" argument. If you want, You can
do
> > > > this by using regrid_data_plane (passing the output of
point2grid
> > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > >
> > > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default values
are
> > > > applied if not given.
> > > >
> > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
-method
> > > > MAXGAUSS argument.
> > > >
> > > > 4 Without -prob_cat_thresh argument, it will be the status of
> > > > data
> > > > existence. With -prob_cat_thresh argument, it will be meeting
the
> > > > condition or not. So the values of the probability are 0 or 1.
> > > >
> > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > requested
> > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > John,
> > > > >
> > > > > I'm reassigning this ticket to Howard Soh who worked on the
> > > > > development of the point2grid tool. He should be able to
better
> > > > > answer these questions.
> > > >
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <Caution-Caution-Caution-Caution-Caution-url: Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > d=95747 >
> > > > > >
> > > > > > John -
> > > > > >
> > > > > > I ran it to see the usage and there are details in the
usage
> > > > > > dialog and also in the XML file you pointed me to which
> > > > > > helped
> > > > > > to answer my own question (I think).
> > > > > > There is an additional interpolation method called
GAUSSIAN
> > > > > > which applies only when using the probability variable (-
> > > > > > prob_cat_thresh 280.0).
> > > > > > So, it
> > > > > > looks like the two optional smoothing arguments
-gaussian_dx
> > > > > > and -gaussian_radius are only used when the method is
> > > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > > smoothing except when using probability variables. Is this
> > > > > > correct?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > MET
> > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify the identity of the sender, and confirm the
> > > > > > authenticity of all links contained within the message
prior
> > > > > > to copying and pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > John,
> > > > > >
> > > > > > While the "-field" command line argument indicates that it
> > > > > > can
> > > > > > be used one or more times, in this first version of the
tool,
> > > > > > it can only be used once.
> > > > > >
> > > > > > If you specify "-field" more that once you'll get this
error
> > > > > > message:
> > > > > > ERROR  :
> > > > > > ERROR  : process_point_file() -> Support only 1 variable.
"2"
> > > > > > variables
> > > > > > were requested.
> > > > > > ERROR  :
> > > > > >
> > > > > > We do plan to support the processing of multiple fields
but
> > > > > > it
> > > > > > has not yet been implemented.
> > > > > >
> > > > > > As for the science behind the interpolation methods, it's
> > > > > > very
> > > > > > simple. Keep track of the point observations falling
inside
> > > > > > each of the target grid boxes. Write to the output the
min,
> > > > > > max, unweighted mean, or median of those values based on
the
> > > > > > method chosen. If you only have 1 point obs falling inside
> > > > > > each grid box, then all those methods will result in the
same
> > > > > > output value.
> > > > > > Whereas if you have many, as in the case of GOES input
files,
> > > > > > then the choice of the derivation method would matter.
> > > > > >
> > > > > > Hope that helps clarify.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > >
> > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via
RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-url:
Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-Caution-
> > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > ?id=
> > > > > > 95747 >
> > > > > > >
> > > > > > > Another question I have is how is -method used in this
> > > > > > > tool?
> > > > > > > In the
> > > > > > User's
> > > > > > > Guide, -method provides options for regridding from one
> > > > > > > grid
> > > > > > > to another such as is used in regrid_data_plane. In the
> > > > > > > case
> > > > > > > of Point2Grid, the process involved is not regridding in
> > > > > > > the
> > > > > > > way that I understand it.
> > > > > > > If I
> > > > > > > understand it correctly, it takes point observations and
> > > > > > > distributes them on to a user provided grid using some
> > > > > > > interpolation method. The next
> > > > > > step I
> > > > > > > think is the smoothing of the field to even out
> > > > > > > irregularities which are caused by errors, noise, etc
> > > > > > > resulting in the fitting of the observed values to some
> > > > > > > function which appear to be a Gaussian function. I guess
I
> > > > > > > also need to understand what happens when the tool is
used
> > > > > > > without any of the optional arguments.
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > To: met_help at ucar.edu
> > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > MET
> > > > > > > V9.0
> > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > John -
> > > > > > >
> > > > > > > Thanks for your quick response which is helpful.
Appreciate
> > > > > > > your dedication so late into the evening! Can I add new
> > > > > > > variables in the
> > > > > > config
> > > > > > > file? Any info on the science behind the interpolation
> > > > > > > methods would be appreciated, but that can wait until
> > > > > > > Monday.
> > > > > > >
> > > > > > > Take care and stay safe.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
MET
> > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify
> > > > > > the
> > > > > > > identity of the sender, and confirm the authenticity of
all
> > > > > > > links
> > > > > > contained
> > > > > > > within the message prior to copying and pasting the
address
> > > > > > > to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > John,
> > > > > > >
> > > > > > > I see you have a question about interpolation options in
> > > > > > > the
> > > > > > > point2grid tool. This XML file contains the calls to the
> > > > > > > point2grid tool that are
> > > > > > run
> > > > > > > for our unit tests:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > un
> > > > > > it
> > > > > > _point2grid.xml
> > > > > > >
> > > > > > > While most of them DO NOT use the "-method" command line
> > > > > > > option, a few at the end do use "-method MAX" to select
the
> > > > > > > maximum value falling within each grid box.
> > > > > > >
> > > > > > > Inspecting the code in this file:
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
> > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > oo
> > > > > > ls
> > > > > > /other/point2grid/point2grid.cc
> > > > > > > I see that the supported options for -method are MIN,
MAX,
> > > > > > > MEDIAN, and UW_MEAN.
> > > > > > >
> > > > > > > Seems like we should add that info to the usage
statement!
> > > > > > > I
> > > > > > > wrote up
> > > > > > this
> > > > > > > GitHub issue to make sure we remember to do so before
the
> > > > > > > 9.1
> > > > > > > release:
> > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > Caution-Caution-Caution-https://github.com/NCAR/MET/issu
> > > > > > > es/1396
> > > > > > >
> > > > > > > Hope that helps clarify.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV via
RT
> > > > > > > < met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
upon.
> > > > > > > > Transaction: Ticket created by
john.w.raby2.civ at mail.mil
> > > > > > > >        Queue: met_help
> > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >        Owner: Nobody
> > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > >       Status: new
> > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > Caution-url:
> > > > > > > > Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > ?id=95747
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > The optional argument for smoothing or interpolation
> > > > > > > > shown
> > > > > > > > in the
> > > > > > User's
> > > > > > > > Guide is Gaussian, but are other methods available? If
> > > > > > > > so,
> > > > > > > > can you
> > > > > > > provide
> > > > > > > > the names and how to specify them as an optional
> > > > > > > > argument?
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > Stay safe.
> > > > > > > >
> > > > > > > > T/
> > > > > > > > John
> > > > > > > >
> > > > > > > > Mr. John W. Raby
> > > > > > > > U.S. Army Research Laboratory White Sands Missile
Range,
> > > > > > > > NM
> > > > > > > > 88002
> > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > Teleworking: Contact Admin Specialist at (410) 278-
6974
> > > > > > > > FAX
> > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED
>
>
>
> CLASSIFICATION: UNCLASSIFIED





------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Wed Jul 15 12:17:03 2020

Yes, you are right. It's not the output you are expecting. We will
improve the documentation on this
(https://github.com/NCAR/MET/issues/1413).

Unlike regrid_data_plane, the MAXGAUSE method with point2grid works
differently: MAX interpolation for the raw data (TMP) and GAUSSIAN
smoothing into the probability variable (TMP_prob_NA or TMP_prob_XX).
MAXGAUSS is the same as the MAX interpolation for the TMP variable for
the point2grid. So the difference is between MAX VS. UW_MEAN
interpolation methods, and GAUSIAN smoothing was not involved.

The right way for the Gaussian smoothing to the raw data will be:
1) regridding by using point2grid with one of interpolation methods
(MAX/MIN/UW_MEAN/MEDIAN)
2) regridding by using regrid_data_plane with MAXGAUSS method and the
same setting as the input and the target grid.

The output of regrid_data_plane (MAXGAUSS) is not right. I'm working
on this.

Cheers,
Howard


On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> Howard -
>
> Today I successfully ran point2grid but used -method MAXGAUSS and it
> ran without error and produced output. I'm uploading the file list,
> output, commands used, log file, and plots to the DoD SAFE so you
can
> take a look. I am also uploading the results of the previous run
where
> you suggested that I use "11" vice "TMP" which fixed the problem and
> ran successfully.
>
> The plots are not what I was expecting somehow. The plotted data
look
> like just points where the original obs were located rather than a
> regularly spaced grid of TMP values. What are your thoughts on this?
>
> Thanks
> ________________________________________
> From: Howard Soh via RT [met_help at ucar.edu]
> Sent: Monday, July 13, 2020 3:33 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Something is not right. I will take a look
>
> DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is 246.351
> to 284.817.
> DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is
-9999
> to -9999.
>
> The input is OK (between 246.351 and 284.817), but the regrided
output
> is all missing value (-9999).
>
> Cheers,
> Howard
>
> The input file has
> On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > The next run with MAXGAUSS was successful, although I have to plot
it
> > to see
> > what it looks like. I've attached the log file. Does it indicate
any
> > issues
> > with how it ran?
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, July 13, 2020 1:14 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS
and
> > GAUSSIAN
> > method is only used for the interpolation. Please set "-method
> > MAXGAUSE". The
> > from_grid and to_grid are the same, there is no difference with
> > additional MAX
> > interpolation.
> >
> > Cheers,
> > Howard
> >
> >
> > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > I tried running MET V9.0 regrid_data_plane following your
example
> > > below, but I'm getting and error:
> > > ERROR  :
> > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > > regridding, only interpolating.
> > >
> > > The following is the command I used to run it:
> > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > V2016.nc
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > V2016.nc
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > ss_test
> > > -v 9
> > >
> > > Any suggestions on how to perform Gaussian smoothing on the
output
> > > TMP
> > > field from point2grid?
> > >
> > > Thanks.
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, July 6, 2020 11:38 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > > the identity of the sender, and confirm the authenticity of all
> > > links
> > > contained within the message prior to copying and pasting the
> > > address
> > > to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Sorry for the late response.
> > >
> > > The input and to_grid must be the same for Gaussian smoothing
and
> > > regrid_data_plane shows the usage message if the to_grid is
> > > missing.
> > > There are
> > > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> > >
> > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> > >
> > > Cheers,
> > > Howard
> > >
> > >
> > >
> > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > Howard -
> > > >
> > > > Thanks for tailoring the command for my particular interests.
> > > > Just
> > > > curious about your regrid_data_plane command. You provided the
> > > > input
> > > > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > > > regrid_data_plane and then you provide the name of the file to
be
> > > > output from regrid_data_plane (pb2nc_TMP.nc
> > > > pb2nc_gaussian_TMP.nc),
> > > > but you didn't provide the to_grid. Is this because it knows
that
> > > > the grids of the input and output files are the same and thus
> > > > ignores the missing to_grid argument?
> > > >
> > > > R/
> > > > John
> > > >
> > > > -----Original Message-----
> > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > <john.w.raby2.civ at mail.mil>
> > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting
> > > > the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > The Gaussian smoothing is available with regrid_data_plane,
too.
> > > > In
> > > > this case, we are interested in Gaussian smoothing without
> > > > regriding.
> > > > The input and the target grid should be the same, which is the
> > > > output of point2grid.
> > > >
> > > > Here is an example (assuming the point data file is generated
by
> > > > pb2nc).
> > > >
> > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > >       G212 pb2nc_TMP.nc \
> > > >       -field 'name="TMP"; level="*";' \
> > > >       -v 1
> > > >
> > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > pb2nc_gaussian_TMP.nc
> > > >       -field 'name="TMP"; level="(*,*)";' \
> > > >       -method GAUSSIAN -v 1
> > > >
> > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
(120.0)
> > > > if
> > > > necessary
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Howard -
> > > > >
> > > > > Thanks for your quick response with details in answer to my
> > > > > question.
> > > > > Right now my interest is with non-probability fields, so
your
> > > > > point in 1. about the possibility of using Gaussian
smoothing
> > > > > for
> > > > > -field argument is interesting.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of
> > > > > all links contained within the message prior to copying and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Two layers for the interpolation with point2grid:
> > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw
> > > > > data
> > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
probability
> > > > > variable. (the existence of valid values if the probability
> > > > > threshold is not given)
> > > > >
> > > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
> > > > > (which
> > > > > is
> > > > > a
> > > > > default) for the layer 1 and "-method GAUSSIAN" for the
layer
> > > > > 2.
> > > > > + "-method MAXGAUSS" is processed as the "-method MAX" for
the
> > > > > layer
> > > > > 1
> > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > + Override the interpolation method of layer 1 with "-method
> > > > > <METHOD>"
> > > > >
> > > > > 1. The Gaussian smoothing at point2grid is activated and
> > > > > applied
> > > > > into the probability variable only if -method GAUSSIAN or
> > > > > -method
> > > > > MAXGAUSS is given as the command line argument. It's not
> > > > > applied
> > > > > into the variable at "-field" argument. If you want, You can
do
> > > > > this by using regrid_data_plane (passing the output of
> > > > > point2grid
> > > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > > >
> > > > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default
values
> > > > > are
> > > > > applied if not given.
> > > > >
> > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
> > > > > -method
> > > > > MAXGAUSS argument.
> > > > >
> > > > > 4 Without -prob_cat_thresh argument, it will be the status
of
> > > > > data
> > > > > existence. With -prob_cat_thresh argument, it will be
meeting
> > > > > the
> > > > > condition or not. So the values of the probability are 0 or
1.
> > > > >
> > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > > requested
> > > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > John,
> > > > > >
> > > > > > I'm reassigning this ticket to Howard Soh who worked on
the
> > > > > > development of the point2grid tool. He should be able to
> > > > > > better
> > > > > > answer these questions.
> > > > >
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via
RT
> > > > > > <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <Caution-Caution-Caution-Caution-Caution-url: Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > > d=95747 >
> > > > > > >
> > > > > > > John -
> > > > > > >
> > > > > > > I ran it to see the usage and there are details in the
> > > > > > > usage
> > > > > > > dialog and also in the XML file you pointed me to which
> > > > > > > helped
> > > > > > > to answer my own question (I think).
> > > > > > > There is an additional interpolation method called
GAUSSIAN
> > > > > > > which applies only when using the probability variable
(-
> > > > > > > prob_cat_thresh 280.0).
> > > > > > > So, it
> > > > > > > looks like the two optional smoothing arguments
> > > > > > > -gaussian_dx
> > > > > > > and -gaussian_radius are only used when the method is
> > > > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > > > smoothing except when using probability variables. Is
this
> > > > > > > correct?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > MET
> > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
> > > > > > > prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > John,
> > > > > > >
> > > > > > > While the "-field" command line argument indicates that
it
> > > > > > > can
> > > > > > > be used one or more times, in this first version of the
> > > > > > > tool,
> > > > > > > it can only be used once.
> > > > > > >
> > > > > > > If you specify "-field" more that once you'll get this
> > > > > > > error
> > > > > > > message:
> > > > > > > ERROR  :
> > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > "2"
> > > > > > > variables
> > > > > > > were requested.
> > > > > > > ERROR  :
> > > > > > >
> > > > > > > We do plan to support the processing of multiple fields
but
> > > > > > > it
> > > > > > > has not yet been implemented.
> > > > > > >
> > > > > > > As for the science behind the interpolation methods,
it's
> > > > > > > very
> > > > > > > simple. Keep track of the point observations falling
inside
> > > > > > > each of the target grid boxes. Write to the output the
min,
> > > > > > > max, unweighted mean, or median of those values based on
> > > > > > > the
> > > > > > > method chosen. If you only have 1 point obs falling
inside
> > > > > > > each grid box, then all those methods will result in the
> > > > > > > same
> > > > > > > output value.
> > > > > > > Whereas if you have many, as in the case of GOES input
> > > > > > > files,
> > > > > > > then the choice of the derivation method would matter.
> > > > > > >
> > > > > > > Hope that helps clarify.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > >
> > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via
RT
> > > > > > > <
> > > > > > > met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > ?id=
> > > > > > > 95747 >
> > > > > > > >
> > > > > > > > Another question I have is how is -method used in this
> > > > > > > > tool?
> > > > > > > > In the
> > > > > > > User's
> > > > > > > > Guide, -method provides options for regridding from
one
> > > > > > > > grid
> > > > > > > > to another such as is used in regrid_data_plane. In
the
> > > > > > > > case
> > > > > > > > of Point2Grid, the process involved is not regridding
in
> > > > > > > > the
> > > > > > > > way that I understand it.
> > > > > > > > If I
> > > > > > > > understand it correctly, it takes point observations
and
> > > > > > > > distributes them on to a user provided grid using some
> > > > > > > > interpolation method. The next
> > > > > > > step I
> > > > > > > > think is the smoothing of the field to even out
> > > > > > > > irregularities which are caused by errors, noise, etc
> > > > > > > > resulting in the fitting of the observed values to
some
> > > > > > > > function which appear to be a Gaussian function. I
guess
> > > > > > > > I
> > > > > > > > also need to understand what happens when the tool is
> > > > > > > > used
> > > > > > > > without any of the optional arguments.
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > To: met_help at ucar.edu
> > > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > #95747]
> > > > > > > > MET
> > > > > > > > V9.0
> > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > John -
> > > > > > > >
> > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > Appreciate
> > > > > > > > your dedication so late into the evening! Can I add
new
> > > > > > > > variables in the
> > > > > > > config
> > > > > > > > file? Any info on the science behind the interpolation
> > > > > > > > methods would be appreciated, but that can wait until
> > > > > > > > Monday.
> > > > > > > >
> > > > > > > > Take care and stay safe.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > > > MET
> > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify
> > > > > > > the
> > > > > > > > identity of the sender, and confirm the authenticity
of
> > > > > > > > all
> > > > > > > > links
> > > > > > > contained
> > > > > > > > within the message prior to copying and pasting the
> > > > > > > > address
> > > > > > > > to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > John,
> > > > > > > >
> > > > > > > > I see you have a question about interpolation options
in
> > > > > > > > the
> > > > > > > > point2grid tool. This XML file contains the calls to
the
> > > > > > > > point2grid tool that are
> > > > > > > run
> > > > > > > > for our unit tests:
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > > un
> > > > > > > it
> > > > > > > _point2grid.xml
> > > > > > > >
> > > > > > > > While most of them DO NOT use the "-method" command
line
> > > > > > > > option, a few at the end do use "-method MAX" to
select
> > > > > > > > the
> > > > > > > > maximum value falling within each grid box.
> > > > > > > >
> > > > > > > > Inspecting the code in this file:
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > > oo
> > > > > > > ls
> > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > I see that the supported options for -method are MIN,
> > > > > > > > MAX,
> > > > > > > > MEDIAN, and UW_MEAN.
> > > > > > > >
> > > > > > > > Seems like we should add that info to the usage
> > > > > > > > statement!
> > > > > > > > I
> > > > > > > > wrote up
> > > > > > > this
> > > > > > > > GitHub issue to make sure we remember to do so before
the
> > > > > > > > 9.1
> > > > > > > > release:
> > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > Caution-Caution-Caution-
https://github.com/NCAR/MET/issu
> > > > > > > > es/1396
> > > > > > > >
> > > > > > > > Hope that helps clarify.
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV
via
> > > > > > > > RT
> > > > > > > > < met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
upon.
> > > > > > > > > Transaction: Ticket created by
> > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > >        Queue: met_help
> > > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > >        Owner: Nobody
> > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > >       Status: new
> > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-url:
> > > > > > > > > Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > ?id=95747
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > > The optional argument for smoothing or interpolation
> > > > > > > > > shown
> > > > > > > > > in the
> > > > > > > User's
> > > > > > > > > Guide is Gaussian, but are other methods available?
If
> > > > > > > > > so,
> > > > > > > > > can you
> > > > > > > > provide
> > > > > > > > > the names and how to specify them as an optional
> > > > > > > > > argument?
> > > > > > > > >
> > > > > > > > > Thanks.
> > > > > > > > >
> > > > > > > > > Stay safe.
> > > > > > > > >
> > > > > > > > > T/
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > Mr. John W. Raby
> > > > > > > > > U.S. Army Research Laboratory White Sands Missile
> > > > > > > > > Range,
> > > > > > > > > NM
> > > > > > > > > 88002
> > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > Teleworking: Contact Admin Specialist at (410) 278-
6974
> > > > > > > > > FAX
> > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: RE: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Wed Jul 15 13:10:58 2020

Thanks for working to improve the documentation. So you can only use
the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid unless
you are working with a probability variable in which case you can use
MAXGAUSS. The default of these is UW_MEAN. The output I generated in
my first test used that method since I didn't specify the -method. In
this case, the plot I generated using plot_data_plane shows a map with
data points at random locations which correspond to the locations of
the original point obs from the output of ascii2nc. My expectation was
that I would see the data points with TMP values in a regularly spaced
grid with 1km spacing. In theory, this NetCDF file could be used in
Grid-Stat to verify a TMP fcst on the same 1km grid. Why doesn't the
plot show a gridded TMP field?

R/
John


________________________________________
From: Howard Soh via RT [met_help at ucar.edu]
Sent: Wednesday, July 15, 2020 12:17 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

Yes, you are right. It's not the output you are expecting. We will
improve the documentation on this (Caution-
https://github.com/NCAR/MET/issues/1413).

Unlike regrid_data_plane, the MAXGAUSE method with point2grid works
differently: MAX interpolation for the raw data (TMP) and GAUSSIAN
smoothing into the probability variable (TMP_prob_NA or TMP_prob_XX).
MAXGAUSS is the same as the MAX interpolation for the TMP variable for
the point2grid. So the difference is between MAX VS. UW_MEAN
interpolation methods, and GAUSIAN smoothing was not involved.

The right way for the Gaussian smoothing to the raw data will be:
1) regridding by using point2grid with one of interpolation methods
(MAX/MIN/UW_MEAN/MEDIAN)
2) regridding by using regrid_data_plane with MAXGAUSS method and the
same setting as the input and the target grid.

The output of regrid_data_plane (MAXGAUSS) is not right. I'm working
on this.

Cheers,
Howard


On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> Howard -
>
> Today I successfully ran point2grid but used -method MAXGAUSS and it
> ran without error and produced output. I'm uploading the file list,
> output, commands used, log file, and plots to the DoD SAFE so you
can
> take a look. I am also uploading the results of the previous run
where
> you suggested that I use "11" vice "TMP" which fixed the problem and
> ran successfully.
>
> The plots are not what I was expecting somehow. The plotted data
look
> like just points where the original obs were located rather than a
> regularly spaced grid of TMP values. What are your thoughts on this?
>
> Thanks
> ________________________________________
> From: Howard Soh via RT [met_help at ucar.edu]
> Sent: Monday, July 13, 2020 3:33 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Something is not right. I will take a look
>
> DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is 246.351
> to 284.817.
> DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is
-9999
> to -9999.
>
> The input is OK (between 246.351 and 284.817), but the regrided
output
> is all missing value (-9999).
>
> Cheers,
> Howard
>
> The input file has
> On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > CLASSIFICATION: UNCLASSIFIED
> >
> > Howard -
> >
> > The next run with MAXGAUSS was successful, although I have to plot
it
> > to see
> > what it looks like. I've attached the log file. Does it indicate
any
> > issues
> > with how it ran?
> >
> > R/
> > John
> >
> > -----Original Message-----
> > From: Howard Soh via RT <met_help at ucar.edu>
> > Sent: Monday, July 13, 2020 1:14 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > <john.w.raby2.civ at mail.mil>
> >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to MAXGAUSS
and
> > GAUSSIAN
> > method is only used for the interpolation. Please set "-method
> > MAXGAUSE". The
> > from_grid and to_grid are the same, there is no difference with
> > additional MAX
> > interpolation.
> >
> > Cheers,
> > Howard
> >
> >
> > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > I tried running MET V9.0 regrid_data_plane following your
example
> > > below, but I'm getting and error:
> > > ERROR  :
> > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > > regridding, only interpolating.
> > >
> > > The following is the command I used to run it:
> > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > V2016.nc
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > V2016.nc
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > ss_test
> > > -v 9
> > >
> > > Any suggestions on how to perform Gaussian smoothing on the
output
> > > TMP
> > > field from point2grid?
> > >
> > > Thanks.
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, July 6, 2020 11:38 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > > the identity of the sender, and confirm the authenticity of all
> > > links
> > > contained within the message prior to copying and pasting the
> > > address
> > > to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Sorry for the late response.
> > >
> > > The input and to_grid must be the same for Gaussian smoothing
and
> > > regrid_data_plane shows the usage message if the to_grid is
> > > missing.
> > > There are
> > > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> > >
> > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
pb2nc_gaussian_TMP.nc
> > > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> > >
> > > Cheers,
> > > Howard
> > >
> > >
> > >
> > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > Howard -
> > > >
> > > > Thanks for tailoring the command for my particular interests.
> > > > Just
> > > > curious about your regrid_data_plane command. You provided the
> > > > input
> > > > file from point2grid (pb2nc_TMP.nc ) as the input_filename to
> > > > regrid_data_plane and then you provide the name of the file to
be
> > > > output from regrid_data_plane (pb2nc_TMP.nc
> > > > pb2nc_gaussian_TMP.nc),
> > > > but you didn't provide the to_grid. Is this because it knows
that
> > > > the grids of the input and output files are the same and thus
> > > > ignores the missing to_grid argument?
> > > >
> > > > R/
> > > > John
> > > >
> > > > -----Original Message-----
> > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > <john.w.raby2.civ at mail.mil>
> > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of
> > > > all links contained within the message prior to copying and
> > > > pasting
> > > > the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > The Gaussian smoothing is available with regrid_data_plane,
too.
> > > > In
> > > > this case, we are interested in Gaussian smoothing without
> > > > regriding.
> > > > The input and the target grid should be the same, which is the
> > > > output of point2grid.
> > > >
> > > > Here is an example (assuming the point data file is generated
by
> > > > pb2nc).
> > > >
> > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > >       G212 pb2nc_TMP.nc \
> > > >       -field 'name="TMP"; level="*";' \
> > > >       -v 1
> > > >
> > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > pb2nc_gaussian_TMP.nc
> > > >       -field 'name="TMP"; level="(*,*)";' \
> > > >       -method GAUSSIAN -v 1
> > > >
> > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
(120.0)
> > > > if
> > > > necessary
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Howard -
> > > > >
> > > > > Thanks for your quick response with details in answer to my
> > > > > question.
> > > > > Right now my interest is with non-probability fields, so
your
> > > > > point in 1. about the possibility of using Gaussian
smoothing
> > > > > for
> > > > > -field argument is interesting.
> > > > >
> > > > > R/
> > > > > John
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of
> > > > > all links contained within the message prior to copying and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Two layers for the interpolation with point2grid:
> > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the raw
> > > > > data
> > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
probability
> > > > > variable. (the existence of valid values if the probability
> > > > > threshold is not given)
> > > > >
> > > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
> > > > > (which
> > > > > is
> > > > > a
> > > > > default) for the layer 1 and "-method GAUSSIAN" for the
layer
> > > > > 2.
> > > > > + "-method MAXGAUSS" is processed as the "-method MAX" for
the
> > > > > layer
> > > > > 1
> > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > + Override the interpolation method of layer 1 with "-method
> > > > > <METHOD>"
> > > > >
> > > > > 1. The Gaussian smoothing at point2grid is activated and
> > > > > applied
> > > > > into the probability variable only if -method GAUSSIAN or
> > > > > -method
> > > > > MAXGAUSS is given as the command line argument. It's not
> > > > > applied
> > > > > into the variable at "-field" argument. If you want, You can
do
> > > > > this by using regrid_data_plane (passing the output of
> > > > > point2grid
> > > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > > >
> > > > > 2. -gaussian_dx and -gaussian_radius are ignored if -method
> > > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default
values
> > > > > are
> > > > > applied if not given.
> > > > >
> > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
> > > > > -method
> > > > > MAXGAUSS argument.
> > > > >
> > > > > 4 Without -prob_cat_thresh argument, it will be the status
of
> > > > > data
> > > > > existence. With -prob_cat_thresh argument, it will be
meeting
> > > > > the
> > > > > condition or not. So the values of the probability are 0 or
1.
> > > > >
> > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > > requested
> > > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > John,
> > > > > >
> > > > > > I'm reassigning this ticket to Howard Soh who worked on
the
> > > > > > development of the point2grid tool. He should be able to
> > > > > > better
> > > > > > answer these questions.
> > > > >
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV via
RT
> > > > > > <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-url:
Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > > d=95747 >
> > > > > > >
> > > > > > > John -
> > > > > > >
> > > > > > > I ran it to see the usage and there are details in the
> > > > > > > usage
> > > > > > > dialog and also in the XML file you pointed me to which
> > > > > > > helped
> > > > > > > to answer my own question (I think).
> > > > > > > There is an additional interpolation method called
GAUSSIAN
> > > > > > > which applies only when using the probability variable
(-
> > > > > > > prob_cat_thresh 280.0).
> > > > > > > So, it
> > > > > > > looks like the two optional smoothing arguments
> > > > > > > -gaussian_dx
> > > > > > > and -gaussian_radius are only used when the method is
> > > > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > > > smoothing except when using probability variables. Is
this
> > > > > > > correct?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > MET
> > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
> > > > > > > prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > John,
> > > > > > >
> > > > > > > While the "-field" command line argument indicates that
it
> > > > > > > can
> > > > > > > be used one or more times, in this first version of the
> > > > > > > tool,
> > > > > > > it can only be used once.
> > > > > > >
> > > > > > > If you specify "-field" more that once you'll get this
> > > > > > > error
> > > > > > > message:
> > > > > > > ERROR  :
> > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > "2"
> > > > > > > variables
> > > > > > > were requested.
> > > > > > > ERROR  :
> > > > > > >
> > > > > > > We do plan to support the processing of multiple fields
but
> > > > > > > it
> > > > > > > has not yet been implemented.
> > > > > > >
> > > > > > > As for the science behind the interpolation methods,
it's
> > > > > > > very
> > > > > > > simple. Keep track of the point observations falling
inside
> > > > > > > each of the target grid boxes. Write to the output the
min,
> > > > > > > max, unweighted mean, or median of those values based on
> > > > > > > the
> > > > > > > method chosen. If you only have 1 point obs falling
inside
> > > > > > > each grid box, then all those methods will result in the
> > > > > > > same
> > > > > > > output value.
> > > > > > > Whereas if you have many, as in the case of GOES input
> > > > > > > files,
> > > > > > > then the choice of the derivation method would matter.
> > > > > > >
> > > > > > > Hope that helps clarify.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > >
> > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV via
RT
> > > > > > > <
> > > > > > > met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > > Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
> > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > ?id=
> > > > > > > 95747 >
> > > > > > > >
> > > > > > > > Another question I have is how is -method used in this
> > > > > > > > tool?
> > > > > > > > In the
> > > > > > > User's
> > > > > > > > Guide, -method provides options for regridding from
one
> > > > > > > > grid
> > > > > > > > to another such as is used in regrid_data_plane. In
the
> > > > > > > > case
> > > > > > > > of Point2Grid, the process involved is not regridding
in
> > > > > > > > the
> > > > > > > > way that I understand it.
> > > > > > > > If I
> > > > > > > > understand it correctly, it takes point observations
and
> > > > > > > > distributes them on to a user provided grid using some
> > > > > > > > interpolation method. The next
> > > > > > > step I
> > > > > > > > think is the smoothing of the field to even out
> > > > > > > > irregularities which are caused by errors, noise, etc
> > > > > > > > resulting in the fitting of the observed values to
some
> > > > > > > > function which appear to be a Gaussian function. I
guess
> > > > > > > > I
> > > > > > > > also need to understand what happens when the tool is
> > > > > > > > used
> > > > > > > > without any of the optional arguments.
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > To: met_help at ucar.edu
> > > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > #95747]
> > > > > > > > MET
> > > > > > > > V9.0
> > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > John -
> > > > > > > >
> > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > Appreciate
> > > > > > > > your dedication so late into the evening! Can I add
new
> > > > > > > > variables in the
> > > > > > > config
> > > > > > > > file? Any info on the science behind the interpolation
> > > > > > > > methods would be appreciated, but that can wait until
> > > > > > > > Monday.
> > > > > > > >
> > > > > > > > Take care and stay safe.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747]
> > > > > > > > MET
> > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify
> > > > > > > the
> > > > > > > > identity of the sender, and confirm the authenticity
of
> > > > > > > > all
> > > > > > > > links
> > > > > > > contained
> > > > > > > > within the message prior to copying and pasting the
> > > > > > > > address
> > > > > > > > to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > John,
> > > > > > > >
> > > > > > > > I see you have a question about interpolation options
in
> > > > > > > > the
> > > > > > > > point2grid tool. This XML file contains the calls to
the
> > > > > > > > point2grid tool that are
> > > > > > > run
> > > > > > > > for our unit tests:
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > > un
> > > > > > > it
> > > > > > > _point2grid.xml
> > > > > > > >
> > > > > > > > While most of them DO NOT use the "-method" command
line
> > > > > > > > option, a few at the end do use "-method MAX" to
select
> > > > > > > > the
> > > > > > > > maximum value falling within each grid box.
> > > > > > > >
> > > > > > > > Inspecting the code in this file:
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
> > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > > oo
> > > > > > > ls
> > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > I see that the supported options for -method are MIN,
> > > > > > > > MAX,
> > > > > > > > MEDIAN, and UW_MEAN.
> > > > > > > >
> > > > > > > > Seems like we should add that info to the usage
> > > > > > > > statement!
> > > > > > > > I
> > > > > > > > wrote up
> > > > > > > this
> > > > > > > > GitHub issue to make sure we remember to do so before
the
> > > > > > > > 9.1
> > > > > > > > release:
> > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > Caution-Caution-Caution-Caution-
https://github.com/NCAR/MET/issu
> > > > > > > > es/1396
> > > > > > > >
> > > > > > > > Hope that helps clarify.
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV
via
> > > > > > > > RT
> > > > > > > > < met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
upon.
> > > > > > > > > Transaction: Ticket created by
> > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > >        Queue: met_help
> > > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > >        Owner: Nobody
> > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > >       Status: new
> > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-url:
> > > > > > > > > Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > ?id=95747
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > > The optional argument for smoothing or interpolation
> > > > > > > > > shown
> > > > > > > > > in the
> > > > > > > User's
> > > > > > > > > Guide is Gaussian, but are other methods available?
If
> > > > > > > > > so,
> > > > > > > > > can you
> > > > > > > > provide
> > > > > > > > > the names and how to specify them as an optional
> > > > > > > > > argument?
> > > > > > > > >
> > > > > > > > > Thanks.
> > > > > > > > >
> > > > > > > > > Stay safe.
> > > > > > > > >
> > > > > > > > > T/
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > Mr. John W. Raby
> > > > > > > > > U.S. Army Research Laboratory White Sands Missile
> > > > > > > > > Range,
> > > > > > > > > NM
> > > > > > > > > 88002
> > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > Teleworking: Contact Admin Specialist at (410) 278-
6974
> > > > > > > > > FAX
> > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> > CLASSIFICATION: UNCLASSIFIED





------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Howard Soh
Time: Wed Jul 15 14:54:51 2020

The plot_data_plane can not show the point observation data. It shows
the plot of the grid data.
In this case, the coverage of the gridded data is 0.23% (97 out of
41616). It's the plot of the grid cells (1km by 1km) even though it
looks like the point data. It caused by few converage (0.23%).

The meaning of following log mesages (from point2grid):

DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG 2: process_point_file() ->  var_count=180, grid: 97 out of 41616

- The input data has 180 TMP data
- One TMP data does not belong to the target grid
- 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616 * 100)
  ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP data

The plot has valid data for 97 cells out of 41616 (which is 204 by
204) grid cells)


Two issues with regrid_data_plane (this specific case because of 0.23%
coverage).

1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR 0.008
(<= 1 / 121).

The default setting of "-vld_thresh" is 0.5.
The max raduis (MAX_R) for the Gaussuan smoothing is 5. The matrix for
Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a valid
regrdded value for the taeget cell, it requires 61 (from 121 / 2)
neighbers with the valid TMP data. Only 97 cells have the valid data
and they are spreaded. None of them satisfies the number of the valid
neighber cells. That's why the output of regrid_data_plae was all
missing values.

2. Bug on computing the Gaussian value.

I will fix this.

I attached the updated NetCDF file (output of regrid_data_plane) and
the plot.

Cheers,
Howard

On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> Thanks for working to improve the documentation. So you can only use
> the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
unless
> you are working with a probability variable in which case you can
use
> MAXGAUSS. The default of these is UW_MEAN. The output I generated in
> my first test used that method since I didn't specify the -method.
In
> this case, the plot I generated using plot_data_plane shows a map
with
> data points at random locations which correspond to the locations of
> the original point obs from the output of ascii2nc. My expectation
was
> that I would see the data points with TMP values in a regularly
spaced
> grid with 1km spacing. In theory, this NetCDF file could be used in
> Grid-Stat to verify a TMP fcst on the same 1km grid. Why doesn't the
> plot show a gridded TMP field?
>
> R/
> John
>
>
> ________________________________________
> From: Howard Soh via RT [met_help at ucar.edu]
> Sent: Wednesday, July 15, 2020 12:17 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify
> the identity of the sender, and confirm the authenticity of all
links
> contained within the message prior to copying and pasting the
address
> to a Web browser.
>
>
>
>
> ----
>
> Yes, you are right. It's not the output you are expecting. We will
> improve the documentation on this (Caution-
> https://github.com/NCAR/MET/issues/1413).
>
> Unlike regrid_data_plane, the MAXGAUSE method with point2grid works
> differently: MAX interpolation for the raw data (TMP) and GAUSSIAN
> smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> MAXGAUSS is the same as the MAX interpolation for the TMP variable
for
> the point2grid. So the difference is between MAX VS. UW_MEAN
> interpolation methods, and GAUSIAN smoothing was not involved.
>
> The right way for the Gaussian smoothing to the raw data will be:
> 1) regridding by using point2grid with one of interpolation methods
> (MAX/MIN/UW_MEAN/MEDIAN)
> 2) regridding by using regrid_data_plane with MAXGAUSS method and
the
> same setting as the input and the target grid.
>
> The output of regrid_data_plane (MAXGAUSS) is not right. I'm working
> on this.
>
> Cheers,
> Howard
>
>
> On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > Howard -
> >
> > Today I successfully ran point2grid but used -method MAXGAUSS and
it
> > ran without error and produced output. I'm uploading the file
list,
> > output, commands used, log file, and plots to the DoD SAFE so you
can
> > take a look. I am also uploading the results of the previous run
> > where
> > you suggested that I use "11" vice "TMP" which fixed the problem
and
> > ran successfully.
> >
> > The plots are not what I was expecting somehow. The plotted data
look
> > like just points where the original obs were located rather than a
> > regularly spaced grid of TMP values. What are your thoughts on
this?
> >
> > Thanks
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Monday, July 13, 2020 3:33 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Something is not right. I will take a look
> >
> > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
246.351
> > to 284.817.
> > DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is
> > -9999
> > to -9999.
> >
> > The input is OK (between 246.351 and 284.817), but the regrided
> > output
> > is all missing value (-9999).
> >
> > Cheers,
> > Howard
> >
> > The input file has
> > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > Howard -
> > >
> > > The next run with MAXGAUSS was successful, although I have to
plot
> > > it
> > > to see
> > > what it looks like. I've attached the log file. Does it indicate
> > > any
> > > issues
> > > with how it ran?
> > >
> > > R/
> > > John
> > >
> > > -----Original Message-----
> > > From: Howard Soh via RT <met_help at ucar.edu>
> > > Sent: Monday, July 13, 2020 1:14 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > <john.w.raby2.civ at mail.mil>
> > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > > the
> > > identity of the sender, and confirm the authenticity of all
links
> > > contained
> > > within the message prior to copying and pasting the address to a
> > > Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
MAXGAUSS
> > > and
> > > GAUSSIAN
> > > method is only used for the interpolation. Please set "-method
> > > MAXGAUSE". The
> > > from_grid and to_grid are the same, there is no difference with
> > > additional MAX
> > > interpolation.
> > >
> > > Cheers,
> > > Howard
> > >
> > >
> > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > Howard -
> > > >
> > > > I tried running MET V9.0 regrid_data_plane following your
example
> > > > below, but I'm getting and error:
> > > > ERROR  :
> > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > > > regridding, only interpolating.
> > > >
> > > > The following is the command I used to run it:
> > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > V2016.nc
> > > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > V2016.nc
> > > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> > > >
> > > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > ss_test
> > > > -v 9
> > > >
> > > > Any suggestions on how to perform Gaussian smoothing on the
> > > > output
> > > > TMP
> > > > field from point2grid?
> > > >
> > > > Thanks.
> > > >
> > > > R/
> > > > John
> > > >
> > > > -----Original Message-----
> > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > <john.w.raby2.civ at mail.mil>
> > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify
> > > > the identity of the sender, and confirm the authenticity of
all
> > > > links
> > > > contained within the message prior to copying and pasting the
> > > > address
> > > > to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Sorry for the late response.
> > > >
> > > > The input and to_grid must be the same for Gaussian smoothing
and
> > > > regrid_data_plane shows the usage message if the to_grid is
> > > > missing.
> > > > There are
> > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> > > >
> > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > pb2nc_gaussian_TMP.nc
> > > > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > >
> > > >
> > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > Howard -
> > > > >
> > > > > Thanks for tailoring the command for my particular
interests.
> > > > > Just
> > > > > curious about your regrid_data_plane command. You provided
the
> > > > > input
> > > > > file from point2grid (pb2nc_TMP.nc ) as the input_filename
to
> > > > > regrid_data_plane and then you provide the name of the file
to
> > > > > be
> > > > > output from regrid_data_plane (pb2nc_TMP.nc
> > > > > pb2nc_gaussian_TMP.nc),
> > > > > but you didn't provide the to_grid. Is this because it knows
> > > > > that
> > > > > the grids of the input and output files are the same and
thus
> > > > > ignores the missing to_grid argument?
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > > -----Original Message-----
> > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > <john.w.raby2.civ at mail.mil>
> > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of
> > > > > all links contained within the message prior to copying and
> > > > > pasting
> > > > > the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > The Gaussian smoothing is available with regrid_data_plane,
> > > > > too.
> > > > > In
> > > > > this case, we are interested in Gaussian smoothing without
> > > > > regriding.
> > > > > The input and the target grid should be the same, which is
the
> > > > > output of point2grid.
> > > > >
> > > > > Here is an example (assuming the point data file is
generated
> > > > > by
> > > > > pb2nc).
> > > > >
> > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > >       G212 pb2nc_TMP.nc \
> > > > >       -field 'name="TMP"; level="*";' \
> > > > >       -v 1
> > > > >
> > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > pb2nc_gaussian_TMP.nc
> > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > >       -method GAUSSIAN -v 1
> > > > >
> > > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
(120.0)
> > > > > if
> > > > > necessary
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > Howard -
> > > > > >
> > > > > > Thanks for your quick response with details in answer to
my
> > > > > > question.
> > > > > > Right now my interest is with non-probability fields, so
your
> > > > > > point in 1. about the possibility of using Gaussian
smoothing
> > > > > > for
> > > > > > -field argument is interesting.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > > ________________________________________
> > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please
> > > > > > verify the identity of the sender, and confirm the
> > > > > > authenticity
> > > > > > of
> > > > > > all links contained within the message prior to copying
and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Two layers for the interpolation with point2grid:
> > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the
raw
> > > > > > data
> > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > probability
> > > > > > variable. (the existence of valid values if the
probability
> > > > > > threshold is not given)
> > > > > >
> > > > > > + "-method GAUSSIAN" is processed as the "-method UW_MEAN"
> > > > > > (which
> > > > > > is
> > > > > > a
> > > > > > default) for the layer 1 and "-method GAUSSIAN" for the
layer
> > > > > > 2.
> > > > > > + "-method MAXGAUSS" is processed as the "-method MAX" for
> > > > > > the
> > > > > > layer
> > > > > > 1
> > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > + Override the interpolation method of layer 1 with "-
method
> > > > > > <METHOD>"
> > > > > >
> > > > > > 1. The Gaussian smoothing at point2grid is activated and
> > > > > > applied
> > > > > > into the probability variable only if -method GAUSSIAN or
> > > > > > -method
> > > > > > MAXGAUSS is given as the command line argument. It's not
> > > > > > applied
> > > > > > into the variable at "-field" argument. If you want, You
can
> > > > > > do
> > > > > > this by using regrid_data_plane (passing the output of
> > > > > > point2grid
> > > > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > > > >
> > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
-method
> > > > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default
values
> > > > > > are
> > > > > > applied if not given.
> > > > > >
> > > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN or
> > > > > > -method
> > > > > > MAXGAUSS argument.
> > > > > >
> > > > > > 4 Without -prob_cat_thresh argument, it will be the status
of
> > > > > > data
> > > > > > existence. With -prob_cat_thresh argument, it will be
meeting
> > > > > > the
> > > > > > condition or not. So the values of the probability are 0
or
> > > > > > 1.
> > > > > >
> > > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > > > requested
> > > > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > John,
> > > > > > >
> > > > > > > I'm reassigning this ticket to Howard Soh who worked on
the
> > > > > > > development of the point2grid tool. He should be able to
> > > > > > > better
> > > > > > > answer these questions.
> > > > > >
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV
via
> > > > > > > RT
> > > > > > > <
> > > > > > > met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > > > d=95747 >
> > > > > > > >
> > > > > > > > John -
> > > > > > > >
> > > > > > > > I ran it to see the usage and there are details in the
> > > > > > > > usage
> > > > > > > > dialog and also in the XML file you pointed me to
which
> > > > > > > > helped
> > > > > > > > to answer my own question (I think).
> > > > > > > > There is an additional interpolation method called
> > > > > > > > GAUSSIAN
> > > > > > > > which applies only when using the probability variable
(-
> > > > > > > > prob_cat_thresh 280.0).
> > > > > > > > So, it
> > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > -gaussian_dx
> > > > > > > > and -gaussian_radius are only used when the method is
> > > > > > > > GAUSSIAN. IS this correct? This means that there is no
> > > > > > > > smoothing except when using probability variables. Is
> > > > > > > > this
> > > > > > > > correct?
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > #95747]
> > > > > > > > MET
> > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > > authenticity of all links contained within the message
> > > > > > > > prior
> > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > John,
> > > > > > > >
> > > > > > > > While the "-field" command line argument indicates
that
> > > > > > > > it
> > > > > > > > can
> > > > > > > > be used one or more times, in this first version of
the
> > > > > > > > tool,
> > > > > > > > it can only be used once.
> > > > > > > >
> > > > > > > > If you specify "-field" more that once you'll get this
> > > > > > > > error
> > > > > > > > message:
> > > > > > > > ERROR  :
> > > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > > "2"
> > > > > > > > variables
> > > > > > > > were requested.
> > > > > > > > ERROR  :
> > > > > > > >
> > > > > > > > We do plan to support the processing of multiple
fields
> > > > > > > > but
> > > > > > > > it
> > > > > > > > has not yet been implemented.
> > > > > > > >
> > > > > > > > As for the science behind the interpolation methods,
it's
> > > > > > > > very
> > > > > > > > simple. Keep track of the point observations falling
> > > > > > > > inside
> > > > > > > > each of the target grid boxes. Write to the output the
> > > > > > > > min,
> > > > > > > > max, unweighted mean, or median of those values based
on
> > > > > > > > the
> > > > > > > > method chosen. If you only have 1 point obs falling
> > > > > > > > inside
> > > > > > > > each grid box, then all those methods will result in
the
> > > > > > > > same
> > > > > > > > output value.
> > > > > > > > Whereas if you have many, as in the case of GOES input
> > > > > > > > files,
> > > > > > > > then the choice of the derivation method would matter.
> > > > > > > >
> > > > > > > > Hope that helps clarify.
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > >
> > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV
via
> > > > > > > > RT
> > > > > > > > <
> > > > > > > > met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > Caution-url:
> > > > > > > > > Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > Caution-Caution-
> > > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > ?id=
> > > > > > > > 95747 >
> > > > > > > > >
> > > > > > > > > Another question I have is how is -method used in
this
> > > > > > > > > tool?
> > > > > > > > > In the
> > > > > > > > User's
> > > > > > > > > Guide, -method provides options for regridding from
one
> > > > > > > > > grid
> > > > > > > > > to another such as is used in regrid_data_plane. In
the
> > > > > > > > > case
> > > > > > > > > of Point2Grid, the process involved is not
regridding
> > > > > > > > > in
> > > > > > > > > the
> > > > > > > > > way that I understand it.
> > > > > > > > > If I
> > > > > > > > > understand it correctly, it takes point observations
> > > > > > > > > and
> > > > > > > > > distributes them on to a user provided grid using
some
> > > > > > > > > interpolation method. The next
> > > > > > > > step I
> > > > > > > > > think is the smoothing of the field to even out
> > > > > > > > > irregularities which are caused by errors, noise,
etc
> > > > > > > > > resulting in the fitting of the observed values to
some
> > > > > > > > > function which appear to be a Gaussian function. I
> > > > > > > > > guess
> > > > > > > > > I
> > > > > > > > > also need to understand what happens when the tool
is
> > > > > > > > > used
> > > > > > > > > without any of the optional arguments.
> > > > > > > > >
> > > > > > > > > Thanks.
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > #95747]
> > > > > > > > > MET
> > > > > > > > > V9.0
> > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > John -
> > > > > > > > >
> > > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > > Appreciate
> > > > > > > > > your dedication so late into the evening! Can I add
new
> > > > > > > > > variables in the
> > > > > > > > config
> > > > > > > > > file? Any info on the science behind the
interpolation
> > > > > > > > > methods would be appreciated, but that can wait
until
> > > > > > > > > Monday.
> > > > > > > > >
> > > > > > > > > Take care and stay safe.
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > > > MET
> > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify
> > > > > > > > the
> > > > > > > > > identity of the sender, and confirm the authenticity
of
> > > > > > > > > all
> > > > > > > > > links
> > > > > > > > contained
> > > > > > > > > within the message prior to copying and pasting the
> > > > > > > > > address
> > > > > > > > > to a Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > John,
> > > > > > > > >
> > > > > > > > > I see you have a question about interpolation
options
> > > > > > > > > in
> > > > > > > > > the
> > > > > > > > > point2grid tool. This XML file contains the calls to
> > > > > > > > > the
> > > > > > > > > point2grid tool that are
> > > > > > > > run
> > > > > > > > > for our unit tests:
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > > > un
> > > > > > > > it
> > > > > > > > _point2grid.xml
> > > > > > > > >
> > > > > > > > > While most of them DO NOT use the "-method" command
> > > > > > > > > line
> > > > > > > > > option, a few at the end do use "-method MAX" to
select
> > > > > > > > > the
> > > > > > > > > maximum value falling within each grid box.
> > > > > > > > >
> > > > > > > > > Inspecting the code in this file:
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > Caution-
> > > > > > > > https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > > > oo
> > > > > > > > ls
> > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > I see that the supported options for -method are
MIN,
> > > > > > > > > MAX,
> > > > > > > > > MEDIAN, and UW_MEAN.
> > > > > > > > >
> > > > > > > > > Seems like we should add that info to the usage
> > > > > > > > > statement!
> > > > > > > > > I
> > > > > > > > > wrote up
> > > > > > > > this
> > > > > > > > > GitHub issue to make sure we remember to do so
before
> > > > > > > > > the
> > > > > > > > > 9.1
> > > > > > > > > release:
> > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > https://github.com/NCAR/MET/issu
> > > > > > > > > es/1396
> > > > > > > > >
> > > > > > > > > Hope that helps clarify.
> > > > > > > > >
> > > > > > > > > Thanks,
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA CIV
> > > > > > > > > via
> > > > > > > > > RT
> > > > > > > > > < met_help at ucar.edu> wrote:
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was acted
> > > > > > > > > > upon.
> > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > >        Queue: met_help
> > > > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > >        Owner: Nobody
> > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > >       Status: new
> > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-url:
> > > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > ?id=95747
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > >
> > > > > > > > > > The optional argument for smoothing or
interpolation
> > > > > > > > > > shown
> > > > > > > > > > in the
> > > > > > > > User's
> > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > If
> > > > > > > > > > so,
> > > > > > > > > > can you
> > > > > > > > > provide
> > > > > > > > > > the names and how to specify them as an optional
> > > > > > > > > > argument?
> > > > > > > > > >
> > > > > > > > > > Thanks.
> > > > > > > > > >
> > > > > > > > > > Stay safe.
> > > > > > > > > >
> > > > > > > > > > T/
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > U.S. Army Research Laboratory White Sands Missile
> > > > > > > > > > Range,
> > > > > > > > > > NM
> > > > > > > > > > 88002
> > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > Teleworking: Contact Admin Specialist at (410)
278-
> > > > > > > > > > 6974
> > > > > > > > > > FAX
> > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > >
> > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jul 20 10:56:25 2020

Hi John,

This is John Halley Gotway. I'm covering the met-help desk for the
next few
days and wanted to check in on this issue. Are you still having
trouble
understanding the output of point2grid, or does it make sense?

Thanks,
John

On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT <met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> The plot_data_plane can not show the point observation data. It
shows the
> plot of the grid data.
> In this case, the coverage of the gridded data is 0.23% (97 out of
41616).
> It's the plot of the grid cells (1km by 1km) even though it looks
like the
> point data. It caused by few converage (0.23%).
>
> The meaning of following log mesages (from point2grid):
>
> DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
> DEBUG 2: process_point_file() ->  var_count=180, grid: 97 out of
41616
>
> - The input data has 180 TMP data
> - One TMP data does not belong to the target grid
> - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616 *
100)
>   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
>
> The plot has valid data for 97 cells out of 41616 (which is 204 by
204)
> grid cells)
>
>
> Two issues with regrid_data_plane (this specific case because of
0.23%
> coverage).
>
> 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008 (<=
> 1 / 121).
>
> The default setting of "-vld_thresh" is 0.5.
> The max raduis (MAX_R) for the Gaussuan smoothing is 5. The matrix
for
> Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a valid
regrdded
> value for the taeget cell, it requires 61 (from 121 / 2) neighbers
with the
> valid TMP data. Only 97 cells have the valid data and they are
spreaded.
> None of them satisfies the number of the valid neighber cells.
That's why
> the output of regrid_data_plae was all missing values.
>
> 2. Bug on computing the Gaussian value.
>
> I will fix this.
>
> I attached the updated NetCDF file (output of regrid_data_plane) and
the
> plot.
>
> Cheers,
> Howard
>
> On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > Thanks for working to improve the documentation. So you can only
use
> > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
unless
> > you are working with a probability variable in which case you can
use
> > MAXGAUSS. The default of these is UW_MEAN. The output I generated
in
> > my first test used that method since I didn't specify the -method.
In
> > this case, the plot I generated using plot_data_plane shows a map
with
> > data points at random locations which correspond to the locations
of
> > the original point obs from the output of ascii2nc. My expectation
was
> > that I would see the data points with TMP values in a regularly
spaced
> > grid with 1km spacing. In theory, this NetCDF file could be used
in
> > Grid-Stat to verify a TMP fcst on the same 1km grid. Why doesn't
the
> > plot show a gridded TMP field?
> >
> > R/
> > John
> >
> >
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Wednesday, July 15, 2020 12:17 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> > the identity of the sender, and confirm the authenticity of all
links
> > contained within the message prior to copying and pasting the
address
> > to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Yes, you are right. It's not the output you are expecting. We will
> > improve the documentation on this (Caution-
> > https://github.com/NCAR/MET/issues/1413).
> >
> > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > differently: MAX interpolation for the raw data (TMP) and GAUSSIAN
> > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > MAXGAUSS is the same as the MAX interpolation for the TMP variable
for
> > the point2grid. So the difference is between MAX VS. UW_MEAN
> > interpolation methods, and GAUSIAN smoothing was not involved.
> >
> > The right way for the Gaussian smoothing to the raw data will be:
> > 1) regridding by using point2grid with one of interpolation
methods
> > (MAX/MIN/UW_MEAN/MEDIAN)
> > 2) regridding by using regrid_data_plane with MAXGAUSS method and
the
> > same setting as the input and the target grid.
> >
> > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > on this.
> >
> > Cheers,
> > Howard
> >
> >
> > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > Howard -
> > >
> > > Today I successfully ran point2grid but used -method MAXGAUSS
and it
> > > ran without error and produced output. I'm uploading the file
list,
> > > output, commands used, log file, and plots to the DoD SAFE so
you can
> > > take a look. I am also uploading the results of the previous run
> > > where
> > > you suggested that I use "11" vice "TMP" which fixed the problem
and
> > > ran successfully.
> > >
> > > The plots are not what I was expecting somehow. The plotted data
look
> > > like just points where the original obs were located rather than
a
> > > regularly spaced grid of TMP values. What are your thoughts on
this?
> > >
> > > Thanks
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Monday, July 13, 2020 3:33 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify
> > > the identity of the sender, and confirm the authenticity of all
links
> > > contained within the message prior to copying and pasting the
address
> > > to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Something is not right. I will take a look
> > >
> > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
246.351
> > > to 284.817.
> > > DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is
> > > -9999
> > > to -9999.
> > >
> > > The input is OK (between 246.351 and 284.817), but the regrided
> > > output
> > > is all missing value (-9999).
> > >
> > > Cheers,
> > > Howard
> > >
> > > The input file has
> > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > Howard -
> > > >
> > > > The next run with MAXGAUSS was successful, although I have to
plot
> > > > it
> > > > to see
> > > > what it looks like. I've attached the log file. Does it
indicate
> > > > any
> > > > issues
> > > > with how it ran?
> > > >
> > > > R/
> > > > John
> > > >
> > > > -----Original Message-----
> > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > <john.w.raby2.civ at mail.mil>
> > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify
> > > > the
> > > > identity of the sender, and confirm the authenticity of all
links
> > > > contained
> > > > within the message prior to copying and pasting the address to
a
> > > > Web
> > > > browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
MAXGAUSS
> > > > and
> > > > GAUSSIAN
> > > > method is only used for the interpolation. Please set "-method
> > > > MAXGAUSE". The
> > > > from_grid and to_grid are the same, there is no difference
with
> > > > additional MAX
> > > > interpolation.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > >
> > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > Howard -
> > > > >
> > > > > I tried running MET V9.0 regrid_data_plane following your
example
> > > > > below, but I'm getting and error:
> > > > > ERROR  :
> > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > > > > regridding, only interpolating.
> > > > >
> > > > > The following is the command I used to run it:
> > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > V2016.nc
> > > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > V2016.nc
> > > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TMP_1200UTC_11NOV2016_gaussian.nc
> > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> > > > >
> > > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > ss_test
> > > > > -v 9
> > > > >
> > > > > Any suggestions on how to perform Gaussian smoothing on the
> > > > > output
> > > > > TMP
> > > > > field from point2grid?
> > > > >
> > > > > Thanks.
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > > -----Original Message-----
> > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > <john.w.raby2.civ at mail.mil>
> > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify
> > > > > the identity of the sender, and confirm the authenticity of
all
> > > > > links
> > > > > contained within the message prior to copying and pasting
the
> > > > > address
> > > > > to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Sorry for the late response.
> > > > >
> > > > > The input and to_grid must be the same for Gaussian
smoothing and
> > > > > regrid_data_plane shows the usage message if the to_grid is
> > > > > missing.
> > > > > There are
> > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> > > > >
> > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > pb2nc_gaussian_TMP.nc
> > > > > -field 'name="TMP"; level="(*,*)";' \ -method GAUSSIAN -v 1
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > >
> > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > Just
> > > > > > curious about your regrid_data_plane command. You provided
the
> > > > > > input
> > > > > > file from point2grid (pb2nc_TMP.nc ) as the input_filename
to
> > > > > > regrid_data_plane and then you provide the name of the
file to
> > > > > > be
> > > > > > output from regrid_data_plane (pb2nc_TMP.nc
> > > > > > pb2nc_gaussian_TMP.nc),
> > > > > > but you didn't provide the to_grid. Is this because it
knows
> > > > > > that
> > > > > > the grids of the input and output files are the same and
thus
> > > > > > ignores the missing to_grid argument?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > > of
> > > > > > all links contained within the message prior to copying
and
> > > > > > pasting
> > > > > > the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > too.
> > > > > > In
> > > > > > this case, we are interested in Gaussian smoothing without
> > > > > > regriding.
> > > > > > The input and the target grid should be the same, which is
the
> > > > > > output of point2grid.
> > > > > >
> > > > > > Here is an example (assuming the point data file is
generated
> > > > > > by
> > > > > > pb2nc).
> > > > > >
> > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > >       G212 pb2nc_TMP.nc \
> > > > > >       -field 'name="TMP"; level="*";' \
> > > > > >       -v 1
> > > > > >
> > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > pb2nc_gaussian_TMP.nc
> > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > >       -method GAUSSIAN -v 1
> > > > > >
> > > > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
(120.0)
> > > > > > if
> > > > > > necessary
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > Howard -
> > > > > > >
> > > > > > > Thanks for your quick response with details in answer to
my
> > > > > > > question.
> > > > > > > Right now my interest is with non-probability fields, so
your
> > > > > > > point in 1. about the possibility of using Gaussian
smoothing
> > > > > > > for
> > > > > > > -field argument is interesting.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please
> > > > > > > verify the identity of the sender, and confirm the
> > > > > > > authenticity
> > > > > > > of
> > > > > > > all links contained within the message prior to copying
and
> > > > > > > pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the
raw
> > > > > > > data
> > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > probability
> > > > > > > variable. (the existence of valid values if the
probability
> > > > > > > threshold is not given)
> > > > > > >
> > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > (which
> > > > > > > is
> > > > > > > a
> > > > > > > default) for the layer 1 and "-method GAUSSIAN" for the
layer
> > > > > > > 2.
> > > > > > > + "-method MAXGAUSS" is processed as the "-method MAX"
for
> > > > > > > the
> > > > > > > layer
> > > > > > > 1
> > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > + Override the interpolation method of layer 1 with "-
method
> > > > > > > <METHOD>"
> > > > > > >
> > > > > > > 1. The Gaussian smoothing at point2grid is activated and
> > > > > > > applied
> > > > > > > into the probability variable only if -method GAUSSIAN
or
> > > > > > > -method
> > > > > > > MAXGAUSS is given as the command line argument. It's not
> > > > > > > applied
> > > > > > > into the variable at "-field" argument. If you want, You
can
> > > > > > > do
> > > > > > > this by using regrid_data_plane (passing the output of
> > > > > > > point2grid
> > > > > > > as the input of regrid_data_plane with GAUSSIAN method).
> > > > > > >
> > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
-method
> > > > > > > GAUSSIAN or -method MAXGAUSS is NOT given. The default
values
> > > > > > > are
> > > > > > > applied if not given.
> > > > > > >
> > > > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN
or
> > > > > > > -method
> > > > > > > MAXGAUSS argument.
> > > > > > >
> > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status of
> > > > > > > data
> > > > > > > existence. With -prob_cat_thresh argument, it will be
meeting
> > > > > > > the
> > > > > > > condition or not. So the values of the probability are 0
or
> > > > > > > 1.
> > > > > > >
> > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > > > > requested
> > > > > > > (between 0. and 1.0 with GAUSSIAN smoothing).
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > John,
> > > > > > > >
> > > > > > > > I'm reassigning this ticket to Howard Soh who worked
on the
> > > > > > > > development of the point2grid tool. He should be able
to
> > > > > > > > better
> > > > > > > > answer these questions.
> > > > > > >
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV
via
> > > > > > > > RT
> > > > > > > > <
> > > > > > > > met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
url:
> > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?i
> > > > > > > > > d=95747 >
> > > > > > > > >
> > > > > > > > > John -
> > > > > > > > >
> > > > > > > > > I ran it to see the usage and there are details in
the
> > > > > > > > > usage
> > > > > > > > > dialog and also in the XML file you pointed me to
which
> > > > > > > > > helped
> > > > > > > > > to answer my own question (I think).
> > > > > > > > > There is an additional interpolation method called
> > > > > > > > > GAUSSIAN
> > > > > > > > > which applies only when using the probability
variable (-
> > > > > > > > > prob_cat_thresh 280.0).
> > > > > > > > > So, it
> > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > -gaussian_dx
> > > > > > > > > and -gaussian_radius are only used when the method
is
> > > > > > > > > GAUSSIAN. IS this correct? This means that there is
no
> > > > > > > > > smoothing except when using probability variables.
Is
> > > > > > > > > this
> > > > > > > > > correct?
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > #95747]
> > > > > > > > > MET
> > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify the identity of the sender, and
confirm the
> > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > prior
> > > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > John,
> > > > > > > > >
> > > > > > > > > While the "-field" command line argument indicates
that
> > > > > > > > > it
> > > > > > > > > can
> > > > > > > > > be used one or more times, in this first version of
the
> > > > > > > > > tool,
> > > > > > > > > it can only be used once.
> > > > > > > > >
> > > > > > > > > If you specify "-field" more that once you'll get
this
> > > > > > > > > error
> > > > > > > > > message:
> > > > > > > > > ERROR  :
> > > > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > > > "2"
> > > > > > > > > variables
> > > > > > > > > were requested.
> > > > > > > > > ERROR  :
> > > > > > > > >
> > > > > > > > > We do plan to support the processing of multiple
fields
> > > > > > > > > but
> > > > > > > > > it
> > > > > > > > > has not yet been implemented.
> > > > > > > > >
> > > > > > > > > As for the science behind the interpolation methods,
it's
> > > > > > > > > very
> > > > > > > > > simple. Keep track of the point observations falling
> > > > > > > > > inside
> > > > > > > > > each of the target grid boxes. Write to the output
the
> > > > > > > > > min,
> > > > > > > > > max, unweighted mean, or median of those values
based on
> > > > > > > > > the
> > > > > > > > > method chosen. If you only have 1 point obs falling
> > > > > > > > > inside
> > > > > > > > > each grid box, then all those methods will result in
the
> > > > > > > > > same
> > > > > > > > > output value.
> > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > files,
> > > > > > > > > then the choice of the derivation method would
matter.
> > > > > > > > >
> > > > > > > > > Hope that helps clarify.
> > > > > > > > >
> > > > > > > > > Thanks,
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV
via
> > > > > > > > > RT
> > > > > > > > > <
> > > > > > > > > met_help at ucar.edu> wrote:
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-url:
> > > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > ?id=
> > > > > > > > > 95747 >
> > > > > > > > > >
> > > > > > > > > > Another question I have is how is -method used in
this
> > > > > > > > > > tool?
> > > > > > > > > > In the
> > > > > > > > > User's
> > > > > > > > > > Guide, -method provides options for regridding
from one
> > > > > > > > > > grid
> > > > > > > > > > to another such as is used in regrid_data_plane.
In the
> > > > > > > > > > case
> > > > > > > > > > of Point2Grid, the process involved is not
regridding
> > > > > > > > > > in
> > > > > > > > > > the
> > > > > > > > > > way that I understand it.
> > > > > > > > > > If I
> > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > and
> > > > > > > > > > distributes them on to a user provided grid using
some
> > > > > > > > > > interpolation method. The next
> > > > > > > > > step I
> > > > > > > > > > think is the smoothing of the field to even out
> > > > > > > > > > irregularities which are caused by errors, noise,
etc
> > > > > > > > > > resulting in the fitting of the observed values to
some
> > > > > > > > > > function which appear to be a Gaussian function. I
> > > > > > > > > > guess
> > > > > > > > > > I
> > > > > > > > > > also need to understand what happens when the tool
is
> > > > > > > > > > used
> > > > > > > > > > without any of the optional arguments.
> > > > > > > > > >
> > > > > > > > > > Thanks.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > #95747]
> > > > > > > > > > MET
> > > > > > > > > > V9.0
> > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > John -
> > > > > > > > > >
> > > > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > > > Appreciate
> > > > > > > > > > your dedication so late into the evening! Can I
add new
> > > > > > > > > > variables in the
> > > > > > > > > config
> > > > > > > > > > file? Any info on the science behind the
interpolation
> > > > > > > > > > methods would be appreciated, but that can wait
until
> > > > > > > > > > Monday.
> > > > > > > > > >
> > > > > > > > > > Take care and stay safe.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
#95747]
> > > > > > > > > > MET
> > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please verify
> > > > > > > > > the
> > > > > > > > > > identity of the sender, and confirm the
authenticity of
> > > > > > > > > > all
> > > > > > > > > > links
> > > > > > > > > contained
> > > > > > > > > > within the message prior to copying and pasting
the
> > > > > > > > > > address
> > > > > > > > > > to a Web browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > I see you have a question about interpolation
options
> > > > > > > > > > in
> > > > > > > > > > the
> > > > > > > > > > point2grid tool. This XML file contains the calls
to
> > > > > > > > > > the
> > > > > > > > > > point2grid tool that are
> > > > > > > > > run
> > > > > > > > > > for our unit tests:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > >
https://github.com/NCAR/MET/blob/master_v9.0/test/xml/
> > > > > > > > > un
> > > > > > > > > it
> > > > > > > > > _point2grid.xml
> > > > > > > > > >
> > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > line
> > > > > > > > > > option, a few at the end do use "-method MAX" to
select
> > > > > > > > > > the
> > > > > > > > > > maximum value falling within each grid box.
> > > > > > > > > >
> > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > >
https://github.com/NCAR/MET/blob/master_v9.0/met/src/t
> > > > > > > > > oo
> > > > > > > > > ls
> > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > I see that the supported options for -method are
MIN,
> > > > > > > > > > MAX,
> > > > > > > > > > MEDIAN, and UW_MEAN.
> > > > > > > > > >
> > > > > > > > > > Seems like we should add that info to the usage
> > > > > > > > > > statement!
> > > > > > > > > > I
> > > > > > > > > > wrote up
> > > > > > > > > this
> > > > > > > > > > GitHub issue to make sure we remember to do so
before
> > > > > > > > > > the
> > > > > > > > > > 9.1
> > > > > > > > > > release:
> > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > https://github.com/NCAR/MET/issu
> > > > > > > > > > es/1396
> > > > > > > > > >
> > > > > > > > > > Hope that helps clarify.
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA
CIV
> > > > > > > > > > via
> > > > > > > > > > RT
> > > > > > > > > > < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > upon.
> > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > >        Queue: met_help
> > > > > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > >       Status: new
> > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-url:
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > ?id=95747
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > >
> > > > > > > > > > > The optional argument for smoothing or
interpolation
> > > > > > > > > > > shown
> > > > > > > > > > > in the
> > > > > > > > > User's
> > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > If
> > > > > > > > > > > so,
> > > > > > > > > > > can you
> > > > > > > > > > provide
> > > > > > > > > > > the names and how to specify them as an optional
> > > > > > > > > > > argument?
> > > > > > > > > > >
> > > > > > > > > > > Thanks.
> > > > > > > > > > >
> > > > > > > > > > > Stay safe.
> > > > > > > > > > >
> > > > > > > > > > > T/
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > Range,
> > > > > > > > > > > NM
> > > > > > > > > > > 88002
> > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > Teleworking: Contact Admin Specialist at (410)
278-
> > > > > > > > > > > 6974
> > > > > > > > > > > FAX
> > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > >
> > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
>
>
>
>

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 20 13:21:26 2020

CLASSIFICATION: UNCLASSIFIED

John -

Welcome back. Yes, I'm still trying to understand the output gridded
file
which places point obs on a grid.  I'm comparing what I'm seeing from
my
testing of point2grid so far with an experiment I did last JAN where I
used
IDV to produce a NetCDF file containing a set of gridded obs on a 1km
grid.
Ref ticket # 93980 for your assist in making the NetCDF file from IDV
MET_compatible. The attached text file contains the point obs which
were used
by IDV to produce the gridded obs. The result was a MET-compatible
file which
I was able to successfully ingest into Grid-Stat. See attached NetCDF
file. If
you look at the plot of the IDV generated gridded obs in the attached
PDF, it
looks different from that which is generated by point2grid. I'm trying
to
reconcile the two to gain further understanding.  See the attached PDF
of the
point2grid output which I produced. The obs used for the point2grid
testing
are not the same ones as those used for the IDV testing. The NETCDF
output
from point2grid is also attached. See list of files below.

That said, I understand that Howard is working on regrid_data_plane to
enable
the use of Gaussian smoothing (MAXGAUSS) on the output of point2grid.
I wonder
if the Gaussian smoothing will achieve an end product which look more
similar
to what I was expecting? Maybe the gridded obs I produced with IDV
isn't ever
going to produce the same thing which point2grid was conceived to
produce due
to the use of a different concept in performing an objective analysis
using
interpolation.

List of attachments:
Case1_TMP_OBS_18_UTC - point obs text file
1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded obs
1km_TMP_1200UTC_11NOV2016.nc - point2grid output
1km_TMP_1200UTC_11NOV2016 - plot of point2grid output

R/
John


-----Original Message-----
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Monday, July 20, 2020 10:56 AM
To: Raby, John W CIV USARMY CCDC ARL (USA) <john.w.raby2.civ at mail.mil>
Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid
(UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the
identity of the sender, and confirm the authenticity of all links
contained
within the message prior to copying and pasting the address to a Web
browser.




----

Hi John,

This is John Halley Gotway. I'm covering the met-help desk for the
next few
days and wanted to check in on this issue. Are you still having
trouble
understanding the output of point2grid, or does it make sense?

Thanks,
John

On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT <met_help at ucar.edu>
wrote:

>
> <Caution-url:
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> The plot_data_plane can not show the point observation data. It
shows
> the plot of the grid data.
> In this case, the coverage of the gridded data is 0.23% (97 out of
41616).
> It's the plot of the grid cells (1km by 1km) even though it looks
like
> the point data. It caused by few converage (0.23%).
>
> The meaning of following log mesages (from point2grid):
>
> DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG
> 2: process_point_file() ->  var_count=180, grid: 97 out of 41616
>
> - The input data has 180 TMP data
> - One TMP data does not belong to the target grid
> - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616 *
100)
>   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
>
> The plot has valid data for 97 cells out of 41616 (which is 204 by
> 204) grid cells)
>
>
> Two issues with regrid_data_plane (this specific case because of
0.23%
> coverage).
>
> 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008
> (<=
> 1 / 121).
>
> The default setting of "-vld_thresh" is 0.5.
> The max raduis (MAX_R) for the Gaussuan smoothing is 5. The matrix
for
> Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a valid
> regrdded value for the taeget cell, it requires 61 (from 121 / 2)
> neighbers with the valid TMP data. Only 97 cells have the valid data
and
> they are spreaded.
> None of them satisfies the number of the valid neighber cells.
That's
> why the output of regrid_data_plae was all missing values.
>
> 2. Bug on computing the Gaussian value.
>
> I will fix this.
>
> I attached the updated NetCDF file (output of regrid_data_plane) and
> the plot.
>
> Cheers,
> Howard
>
> On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > Thanks for working to improve the documentation. So you can only
use
> > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
> > unless you are working with a probability variable in which case
you
> > can use MAXGAUSS. The default of these is UW_MEAN. The output I
> > generated in my first test used that method since I didn't specify
> > the -method. In this case, the plot I generated using
> > plot_data_plane shows a map with data points at random locations
> > which correspond to the locations of the original point obs from
the
> > output of ascii2nc. My expectation was that I would see the data
> > points with TMP values in a regularly spaced grid with 1km
spacing.
> > In theory, this NetCDF file could be used in Grid-Stat to verify a
> > TMP fcst on the same 1km grid. Why doesn't the plot show a gridded
TMP
> > field?
> >
> > R/
> > John
> >
> >
> > ________________________________________
> > From: Howard Soh via RT [met_help at ucar.edu]
> > Sent: Wednesday, July 15, 2020 12:17 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
> > verify the identity of the sender, and confirm the authenticity of
> > all links contained within the message prior to copying and
pasting
> > the address to a Web browser.
> >
> >
> >
> >
> > ----
> >
> > Yes, you are right. It's not the output you are expecting. We will
> > improve the documentation on this (Caution-
> > Caution-https://github.com/NCAR/MET/issues/1413).
> >
> > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > differently: MAX interpolation for the raw data (TMP) and GAUSSIAN
> > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > MAXGAUSS is the same as the MAX interpolation for the TMP variable
> > for the point2grid. So the difference is between MAX VS. UW_MEAN
> > interpolation methods, and GAUSIAN smoothing was not involved.
> >
> > The right way for the Gaussian smoothing to the raw data will be:
> > 1) regridding by using point2grid with one of interpolation
methods
> > (MAX/MIN/UW_MEAN/MEDIAN)
> > 2) regridding by using regrid_data_plane with MAXGAUSS method and
> > the same setting as the input and the target grid.
> >
> > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > on this.
> >
> > Cheers,
> > Howard
> >
> >
> > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > Howard -
> > >
> > > Today I successfully ran point2grid but used -method MAXGAUSS
and
> > > it ran without error and produced output. I'm uploading the file
> > > list, output, commands used, log file, and plots to the DoD SAFE
> > > so you can take a look. I am also uploading the results of the
> > > previous run where you suggested that I use "11" vice "TMP"
which
> > > fixed the problem and ran successfully.
> > >
> > > The plots are not what I was expecting somehow. The plotted data
> > > look like just points where the original obs were located rather
> > > than a regularly spaced grid of TMP values. What are your
thoughts on
> > > this?
> > >
> > > Thanks
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Monday, July 13, 2020 3:33 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
> > > pasting the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Something is not right. I will take a look
> > >
> > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
> > > 246.351 to 284.817.
> > > DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";) is
> > > -9999
> > > to -9999.
> > >
> > > The input is OK (between 246.351 and 284.817), but the regrided
> > > output is all missing value (-9999).
> > >
> > > Cheers,
> > > Howard
> > >
> > > The input file has
> > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > Howard -
> > > >
> > > > The next run with MAXGAUSS was successful, although I have to
> > > > plot it to see what it looks like. I've attached the log file.
> > > > Does it indicate any issues with how it ran?
> > > >
> > > > R/
> > > > John
> > > >
> > > > -----Original Message-----
> > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > <john.w.raby2.civ at mail.mil>
> > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity
> > > > of all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > MAXGAUSS and GAUSSIAN method is only used for the
interpolation.
> > > > Please set "-method MAXGAUSE". The from_grid and to_grid are
the
> > > > same, there is no difference with additional MAX
interpolation.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > >
> > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > Howard -
> > > > >
> > > > > I tried running MET V9.0 regrid_data_plane following your
> > > > > example below, but I'm getting and error:
> > > > > ERROR  :
> > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid for
> > > > > regridding, only interpolating.
> > > > >
> > > > > The following is the command I used to run it:
> > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > V2016.nc
> > > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > V2016.nc
> > > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> P_1200UTC_11NOV2016_gaussian.nc
> > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN -log
> > > > >
> > > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > ss_test
> > > > > -v 9
> > > > >
> > > > > Any suggestions on how to perform Gaussian smoothing on the
> > > > > output TMP field from point2grid?
> > > > >
> > > > > Thanks.
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > > -----Original Message-----
> > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > <john.w.raby2.civ at mail.mil>
> > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
> > > > > Please verify the identity of the sender, and confirm the
> > > > > authenticity of all links contained within the message prior
> > > > > to copying and pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Sorry for the late response.
> > > > >
> > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > and regrid_data_plane shows the usage message if the to_grid
> > > > > is missing.
> > > > > There are
> > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the to_grid.
> > > > >
> > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP"; level="(*,*)";' \
> > > > > -method GAUSSIAN -v 1
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > >
> > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > Just
> > > > > > curious about your regrid_data_plane command. You provided
> > > > > > the input file from point2grid (pb2nc_TMP.nc ) as the
> > > > > > input_filename to regrid_data_plane and then you provide
the
> > > > > > name of the file to be output from regrid_data_plane
> > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
provide
> > > > > > the to_grid. Is this because it knows that the grids of
the
> > > > > > input and output files are the same and thus ignores the
> > > > > > missing to_grid argument?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify the identity of the sender, and confirm the
> > > > > > authenticity of all links contained within the message
prior
> > > > > > to copying and pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > too.
> > > > > > In
> > > > > > this case, we are interested in Gaussian smoothing without
> > > > > > regriding.
> > > > > > The input and the target grid should be the same, which is
> > > > > > the output of point2grid.
> > > > > >
> > > > > > Here is an example (assuming the point data file is
> > > > > > generated by pb2nc).
> > > > > >
> > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > >       G212 pb2nc_TMP.nc \
> > > > > >       -field 'name="TMP"; level="*";' \
> > > > > >       -v 1
> > > > > >
> > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > pb2nc_gaussian_TMP.nc
> > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > >       -method GAUSSIAN -v 1
> > > > > >
> > > > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
> > > > > > (120.0) if necessary
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > Howard -
> > > > > > >
> > > > > > > Thanks for your quick response with details in answer to
> > > > > > > my question.
> > > > > > > Right now my interest is with non-probability fields, so
> > > > > > > your point in 1. about the possibility of using Gaussian
> > > > > > > smoothing for -field argument is interesting.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > > ________________________________________
> > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
> > > > > > > V9.0 Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please
> > > > > > > verify the identity of the sender, and confirm the
> > > > > > > authenticity of all links contained within the message
> > > > > > > prior to copying and pasting the address to a Web
browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for the
> > > > > > > raw data
> > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > probability variable. (the existence of valid values if
> > > > > > > the probability threshold is not given)
> > > > > > >
> > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > (which
> > > > > > > is
> > > > > > > a
> > > > > > > default) for the layer 1 and "-method GAUSSIAN" for the
> > > > > > > layer 2.
> > > > > > > + "-method MAXGAUSS" is processed as the "-method MAX"
for
> > > > > > > the
> > > > > > > layer
> > > > > > > 1
> > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > + Override the interpolation method of layer 1 with
> > > > > > > + "-method
> > > > > > > <METHOD>"
> > > > > > >
> > > > > > > 1. The Gaussian smoothing at point2grid is activated and
> > > > > > > applied into the probability variable only if -method
> > > > > > > GAUSSIAN or -method MAXGAUSS is given as the command
line
> > > > > > > argument. It's not applied into the variable at "-field"
> > > > > > > argument. If you want, You can do this by using
> > > > > > > regrid_data_plane (passing the output of point2grid as
the
> > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > >
> > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
> > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given. The
> > > > > > > default values are applied if not given.
> > > > > > >
> > > > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN
or
> > > > > > > -method MAXGAUSS argument.
> > > > > > >
> > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status
> > > > > > > of data existence. With -prob_cat_thresh argument, it
will
> > > > > > > be meeting the condition or not. So the values of the
> > > > > > > probability are 0 or 1.
> > > > > > >
> > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing is
> > > > > > > requested (between 0. and 1.0 with GAUSSIAN smoothing).
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > John,
> > > > > > > >
> > > > > > > > I'm reassigning this ticket to Howard Soh who worked
on
> > > > > > > > the development of the point2grid tool. He should be
> > > > > > > > able to better answer these questions.
> > > > > > >
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA CIV
> > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > lay.html?i
> > > > > > > > > d=95747 >
> > > > > > > > >
> > > > > > > > > John -
> > > > > > > > >
> > > > > > > > > I ran it to see the usage and there are details in
the
> > > > > > > > > usage dialog and also in the XML file you pointed me
> > > > > > > > > to which helped to answer my own question (I think).
> > > > > > > > > There is an additional interpolation method called
> > > > > > > > > GAUSSIAN which applies only when using the
probability
> > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > So, it
> > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > -gaussian_dx and -gaussian_radius are only used when
> > > > > > > > > the method is GAUSSIAN. IS this correct? This means
> > > > > > > > > that there is no smoothing except when using
> > > > > > > > > probability variables. Is this correct?
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > #95747] MET
> > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > the authenticity of all links contained within the
> > > > > > > > > message prior to copying and pasting the address to
a
> > > > > > > > > Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > John,
> > > > > > > > >
> > > > > > > > > While the "-field" command line argument indicates
> > > > > > > > > that it can be used one or more times, in this first
> > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > >
> > > > > > > > > If you specify "-field" more that once you'll get
this
> > > > > > > > > error
> > > > > > > > > message:
> > > > > > > > > ERROR  :
> > > > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > > > "2"
> > > > > > > > > variables
> > > > > > > > > were requested.
> > > > > > > > > ERROR  :
> > > > > > > > >
> > > > > > > > > We do plan to support the processing of multiple
> > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > >
> > > > > > > > > As for the science behind the interpolation methods,
> > > > > > > > > it's very simple. Keep track of the point
observations
> > > > > > > > > falling inside each of the target grid boxes. Write
to
> > > > > > > > > the output the min, max, unweighted mean, or median
of
> > > > > > > > > those values based on the method chosen. If you only
> > > > > > > > > have 1 point obs falling inside each grid box, then
> > > > > > > > > all those methods will result in the same output
> > > > > > > > > value.
> > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > files, then the choice of the derivation method
would
> > > > > > > > > matter.
> > > > > > > > >
> > > > > > > > > Hope that helps clarify.
> > > > > > > > >
> > > > > > > > > Thanks,
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA CIV
> > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-Caution-url:
> > > > > > > > > > Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-Caution-
> > > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > ?id=
> > > > > > > > > 95747 >
> > > > > > > > > >
> > > > > > > > > > Another question I have is how is -method used in
> > > > > > > > > > this tool?
> > > > > > > > > > In the
> > > > > > > > > User's
> > > > > > > > > > Guide, -method provides options for regridding
from
> > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > regrid_data_plane. In the case of Point2Grid, the
> > > > > > > > > > process involved is not regridding in the way that
I
> > > > > > > > > > understand it.
> > > > > > > > > > If I
> > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > and distributes them on to a user provided grid
> > > > > > > > > > using some interpolation method. The next
> > > > > > > > > step I
> > > > > > > > > > think is the smoothing of the field to even out
> > > > > > > > > > irregularities which are caused by errors, noise,
> > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > to some function which appear to be a Gaussian
> > > > > > > > > > function. I guess I also need to understand what
> > > > > > > > > > happens when the tool is used without any of the
> > > > > > > > > > optional arguments.
> > > > > > > > > >
> > > > > > > > > > Thanks.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > #95747] MET
> > > > > > > > > > V9.0
> > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > John -
> > > > > > > > > >
> > > > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > > > Appreciate
> > > > > > > > > > your dedication so late into the evening! Can I
add
> > > > > > > > > > new variables in the
> > > > > > > > > config
> > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > interpolation methods would be appreciated, but
that
> > > > > > > > > > can wait until Monday.
> > > > > > > > > >
> > > > > > > > > > Take care and stay safe.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > #95747] MET
> > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please verify
> > > > > > > > > the
> > > > > > > > > > identity of the sender, and confirm the
authenticity
> > > > > > > > > > of all links
> > > > > > > > > contained
> > > > > > > > > > within the message prior to copying and pasting
the
> > > > > > > > > > address to a Web browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > I see you have a question about interpolation
> > > > > > > > > > options in the point2grid tool. This XML file
> > > > > > > > > > contains the calls to the point2grid tool that are
> > > > > > > > > run
> > > > > > > > > > for our unit tests:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > est/xml/
> > > > > > > > > un
> > > > > > > > > it
> > > > > > > > > _point2grid.xml
> > > > > > > > > >
> > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > line option, a few at the end do use "-method MAX"
> > > > > > > > > > to select the maximum value falling within each
grid
> > > > > > > > > > box.
> > > > > > > > > >
> > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
> > > > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > et/src/t
> > > > > > > > > oo
> > > > > > > > > ls
> > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > I see that the supported options for -method are
> > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > >
> > > > > > > > > > Seems like we should add that info to the usage
> > > > > > > > > > statement!
> > > > > > > > > > I
> > > > > > > > > > wrote up
> > > > > > > > > this
> > > > > > > > > > GitHub issue to make sure we remember to do so
> > > > > > > > > > before the
> > > > > > > > > > 9.1
> > > > > > > > > > release:
> > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > > > > > > es/1396
> > > > > > > > > >
> > > > > > > > > > Hope that helps clarify.
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA
CIV
> > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > upon.
> > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > >        Queue: met_help
> > > > > > > > > > >      Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > >       Status: new
> > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-Caution-url:
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > splay.html
> > > > > > > > > > ?id=95747
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > >
> > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > interpolation shown in the
> > > > > > > > > User's
> > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > If
> > > > > > > > > > > so,
> > > > > > > > > > > can you
> > > > > > > > > > provide
> > > > > > > > > > > the names and how to specify them as an optional
> > > > > > > > > > > argument?
> > > > > > > > > > >
> > > > > > > > > > > Thanks.
> > > > > > > > > > >
> > > > > > > > > > > Stay safe.
> > > > > > > > > > >
> > > > > > > > > > > T/
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > Range, NM
> > > > > > > > > > > 88002
> > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > Teleworking: Contact Admin Specialist at (410)
> > > > > > > > > > > 278-
> > > > > > > > > > > 6974
> > > > > > > > > > > FAX
> > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > >
> > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
>
>
>
>

CLASSIFICATION: UNCLASSIFIED

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 20 13:21:26 2020

Id	Alt	Lat	Lon	TMP
AMOSD	165.20000	33.35713	-117.27570	287.59445
ANESD	555.70001	32.82697	-116.77332	285.92776
AR568	73.20000	32.94533	-117.24500	287.59445
ARHSD	450.79999	33.03983	-116.95800	287.03888
AS488	469.50000	33.02133	-116.94017	287.03888
AS678	228.00000	33.40050	-117.24533	285.92776
AT057	105.00000	32.79433	-117.07650	286.48334
AT094	1367.59998	33.06156	-116.57361	279.26111
AT107	621.00000	33.41511	-117.12317	285.92776
AT125	67.70000	33.24633	-117.25333	285.92776
AVOSD	197.80000	33.39250	-117.25500	285.37222
BAJC1	730.00000	32.66750	-116.69940	287.03888
BLCSD	578.20001	33.09856	-116.82758	288.70557
BMSSD	994.59998	32.76200	-116.48900	286.48334
BNASD	402.29999	32.93497	-116.87588	285.92776
BOCSD	1129.59998	32.94035	-116.63668	285.37222
BRJSD	372.20001	32.60110	-116.68600	290.37222
BVYSD	363.29999	32.86830	-116.84252	286.48334
C3287	109.70000	32.81480	-117.05310	287.03888
C7369	177.00000	32.76687	-117.01615	285.92776
C8890	314.00000	33.17717	-117.10450	288.14999
CCRSD	786.09998	33.28620	-116.89700	286.48334
C9939	655.00000	33.20167	-116.93733	288.14999
CGDSD	516.29999	33.28464	-117.01553	284.26111
CHHSD	1405.09998	33.38710	-116.68700	282.03888
CI147	176.80000	32.63000	-116.93800	287.31668
CI150	135.60001	32.88600	-117.14200	287.09445
CI184	114.90000	32.73000	-117.13900	287.48334
CIRSD	199.30000	33.28556	-117.14574	286.48334
CLMSD	440.10001	33.01561	-116.87001	285.37222
CMCSD	868.09998	32.64600	-116.49700	285.92776
CMNC1	1049.40002	32.72110	-116.46390	285.92776
CMNSD	979.00000	32.71464	-116.47076	285.37222
COCSD	238.00000	32.85138	-116.89515	285.92776
COYSD	1162.80005	33.25844	-116.58518	283.14999
CPOSD	741.59998	32.59898	-116.49292	286.48334
CRVSD	550.50000	32.78356	-116.72300	285.37222
CSTSD	452.89999	32.81154	-116.85324	287.03888
CTMSD	1220.40002	32.78290	-116.55000	285.37222
D0044	607.79999	32.84017	-116.76883	285.92776
D0395	14.00000	33.19917	-117.38700	286.48334
D2370	53.00000	33.06267	-117.28933	287.59445
D3101	105.90000	32.92109	-117.25306	287.03888
D4365	170.00000	32.45833	-116.82133	290.37222
D4414	131.00000	33.06236	-117.20023	286.48334
D4456	513.00000	33.07083	-116.85083	289.26111
D4679	221.60001	32.92900	-117.09200	286.48334
D5256	149.00000	32.77170	-117.03331	287.03888
D5683	186.80000	33.09417	-117.22050	287.03888
D5934	2.00000	32.52317	-117.12400	288.14999
D6351	184.00000	33.14790	-117.16311	287.59445
D6379	73.00000	33.17667	-117.27533	287.59445
D8659	320.00000	33.32500	-117.08517	285.92776
DEHSD	232.30000	32.78691	-116.83534	285.92776
DENSD	1034.19995	32.85883	-116.61182	285.37222
DLZSD	184.70000	33.45839	-117.29735	286.48334
DMHSD	89.90000	32.96400	-117.22300	287.03888
DZRSD	434.60001	32.61500	-116.75800	289.26111
ESCSD	220.39999	33.15917	-117.03133	289.26111
EWNSD	988.20001	33.22650	-116.61200	283.70557
FBKSD	290.50000	33.35471	-117.16720	287.03888
FTVSD	559.29999	33.25614	-116.98201	287.03888
GOSSD	468.20001	33.07279	-116.84494	289.81668
GTYSD	1186.30005	32.85520	-116.57400	284.26111
H0294	200.60500	32.75000	-116.93000	286.75000
HARSD	381.89999	32.85794	-116.82188	287.03888
HDMSD	114.30000	33.04390	-117.13500	286.48334
HIDSD	433.39999	33.23996	-117.04519	286.48334
HMDSD	414.79999	33.22651	-117.09415	285.37222
HP002	1146.09998	32.70150	-116.76340	283.59445
HVYSD	413.89999	33.02928	-116.93516	287.03888
JULSD	1280.50000	33.07638	-116.59076	279.81668
KCKSD	310.00000	33.31311	-117.08583	285.92776
KCRQ	100.00000	33.13000	-117.28000	287.14999
KCZZ	802.00000	32.62000	-116.47000	286.14999
KMYF	129.00000	32.82000	-117.13000	287.14999
KNFG	24.00000	33.30000	-117.35000	286.14999
KNRS	7.00000	32.57000	-117.12000	291.14999
KNXF	27.00000	33.29000	-117.46000	287.14999
KNZY	8.00000	32.70000	-117.22000	290.14999
LCMSD	1456.59998	32.99740	-116.59500	278.70557
LFRSD	1444.80005	32.93310	-116.52800	279.81668
LKWSD	500.50000	33.17799	-116.99470	287.03888
LLDSD	419.10001	32.80554	-116.78794	286.48334
MGDSD	1020.79999	33.16756	-116.75917	285.37222
MGRSD	485.89999	32.98626	-116.90810	286.48334
MGYSD	854.40002	33.22329	-116.69879	284.81668
MORSD	940.29999	32.68839	-116.51179	285.37222
MSDSD	113.10000	32.81400	-117.24100	287.03888
MVNSD	62.80000	32.78100	-117.13800	287.03888
NBCSD	942.09998	32.98820	-116.66800	284.26111
NDCSD	1029.30005	32.88320	-116.64700	284.81668
NHGSD	553.79999	33.32220	-116.96200	286.48334
NPTSD	771.79999	32.62980	-116.58500	287.03888
OLVSD	355.70001	33.08010	-117.13000	286.48334
PAUSD	125.30000	33.37111	-117.07894	285.37222
POTSD	719.29999	32.60439	-116.60744	287.03888
PSQC1	77.70000	33.09140	-117.01220	285.92776
PTMSD	393.79999	33.41921	-117.08186	286.48334
PVDSD	561.09998	32.81324	-116.73412	284.81668
RAMSD	461.50000	33.07300	-116.85802	289.26111
RBHSD	569.70001	33.41396	-117.12595	287.03888
RBVSD	333.50000	33.40363	-117.16473	287.59445
RCHSD	1258.19995	33.21030	-116.50836	280.92776
RIOSD	204.20000	32.84329	-116.88113	287.03888
RNBSD	322.50000	33.41892	-117.14413	287.03888
RPOSD	803.09998	32.64791	-116.63145	287.03888
RSFSD	75.90000	33.03333	-117.18956	287.03888
SDALP	518.20001	32.83600	-116.77700	285.37222
SDAPN	387.10001	32.83194	-116.81333	285.92776
SDEHM	469.39999	33.22600	-117.11500	285.37222
SDL13	106.70000	32.67500	-117.23900	287.59445
SDLAM	149.39999	32.77170	-117.03330	287.03888
SDLKD	190.50000	32.84300	-116.89200	286.48334
SDMEG	1304.50000	33.18800	-116.76100	282.59445
SDRMT	487.70001	33.40000	-117.19000	287.59445
SPPSD	513.29999	32.99948	-116.78854	288.70557
SRHSD	1222.30005	32.80849	-116.50830	284.81668
SSOSD	712.29999	33.07452	-116.81370	287.59445
SSSSD	988.79999	33.34400	-116.73200	284.81668
SVCSD	171.30000	32.91264	-116.95121	286.48334
SVLSD	685.20001	32.72080	-116.70200	288.14999
SYRSD	897.00000	33.13220	-116.68600	285.37222
TLKSD	380.70001	33.22617	-117.07559	285.92776
TWOSD	233.80000	33.19377	-117.15301	287.59445
VLCC1	452.00000	33.23690	-117.00860	287.59445
VLCSD	436.50000	33.23563	-117.00680	286.48334
WALSD	473.70001	32.83013	-116.80397	285.92776
WCVSD	777.90002	32.73200	-116.76200	287.03888
WDCSD	1181.69995	32.84950	-116.63000	284.81668
WPTSD	701.70001	32.61225	-116.61333	287.59445
WRBSD	228.30000	33.03110	-117.12200	286.48334
WSYSD	1003.09998	33.08680	-116.69000	283.70557
YSAC1	890.00000	33.11080	-116.67110	284.81668
YSASD	917.40002	33.11312	-116.67182	284.26111

------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jul 20 14:13:35 2020

John,

I see that you're processing 136 point observations of temperature.
And
you're placing them onto a 204x204 grid... totalling 41,616 grid
points.

Honestly, we haven't worked on the point2grid tool for many years
since
it's so problematic. There are just so many ways a person might want
to
interpolate point observations onto a grid, that it isn't conceivable
to
support them all. And frankly, none of them are very good. Using 136
point
observations to populate values at 41,616 grid points just isn't going
to
produce meaningful output. I see that the IDV output appears to have
structure, but it's all artifacts of the interpolation process. And
whatever method IDV is using is not currently supported in the
point2grid
tool. The point2grid tool works best when you have at least one point
observation falling into each grid cell, which is the case with dense
satellite data or a dense network of point stations.

With your existing data, I'd recommend using a much, much coarser
target
grid. You could try shifting from a 1-km grid to a 10-km grid, to see
if
the results make more sense.

Thanks,
John

On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> CLASSIFICATION: UNCLASSIFIED
>
> John -
>
> Welcome back. Yes, I'm still trying to understand the output gridded
file
> which places point obs on a grid.  I'm comparing what I'm seeing
from my
> testing of point2grid so far with an experiment I did last JAN where
I
> used
> IDV to produce a NetCDF file containing a set of gridded obs on a
1km
> grid.
> Ref ticket # 93980 for your assist in making the NetCDF file from
IDV
> MET_compatible. The attached text file contains the point obs which
were
> used
> by IDV to produce the gridded obs. The result was a MET-compatible
file
> which
> I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> file. If
> you look at the plot of the IDV generated gridded obs in the
attached PDF,
> it
> looks different from that which is generated by point2grid. I'm
trying to
> reconcile the two to gain further understanding.  See the attached
PDF of
> the
> point2grid output which I produced. The obs used for the point2grid
> testing
> are not the same ones as those used for the IDV testing. The NETCDF
output
> from point2grid is also attached. See list of files below.
>
> That said, I understand that Howard is working on regrid_data_plane
to
> enable
> the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> wonder
> if the Gaussian smoothing will achieve an end product which look
more
> similar
> to what I was expecting? Maybe the gridded obs I produced with IDV
isn't
> ever
> going to produce the same thing which point2grid was conceived to
produce
> due
> to the use of a different concept in performing an objective
analysis
> using
> interpolation.
>
> List of attachments:
> Case1_TMP_OBS_18_UTC - point obs text file
> 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded obs
> 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
>
> R/
> John
>
>
> -----Original Message-----
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Monday, July 20, 2020 10:56 AM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> Hi John,
>
> This is John Halley Gotway. I'm covering the met-help desk for the
next
> few
> days and wanted to check in on this issue. Are you still having
trouble
> understanding the output of point2grid, or does it make sense?
>
> Thanks,
> John
>
> On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <Caution-url:
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > The plot_data_plane can not show the point observation data. It
shows
> > the plot of the grid data.
> > In this case, the coverage of the gridded data is 0.23% (97 out of
> 41616).
> > It's the plot of the grid cells (1km by 1km) even though it looks
like
> > the point data. It caused by few converage (0.23%).
> >
> > The meaning of following log mesages (from point2grid):
> >
> > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG
> > 2: process_point_file() ->  var_count=180, grid: 97 out of 41616
> >
> > - The input data has 180 TMP data
> > - One TMP data does not belong to the target grid
> > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616 *
100)
> >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
> >
> > The plot has valid data for 97 cells out of 41616 (which is 204 by
> > 204) grid cells)
> >
> >
> > Two issues with regrid_data_plane (this specific case because of
0.23%
> > coverage).
> >
> > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008
> > (<=
> > 1 / 121).
> >
> > The default setting of "-vld_thresh" is 0.5.
> > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The matrix
for
> > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a valid
> > regrdded value for the taeget cell, it requires 61 (from 121 / 2)
> > neighbers with the valid TMP data. Only 97 cells have the valid
data and
> > they are spreaded.
> > None of them satisfies the number of the valid neighber cells.
That's
> > why the output of regrid_data_plae was all missing values.
> >
> > 2. Bug on computing the Gaussian value.
> >
> > I will fix this.
> >
> > I attached the updated NetCDF file (output of regrid_data_plane)
and
> > the plot.
> >
> > Cheers,
> > Howard
> >
> > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > Thanks for working to improve the documentation. So you can only
use
> > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
> > > unless you are working with a probability variable in which case
you
> > > can use MAXGAUSS. The default of these is UW_MEAN. The output I
> > > generated in my first test used that method since I didn't
specify
> > > the -method. In this case, the plot I generated using
> > > plot_data_plane shows a map with data points at random locations
> > > which correspond to the locations of the original point obs from
the
> > > output of ascii2nc. My expectation was that I would see the data
> > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > In theory, this NetCDF file could be used in Grid-Stat to verify
a
> > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded TMP
> > > field?
> > >
> > > R/
> > > John
> > >
> > >
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
pasting
> > > the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Yes, you are right. It's not the output you are expecting. We
will
> > > improve the documentation on this (Caution-
> > > Caution-https://github.com/NCAR/MET/issues/1413).
> > >
> > > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > for the point2grid. So the difference is between MAX VS. UW_MEAN
> > > interpolation methods, and GAUSIAN smoothing was not involved.
> > >
> > > The right way for the Gaussian smoothing to the raw data will
be:
> > > 1) regridding by using point2grid with one of interpolation
methods
> > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > 2) regridding by using regrid_data_plane with MAXGAUSS method
and
> > > the same setting as the input and the target grid.
> > >
> > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > > on this.
> > >
> > > Cheers,
> > > Howard
> > >
> > >
> > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Howard -
> > > >
> > > > Today I successfully ran point2grid but used -method MAXGAUSS
and
> > > > it ran without error and produced output. I'm uploading the
file
> > > > list, output, commands used, log file, and plots to the DoD
SAFE
> > > > so you can take a look. I am also uploading the results of the
> > > > previous run where you suggested that I use "11" vice "TMP"
which
> > > > fixed the problem and ran successfully.
> > > >
> > > > The plots are not what I was expecting somehow. The plotted
data
> > > > look like just points where the original obs were located
rather
> > > > than a regularly spaced grid of TMP values. What are your
thoughts
> on
> > > > this?
> > > >
> > > > Thanks
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity of
> > > > all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Something is not right. I will take a look
> > > >
> > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
> > > > 246.351 to 284.817.
> > > > DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";)
is
> > > > -9999
> > > > to -9999.
> > > >
> > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > output is all missing value (-9999).
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > The input file has
> > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > Howard -
> > > > >
> > > > > The next run with MAXGAUSS was successful, although I have
to
> > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > Does it indicate any issues with how it ran?
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > > -----Original Message-----
> > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > <john.w.raby2.civ at mail.mil>
> > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of all links contained within the message prior to copying
and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > MAXGAUSS and GAUSSIAN method is only used for the
interpolation.
> > > > > Please set "-method MAXGAUSE". The from_grid and to_grid are
the
> > > > > same, there is no difference with additional MAX
interpolation.
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > I tried running MET V9.0 regrid_data_plane following your
> > > > > > example below, but I'm getting and error:
> > > > > > ERROR  :
> > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid
for
> > > > > > regridding, only interpolating.
> > > > > >
> > > > > > The following is the command I used to run it:
> > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > V2016.nc
> > > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > V2016.nc
> > > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > P_1200UTC_11NOV2016_gaussian.nc
> > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > >
> > > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > ss_test
> > > > > > -v 9
> > > > > >
> > > > > > Any suggestions on how to perform Gaussian smoothing on
the
> > > > > > output TMP field from point2grid?
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify the identity of the sender, and confirm the
> > > > > > authenticity of all links contained within the message
prior
> > > > > > to copying and pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Sorry for the late response.
> > > > > >
> > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > is missing.
> > > > > > There are
> > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > >
> > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP"; level="(*,*)";'
\
> > > > > > -method GAUSSIAN -v 1
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > > Just
> > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as the
> > > > > > > input_filename to regrid_data_plane and then you provide
the
> > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
provide
> > > > > > > the to_grid. Is this because it knows that the grids of
the
> > > > > > > input and output files are the same and thus ignores the
> > > > > > > missing to_grid argument?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > too.
> > > > > > > In
> > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > regriding.
> > > > > > > The input and the target grid should be the same, which
is
> > > > > > > the output of point2grid.
> > > > > > >
> > > > > > > Here is an example (assuming the point data file is
> > > > > > > generated by pb2nc).
> > > > > > >
> > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > >       -v 1
> > > > > > >
> > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > >       -method GAUSSIAN -v 1
> > > > > > >
> > > > > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
> > > > > > > (120.0) if necessary
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > Thanks for your quick response with details in answer
to
> > > > > > > > my question.
> > > > > > > > Right now my interest is with non-probability fields,
so
> > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > smoothing for -field argument is interesting.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
> > > > > > > > V9.0 Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please
> > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > authenticity of all links contained within the message
> > > > > > > > prior to copying and pasting the address to a Web
browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for
the
> > > > > > > > raw data
> > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > > probability variable. (the existence of valid values
if
> > > > > > > > the probability threshold is not given)
> > > > > > > >
> > > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > > (which
> > > > > > > > is
> > > > > > > > a
> > > > > > > > default) for the layer 1 and "-method GAUSSIAN" for
the
> > > > > > > > layer 2.
> > > > > > > > + "-method MAXGAUSS" is processed as the "-method MAX"
for
> > > > > > > > the
> > > > > > > > layer
> > > > > > > > 1
> > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > + Override the interpolation method of layer 1 with
> > > > > > > > + "-method
> > > > > > > > <METHOD>"
> > > > > > > >
> > > > > > > > 1. The Gaussian smoothing at point2grid is activated
and
> > > > > > > > applied into the probability variable only if -method
> > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the command
line
> > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > argument. If you want, You can do this by using
> > > > > > > > regrid_data_plane (passing the output of point2grid as
the
> > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > >
> > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
> > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given. The
> > > > > > > > default values are applied if not given.
> > > > > > > >
> > > > > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN
or
> > > > > > > > -method MAXGAUSS argument.
> > > > > > > >
> > > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status
> > > > > > > > of data existence. With -prob_cat_thresh argument, it
will
> > > > > > > > be meeting the condition or not. So the values of the
> > > > > > > > probability are 0 or 1.
> > > > > > > >
> > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing
is
> > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > John,
> > > > > > > > >
> > > > > > > > > I'm reassigning this ticket to Howard Soh who worked
on
> > > > > > > > > the development of the point2grid tool. He should be
> > > > > > > > > able to better answer these questions.
> > > > > > > >
> > > > > > > > >
> > > > > > > > > Thanks,
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA
CIV
> > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> <Caution-Caution-Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > lay.html?i
> > > > > > > > > > d=95747 >
> > > > > > > > > >
> > > > > > > > > > John -
> > > > > > > > > >
> > > > > > > > > > I ran it to see the usage and there are details in
the
> > > > > > > > > > usage dialog and also in the XML file you pointed
me
> > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > There is an additional interpolation method called
> > > > > > > > > > GAUSSIAN which applies only when using the
probability
> > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > So, it
> > > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > > -gaussian_dx and -gaussian_radius are only used
when
> > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > #95747] MET
> > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > the authenticity of all links contained within the
> > > > > > > > > > message prior to copying and pasting the address
to a
> > > > > > > > > > Web browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > While the "-field" command line argument indicates
> > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > >
> > > > > > > > > > If you specify "-field" more that once you'll get
this
> > > > > > > > > > error
> > > > > > > > > > message:
> > > > > > > > > > ERROR  :
> > > > > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > > > > "2"
> > > > > > > > > > variables
> > > > > > > > > > were requested.
> > > > > > > > > > ERROR  :
> > > > > > > > > >
> > > > > > > > > > We do plan to support the processing of multiple
> > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > >
> > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > it's very simple. Keep track of the point
observations
> > > > > > > > > > falling inside each of the target grid boxes.
Write to
> > > > > > > > > > the output the min, max, unweighted mean, or
median of
> > > > > > > > > > those values based on the method chosen. If you
only
> > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > all those methods will result in the same output
> > > > > > > > > > value.
> > > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > > files, then the choice of the derivation method
would
> > > > > > > > > > matter.
> > > > > > > > > >
> > > > > > > > > > Hope that helps clarify.
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA
CIV
> > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > Caution-Caution-url:
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > ?id=
> > > > > > > > > > 95747 >
> > > > > > > > > > >
> > > > > > > > > > > Another question I have is how is -method used
in
> > > > > > > > > > > this tool?
> > > > > > > > > > > In the
> > > > > > > > > > User's
> > > > > > > > > > > Guide, -method provides options for regridding
from
> > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > regrid_data_plane. In the case of Point2Grid,
the
> > > > > > > > > > > process involved is not regridding in the way
that I
> > > > > > > > > > > understand it.
> > > > > > > > > > > If I
> > > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > > and distributes them on to a user provided grid
> > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > step I
> > > > > > > > > > > think is the smoothing of the field to even out
> > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > to some function which appear to be a Gaussian
> > > > > > > > > > > function. I guess I also need to understand what
> > > > > > > > > > > happens when the tool is used without any of the
> > > > > > > > > > > optional arguments.
> > > > > > > > > > >
> > > > > > > > > > > Thanks.
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0
> > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > John -
> > > > > > > > > > >
> > > > > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > > > > Appreciate
> > > > > > > > > > > your dedication so late into the evening! Can I
add
> > > > > > > > > > > new variables in the
> > > > > > > > > > config
> > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > interpolation methods would be appreciated, but
that
> > > > > > > > > > > can wait until Monday.
> > > > > > > > > > >
> > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > > Please verify
> > > > > > > > > > the
> > > > > > > > > > > identity of the sender, and confirm the
authenticity
> > > > > > > > > > > of all links
> > > > > > > > > > contained
> > > > > > > > > > > within the message prior to copying and pasting
the
> > > > > > > > > > > address to a Web browser.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > ----
> > > > > > > > > > >
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > I see you have a question about interpolation
> > > > > > > > > > > options in the point2grid tool. This XML file
> > > > > > > > > > > contains the calls to the point2grid tool that
are
> > > > > > > > > > run
> > > > > > > > > > > for our unit tests:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > est/xml/
> > > > > > > > > > un
> > > > > > > > > > it
> > > > > > > > > > _point2grid.xml
> > > > > > > > > > >
> > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > line option, a few at the end do use "-method
MAX"
> > > > > > > > > > > to select the maximum value falling within each
grid
> > > > > > > > > > > box.
> > > > > > > > > > >
> > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > et/src/t
> > > > > > > > > > oo
> > > > > > > > > > ls
> > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > I see that the supported options for -method are
> > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > >
> > > > > > > > > > > Seems like we should add that info to the usage
> > > > > > > > > > > statement!
> > > > > > > > > > > I
> > > > > > > > > > > wrote up
> > > > > > > > > > this
> > > > > > > > > > > GitHub issue to make sure we remember to do so
> > > > > > > > > > > before the
> > > > > > > > > > > 9.1
> > > > > > > > > > > release:
> > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > Caution-https://github.com/NCAR/MET/issu
> > > > > > > > > > > es/1396
> > > > > > > > > > >
> > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA
CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > > upon.
> > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > >       Status: new
> > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > splay.html
> > > > > > > > > > > ?id=95747
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > >
> > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > User's
> > > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > > If
> > > > > > > > > > > > so,
> > > > > > > > > > > > can you
> > > > > > > > > > > provide
> > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > argument?
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks.
> > > > > > > > > > > >
> > > > > > > > > > > > Stay safe.
> > > > > > > > > > > >
> > > > > > > > > > > > T/
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > Range, NM
> > > > > > > > > > > > 88002
> > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > Teleworking: Contact Admin Specialist at (410)
> > > > > > > > > > > > 278-
> > > > > > > > > > > > 6974
> > > > > > > > > > > > FAX
> > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > >
> > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> >
>
> CLASSIFICATION: UNCLASSIFIED
>
>

------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 20 15:19:53 2020

Thanks for sharing your thoughts and knowledge on this. Believe it or
not, this discussion is helpful from a research perspective. I've been
trying to develop ways to generated gridded obs on ever-finer grids
because of Army requirements. Having your sense of what works best in
point2grid is helpful. It is valid concern about what to do in data
sparse regions where you don't have obs. I've been verifying the URMA
gridded obs on a 2.5 km grid over CONUS and they are grappling with
the same issues. Gridded obs verify the best (against point obs) if
there is a dense network of point obs which are used as input. For
model verification, the real challenge is in areas where there are no
point obs which were used as input. The gridded obs in these point ob-
sparse areas are as good as the method used to interpolate between
sparse point obs.

Was the intent of developing point2grid to provide gridded obs for
verification? What happens if you use Grid-Stat and use the point2grid
produced gridded obs such as those I showed you for verification. Will
this generate erroneous results in the areas where you don't have
point obs or will it indicate where there were insufficient point obs
and flag it? Could one sort out were there are valid error stats and
limit the verification to just those areas where there is good
coverage of point obs or will the presence of large areas with poor
coverage result in contaminating the error statistics? Consider
running point2grid for a first pass and then based on that, possibly
dividing up the domain into subdomains using masking regions where the
gridded analysis(obs) is judged to be of good quality?

For the IDV gridded obs, I was able to specify the use of the
optimized Barnes Objective Analysis technique and could specify the
factors used to produce the analysis. The factors which can be changed
are the number of passes (2-4), the search radius (length scale) and
the gain which determines the amount of reduction applied to the
length scale for the second pass. I did some testing and obtained the
best overall analysis with 4-passes, 10 grid square search radius and
gain – 0.2 which is the minimum gain suggested by the IDV User’s
Guide. It also provides a way for you to specify the output grid
spacing. While I was able to generate output on a 1km grid, the
process is much more time consuming than running point2grid, hence my
interest in point2grid.

Based on your thought that the issue with my point obs is their
sparseness and that the recommended way to workaround this is to
upscale the grid from 1km to 10 km, I would have to conclude that it
might be better to use the URMA obs on a 2.5 km grid and downscale
them (regrid) to verify our model output on a 1km grid.  The objective
analysis technique used by RTMA/URMA is more sophisticated in that
they use a gridded background field from Rapid Refresh (RAP) model and
the High Resolution Rapid Refresh (HRRR) model. Their output is
blended with the North American Mesoscale Model (NAM) and downscaled
as a first guess background field on a 2.5 km grid. The first guess
field is then adjusted through a two-dimensional variational data
assimilation technique (2dvar) to analyze point weather observations
from the national networks of point obs. That said, I still consider
the capability to allow a user to generate a gridded filed form point
obs to be interesting. I have access to some experimental measurements
which are not being input into the national databases such as MADIS.
If I could ingest those measurements as well as the points obs we get
from MADIS into point2grid, then the quality of the gridded obs output
might exceed that from URMA for a limited area around the experimental
measurements. It's worth some experimentation.

R/
John

________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Monday, July 20, 2020 2:13 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

I see that you're processing 136 point observations of temperature.
And
you're placing them onto a 204x204 grid... totalling 41,616 grid
points.

Honestly, we haven't worked on the point2grid tool for many years
since
it's so problematic. There are just so many ways a person might want
to
interpolate point observations onto a grid, that it isn't conceivable
to
support them all. And frankly, none of them are very good. Using 136
point
observations to populate values at 41,616 grid points just isn't going
to
produce meaningful output. I see that the IDV output appears to have
structure, but it's all artifacts of the interpolation process. And
whatever method IDV is using is not currently supported in the
point2grid
tool. The point2grid tool works best when you have at least one point
observation falling into each grid cell, which is the case with dense
satellite data or a dense network of point stations.

With your existing data, I'd recommend using a much, much coarser
target
grid. You could try shifting from a 1-km grid to a 10-km grid, to see
if
the results make more sense.

Thanks,
John

On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> CLASSIFICATION: UNCLASSIFIED
>
> John -
>
> Welcome back. Yes, I'm still trying to understand the output gridded
file
> which places point obs on a grid.  I'm comparing what I'm seeing
from my
> testing of point2grid so far with an experiment I did last JAN where
I
> used
> IDV to produce a NetCDF file containing a set of gridded obs on a
1km
> grid.
> Ref ticket # 93980 for your assist in making the NetCDF file from
IDV
> MET_compatible. The attached text file contains the point obs which
were
> used
> by IDV to produce the gridded obs. The result was a MET-compatible
file
> which
> I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> file. If
> you look at the plot of the IDV generated gridded obs in the
attached PDF,
> it
> looks different from that which is generated by point2grid. I'm
trying to
> reconcile the two to gain further understanding.  See the attached
PDF of
> the
> point2grid output which I produced. The obs used for the point2grid
> testing
> are not the same ones as those used for the IDV testing. The NETCDF
output
> from point2grid is also attached. See list of files below.
>
> That said, I understand that Howard is working on regrid_data_plane
to
> enable
> the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> wonder
> if the Gaussian smoothing will achieve an end product which look
more
> similar
> to what I was expecting? Maybe the gridded obs I produced with IDV
isn't
> ever
> going to produce the same thing which point2grid was conceived to
produce
> due
> to the use of a different concept in performing an objective
analysis
> using
> interpolation.
>
> List of attachments:
> Case1_TMP_OBS_18_UTC - point obs text file
> 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded obs
> 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
>
> R/
> John
>
>
> -----Original Message-----
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Monday, July 20, 2020 10:56 AM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
> contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> Hi John,
>
> This is John Halley Gotway. I'm covering the met-help desk for the
next
> few
> days and wanted to check in on this issue. Are you still having
trouble
> understanding the output of point2grid, or does it make sense?
>
> Thanks,
> John
>
> On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <Caution-Caution-url:
> > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > The plot_data_plane can not show the point observation data. It
shows
> > the plot of the grid data.
> > In this case, the coverage of the gridded data is 0.23% (97 out of
> 41616).
> > It's the plot of the grid cells (1km by 1km) even though it looks
like
> > the point data. It caused by few converage (0.23%).
> >
> > The meaning of following log mesages (from point2grid):
> >
> > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG
> > 2: process_point_file() ->  var_count=180, grid: 97 out of 41616
> >
> > - The input data has 180 TMP data
> > - One TMP data does not belong to the target grid
> > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616 *
100)
> >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
> >
> > The plot has valid data for 97 cells out of 41616 (which is 204 by
> > 204) grid cells)
> >
> >
> > Two issues with regrid_data_plane (this specific case because of
0.23%
> > coverage).
> >
> > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008
> > (<=
> > 1 / 121).
> >
> > The default setting of "-vld_thresh" is 0.5.
> > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The matrix
for
> > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a valid
> > regrdded value for the taeget cell, it requires 61 (from 121 / 2)
> > neighbers with the valid TMP data. Only 97 cells have the valid
data and
> > they are spreaded.
> > None of them satisfies the number of the valid neighber cells.
That's
> > why the output of regrid_data_plae was all missing values.
> >
> > 2. Bug on computing the Gaussian value.
> >
> > I will fix this.
> >
> > I attached the updated NetCDF file (output of regrid_data_plane)
and
> > the plot.
> >
> > Cheers,
> > Howard
> >
> > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > Thanks for working to improve the documentation. So you can only
use
> > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
> > > unless you are working with a probability variable in which case
you
> > > can use MAXGAUSS. The default of these is UW_MEAN. The output I
> > > generated in my first test used that method since I didn't
specify
> > > the -method. In this case, the plot I generated using
> > > plot_data_plane shows a map with data points at random locations
> > > which correspond to the locations of the original point obs from
the
> > > output of ascii2nc. My expectation was that I would see the data
> > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > In theory, this NetCDF file could be used in Grid-Stat to verify
a
> > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded TMP
> > > field?
> > >
> > > R/
> > > John
> > >
> > >
> > > ________________________________________
> > > From: Howard Soh via RT [met_help at ucar.edu]
> > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
> > > verify the identity of the sender, and confirm the authenticity
of
> > > all links contained within the message prior to copying and
pasting
> > > the address to a Web browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Yes, you are right. It's not the output you are expecting. We
will
> > > improve the documentation on this (Caution-
> > > Caution-Caution-https://github.com/NCAR/MET/issues/1413).
> > >
> > > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > for the point2grid. So the difference is between MAX VS. UW_MEAN
> > > interpolation methods, and GAUSIAN smoothing was not involved.
> > >
> > > The right way for the Gaussian smoothing to the raw data will
be:
> > > 1) regridding by using point2grid with one of interpolation
methods
> > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > 2) regridding by using regrid_data_plane with MAXGAUSS method
and
> > > the same setting as the input and the target grid.
> > >
> > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > > on this.
> > >
> > > Cheers,
> > > Howard
> > >
> > >
> > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Howard -
> > > >
> > > > Today I successfully ran point2grid but used -method MAXGAUSS
and
> > > > it ran without error and produced output. I'm uploading the
file
> > > > list, output, commands used, log file, and plots to the DoD
SAFE
> > > > so you can take a look. I am also uploading the results of the
> > > > previous run where you suggested that I use "11" vice "TMP"
which
> > > > fixed the problem and ran successfully.
> > > >
> > > > The plots are not what I was expecting somehow. The plotted
data
> > > > look like just points where the original obs were located
rather
> > > > than a regularly spaced grid of TMP values. What are your
thoughts
> on
> > > > this?
> > > >
> > > > Thanks
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity of
> > > > all links contained within the message prior to copying and
> > > > pasting the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Something is not right. I will take a look
> > > >
> > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
> > > > 246.351 to 284.817.
> > > > DEBUG 2: Range of regridded data (name="TMP"; level="(*,*)";)
is
> > > > -9999
> > > > to -9999.
> > > >
> > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > output is all missing value (-9999).
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > The input file has
> > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > > Howard -
> > > > >
> > > > > The next run with MAXGAUSS was successful, although I have
to
> > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > Does it indicate any issues with how it ran?
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > > -----Original Message-----
> > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > <john.w.raby2.civ at mail.mil>
> > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > of all links contained within the message prior to copying
and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > MAXGAUSS and GAUSSIAN method is only used for the
interpolation.
> > > > > Please set "-method MAXGAUSE". The from_grid and to_grid are
the
> > > > > same, there is no difference with additional MAX
interpolation.
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > I tried running MET V9.0 regrid_data_plane following your
> > > > > > example below, but I'm getting and error:
> > > > > > ERROR  :
> > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid
for
> > > > > > regridding, only interpolating.
> > > > > >
> > > > > > The following is the command I used to run it:
> > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > V2016.nc
> > > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > V2016.nc
> > > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > P_1200UTC_11NOV2016_gaussian.nc
> > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > >
> > > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > ss_test
> > > > > > -v 9
> > > > > >
> > > > > > Any suggestions on how to perform Gaussian smoothing on
the
> > > > > > output TMP field from point2grid?
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
> > > > > > Please verify the identity of the sender, and confirm the
> > > > > > authenticity of all links contained within the message
prior
> > > > > > to copying and pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Sorry for the late response.
> > > > > >
> > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > is missing.
> > > > > > There are
> > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > >
> > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP"; level="(*,*)";'
\
> > > > > > -method GAUSSIAN -v 1
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > > Just
> > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as the
> > > > > > > input_filename to regrid_data_plane and then you provide
the
> > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
provide
> > > > > > > the to_grid. Is this because it knows that the grids of
the
> > > > > > > input and output files are the same and thus ignores the
> > > > > > > missing to_grid argument?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > too.
> > > > > > > In
> > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > regriding.
> > > > > > > The input and the target grid should be the same, which
is
> > > > > > > the output of point2grid.
> > > > > > >
> > > > > > > Here is an example (assuming the point data file is
> > > > > > > generated by pb2nc).
> > > > > > >
> > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > >       -v 1
> > > > > > >
> > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > >       -method GAUSSIAN -v 1
> > > > > > >
> > > > > > > Note: override -gaussian_dx (81.271) & -gaussian_radius
> > > > > > > (120.0) if necessary
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > Thanks for your quick response with details in answer
to
> > > > > > > > my question.
> > > > > > > > Right now my interest is with non-probability fields,
so
> > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > smoothing for -field argument is interesting.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > > ________________________________________
> > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
> > > > > > > > V9.0 Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please
> > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > authenticity of all links contained within the message
> > > > > > > > prior to copying and pasting the address to a Web
browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for
the
> > > > > > > > raw data
> > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > > probability variable. (the existence of valid values
if
> > > > > > > > the probability threshold is not given)
> > > > > > > >
> > > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > > (which
> > > > > > > > is
> > > > > > > > a
> > > > > > > > default) for the layer 1 and "-method GAUSSIAN" for
the
> > > > > > > > layer 2.
> > > > > > > > + "-method MAXGAUSS" is processed as the "-method MAX"
for
> > > > > > > > the
> > > > > > > > layer
> > > > > > > > 1
> > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > + Override the interpolation method of layer 1 with
> > > > > > > > + "-method
> > > > > > > > <METHOD>"
> > > > > > > >
> > > > > > > > 1. The Gaussian smoothing at point2grid is activated
and
> > > > > > > > applied into the probability variable only if -method
> > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the command
line
> > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > argument. If you want, You can do this by using
> > > > > > > > regrid_data_plane (passing the output of point2grid as
the
> > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > >
> > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
> > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given. The
> > > > > > > > default values are applied if not given.
> > > > > > > >
> > > > > > > > 3 -prob_cat_thresh applied regardless -method GAUSSIAN
or
> > > > > > > > -method MAXGAUSS argument.
> > > > > > > >
> > > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status
> > > > > > > > of data existence. With -prob_cat_thresh argument, it
will
> > > > > > > > be meeting the condition or not. So the values of the
> > > > > > > > probability are 0 or 1.
> > > > > > > >
> > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN smoothing
is
> > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > John,
> > > > > > > > >
> > > > > > > > > I'm reassigning this ticket to Howard Soh who worked
on
> > > > > > > > > the development of the point2grid tool. He should be
> > > > > > > > > able to better answer these questions.
> > > > > > > >
> > > > > > > > >
> > > > > > > > > Thanks,
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA
CIV
> > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
url:
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > lay.html?i
> > > > > > > > > > d=95747 >
> > > > > > > > > >
> > > > > > > > > > John -
> > > > > > > > > >
> > > > > > > > > > I ran it to see the usage and there are details in
the
> > > > > > > > > > usage dialog and also in the XML file you pointed
me
> > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > There is an additional interpolation method called
> > > > > > > > > > GAUSSIAN which applies only when using the
probability
> > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > So, it
> > > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > > -gaussian_dx and -gaussian_radius are only used
when
> > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > #95747] MET
> > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > the authenticity of all links contained within the
> > > > > > > > > > message prior to copying and pasting the address
to a
> > > > > > > > > > Web browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > While the "-field" command line argument indicates
> > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > >
> > > > > > > > > > If you specify "-field" more that once you'll get
this
> > > > > > > > > > error
> > > > > > > > > > message:
> > > > > > > > > > ERROR  :
> > > > > > > > > > ERROR  : process_point_file() -> Support only 1
variable.
> > > > > > > > > > "2"
> > > > > > > > > > variables
> > > > > > > > > > were requested.
> > > > > > > > > > ERROR  :
> > > > > > > > > >
> > > > > > > > > > We do plan to support the processing of multiple
> > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > >
> > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > it's very simple. Keep track of the point
observations
> > > > > > > > > > falling inside each of the target grid boxes.
Write to
> > > > > > > > > > the output the min, max, unweighted mean, or
median of
> > > > > > > > > > those values based on the method chosen. If you
only
> > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > all those methods will result in the same output
> > > > > > > > > > value.
> > > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > > files, then the choice of the derivation method
would
> > > > > > > > > > matter.
> > > > > > > > > >
> > > > > > > > > > Hope that helps clarify.
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA
CIV
> > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > Caution-Caution-Caution-url:
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
> > > > > > > > > > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > ?id=
> > > > > > > > > > 95747 >
> > > > > > > > > > >
> > > > > > > > > > > Another question I have is how is -method used
in
> > > > > > > > > > > this tool?
> > > > > > > > > > > In the
> > > > > > > > > > User's
> > > > > > > > > > > Guide, -method provides options for regridding
from
> > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > regrid_data_plane. In the case of Point2Grid,
the
> > > > > > > > > > > process involved is not regridding in the way
that I
> > > > > > > > > > > understand it.
> > > > > > > > > > > If I
> > > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > > and distributes them on to a user provided grid
> > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > step I
> > > > > > > > > > > think is the smoothing of the field to even out
> > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > to some function which appear to be a Gaussian
> > > > > > > > > > > function. I guess I also need to understand what
> > > > > > > > > > > happens when the tool is used without any of the
> > > > > > > > > > > optional arguments.
> > > > > > > > > > >
> > > > > > > > > > > Thanks.
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0
> > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > John -
> > > > > > > > > > >
> > > > > > > > > > > Thanks for your quick response which is helpful.
> > > > > > > > > > > Appreciate
> > > > > > > > > > > your dedication so late into the evening! Can I
add
> > > > > > > > > > > new variables in the
> > > > > > > > > > config
> > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > interpolation methods would be appreciated, but
that
> > > > > > > > > > > can wait until Monday.
> > > > > > > > > > >
> > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > > Please verify
> > > > > > > > > > the
> > > > > > > > > > > identity of the sender, and confirm the
authenticity
> > > > > > > > > > > of all links
> > > > > > > > > > contained
> > > > > > > > > > > within the message prior to copying and pasting
the
> > > > > > > > > > > address to a Web browser.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > ----
> > > > > > > > > > >
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > I see you have a question about interpolation
> > > > > > > > > > > options in the point2grid tool. This XML file
> > > > > > > > > > > contains the calls to the point2grid tool that
are
> > > > > > > > > > run
> > > > > > > > > > > for our unit tests:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > est/xml/
> > > > > > > > > > un
> > > > > > > > > > it
> > > > > > > > > > _point2grid.xml
> > > > > > > > > > >
> > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > line option, a few at the end do use "-method
MAX"
> > > > > > > > > > > to select the maximum value falling within each
grid
> > > > > > > > > > > box.
> > > > > > > > > > >
> > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-
> > > > > > > > > > Caution-Caution-
https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > et/src/t
> > > > > > > > > > oo
> > > > > > > > > > ls
> > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > I see that the supported options for -method are
> > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > >
> > > > > > > > > > > Seems like we should add that info to the usage
> > > > > > > > > > > statement!
> > > > > > > > > > > I
> > > > > > > > > > > wrote up
> > > > > > > > > > this
> > > > > > > > > > > GitHub issue to make sure we remember to do so
> > > > > > > > > > > before the
> > > > > > > > > > > 9.1
> > > > > > > > > > > release:
> > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > Caution-Caution-https://github.com/NCAR/MET/issu
> > > > > > > > > > > es/1396
> > > > > > > > > > >
> > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W USA
CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > > upon.
> > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > >       Status: new
> > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > splay.html
> > > > > > > > > > > ?id=95747
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > >
> > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > User's
> > > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > > If
> > > > > > > > > > > > so,
> > > > > > > > > > > > can you
> > > > > > > > > > > provide
> > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > argument?
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks.
> > > > > > > > > > > >
> > > > > > > > > > > > Stay safe.
> > > > > > > > > > > >
> > > > > > > > > > > > T/
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > Range, NM
> > > > > > > > > > > > 88002
> > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > Teleworking: Contact Admin Specialist at (410)
> > > > > > > > > > > > 278-
> > > > > > > > > > > > 6974
> > > > > > > > > > > > FAX
> > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > >
> > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > > >
> > > > > CLASSIFICATION: UNCLASSIFIED
> >
> >
> >
> >
>
> CLASSIFICATION: UNCLASSIFIED
>
>



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jul 20 15:38:13 2020

John,

Yes, I agree that using URMA is a good idea. There are several people
working on that analysis, and generally, I'd trust their results more
than
the output of the point2grid tool. This tool is still very new, and,
as
you've illustrated, there are many other methods that could be added
to it.
I'll leave it to the scientists and project managers to determine our
development priorities within our funding constraints.

I want to point out though that Python embedding could be a very nice
option here. I would imagine that Python packages exist which support
many
methods of putting point data onto a grid. When researching potential
ways
of doing so, you could have MET execute a custom Python script and
read the
resulting gridded data into grid_stat or mode.

John

On Mon, Jul 20, 2020 at 3:20 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Thanks for sharing your thoughts and knowledge on this. Believe it
or not,
> this discussion is helpful from a research perspective. I've been
trying to
> develop ways to generated gridded obs on ever-finer grids because of
Army
> requirements. Having your sense of what works best in point2grid is
> helpful. It is valid concern about what to do in data sparse regions
where
> you don't have obs. I've been verifying the URMA gridded obs on a
2.5 km
> grid over CONUS and they are grappling with the same issues. Gridded
obs
> verify the best (against point obs) if there is a dense network of
point
> obs which are used as input. For model verification, the real
challenge is
> in areas where there are no point obs which were used as input. The
gridded
> obs in these point ob-sparse areas are as good as the method used to
> interpolate between sparse point obs.
>
> Was the intent of developing point2grid to provide gridded obs for
> verification? What happens if you use Grid-Stat and use the
point2grid
> produced gridded obs such as those I showed you for verification.
Will this
> generate erroneous results in the areas where you don't have point
obs or
> will it indicate where there were insufficient point obs and flag
it? Could
> one sort out were there are valid error stats and limit the
verification to
> just those areas where there is good coverage of point obs or will
the
> presence of large areas with poor coverage result in contaminating
the
> error statistics? Consider running point2grid for a first pass and
then
> based on that, possibly dividing up the domain into subdomains using
> masking regions where the gridded analysis(obs) is judged to be of
good
> quality?
>
> For the IDV gridded obs, I was able to specify the use of the
optimized
> Barnes Objective Analysis technique and could specify the factors
used to
> produce the analysis. The factors which can be changed are the
number of
> passes (2-4), the search radius (length scale) and the gain which
> determines the amount of reduction applied to the length scale for
the
> second pass. I did some testing and obtained the best overall
analysis with
> 4-passes, 10 grid square search radius and gain – 0.2 which is the
minimum
> gain suggested by the IDV User’s Guide. It also provides a way for
you to
> specify the output grid spacing. While I was able to generate output
on a
> 1km grid, the process is much more time consuming than running
point2grid,
> hence my interest in point2grid.
>
> Based on your thought that the issue with my point obs is their
sparseness
> and that the recommended way to workaround this is to upscale the
grid from
> 1km to 10 km, I would have to conclude that it might be better to
use the
> URMA obs on a 2.5 km grid and downscale them (regrid) to verify our
model
> output on a 1km grid.  The objective analysis technique used by
RTMA/URMA
> is more sophisticated in that they use a gridded background field
from
> Rapid Refresh (RAP) model and the High Resolution Rapid Refresh
(HRRR)
> model. Their output is blended with the North American Mesoscale
Model
> (NAM) and downscaled as a first guess background field on a 2.5 km
grid.
> The first guess field is then adjusted through a two-dimensional
> variational data assimilation technique (2dvar) to analyze point
weather
> observations from the national networks of point obs. That said, I
still
> consider the capability to allow a user to generate a gridded filed
form
> point obs to be interesting. I have access to !
>  some experimental measurements which are not being input into the
> national databases such as MADIS. If I could ingest those
measurements as
> well as the points obs we get from MADIS into point2grid, then the
quality
> of the gridded obs output might exceed that from URMA for a limited
area
> around the experimental measurements. It's worth some
experimentation.
>
> R/
> John
>
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, July 20, 2020 2:13 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> I see that you're processing 136 point observations of temperature.
And
> you're placing them onto a 204x204 grid... totalling 41,616 grid
points.
>
> Honestly, we haven't worked on the point2grid tool for many years
since
> it's so problematic. There are just so many ways a person might want
to
> interpolate point observations onto a grid, that it isn't
conceivable to
> support them all. And frankly, none of them are very good. Using 136
point
> observations to populate values at 41,616 grid points just isn't
going to
> produce meaningful output. I see that the IDV output appears to have
> structure, but it's all artifacts of the interpolation process. And
> whatever method IDV is using is not currently supported in the
point2grid
> tool. The point2grid tool works best when you have at least one
point
> observation falling into each grid cell, which is the case with
dense
> satellite data or a dense network of point stations.
>
> With your existing data, I'd recommend using a much, much coarser
target
> grid. You could try shifting from a 1-km grid to a 10-km grid, to
see if
> the results make more sense.
>
> Thanks,
> John
>
> On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-url: Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> > John -
> >
> > Welcome back. Yes, I'm still trying to understand the output
gridded file
> > which places point obs on a grid.  I'm comparing what I'm seeing
from my
> > testing of point2grid so far with an experiment I did last JAN
where I
> > used
> > IDV to produce a NetCDF file containing a set of gridded obs on a
1km
> > grid.
> > Ref ticket # 93980 for your assist in making the NetCDF file from
IDV
> > MET_compatible. The attached text file contains the point obs
which were
> > used
> > by IDV to produce the gridded obs. The result was a MET-compatible
file
> > which
> > I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> > file. If
> > you look at the plot of the IDV generated gridded obs in the
attached
> PDF,
> > it
> > looks different from that which is generated by point2grid. I'm
trying to
> > reconcile the two to gain further understanding.  See the attached
PDF of
> > the
> > point2grid output which I produced. The obs used for the
point2grid
> > testing
> > are not the same ones as those used for the IDV testing. The
NETCDF
> output
> > from point2grid is also attached. See list of files below.
> >
> > That said, I understand that Howard is working on
regrid_data_plane to
> > enable
> > the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> > wonder
> > if the Gaussian smoothing will achieve an end product which look
more
> > similar
> > to what I was expecting? Maybe the gridded obs I produced with IDV
isn't
> > ever
> > going to produce the same thing which point2grid was conceived to
produce
> > due
> > to the use of a different concept in performing an objective
analysis
> > using
> > interpolation.
> >
> > List of attachments:
> > Case1_TMP_OBS_18_UTC - point obs text file
> > 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> > 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded obs
> > 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> > 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
> >
> > R/
> > John
> >
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Monday, July 20, 2020 10:56 AM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > Hi John,
> >
> > This is John Halley Gotway. I'm covering the met-help desk for the
next
> > few
> > days and wanted to check in on this issue. Are you still having
trouble
> > understanding the output of point2grid, or does it make sense?
> >
> > Thanks,
> > John
> >
> > On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <Caution-Caution-url:
> > > Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > The plot_data_plane can not show the point observation data. It
shows
> > > the plot of the grid data.
> > > In this case, the coverage of the gridded data is 0.23% (97 out
of
> > 41616).
> > > It's the plot of the grid cells (1km by 1km) even though it
looks like
> > > the point data. It caused by few converage (0.23%).
> > >
> > > The meaning of following log mesages (from point2grid):
> > >
> > > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG
> > > 2: process_point_file() ->  var_count=180, grid: 97 out of 41616
> > >
> > > - The input data has 180 TMP data
> > > - One TMP data does not belong to the target grid
> > > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616
* 100)
> > >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
> > >
> > > The plot has valid data for 97 cells out of 41616 (which is 204
by
> > > 204) grid cells)
> > >
> > >
> > > Two issues with regrid_data_plane (this specific case because of
0.23%
> > > coverage).
> > >
> > > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008
> > > (<=
> > > 1 / 121).
> > >
> > > The default setting of "-vld_thresh" is 0.5.
> > > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The
matrix for
> > > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a
valid
> > > regrdded value for the taeget cell, it requires 61 (from 121 /
2)
> > > neighbers with the valid TMP data. Only 97 cells have the valid
data
> and
> > > they are spreaded.
> > > None of them satisfies the number of the valid neighber cells.
That's
> > > why the output of regrid_data_plae was all missing values.
> > >
> > > 2. Bug on computing the Gaussian value.
> > >
> > > I will fix this.
> > >
> > > I attached the updated NetCDF file (output of regrid_data_plane)
and
> > > the plot.
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Thanks for working to improve the documentation. So you can
only use
> > > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
> > > > unless you are working with a probability variable in which
case you
> > > > can use MAXGAUSS. The default of these is UW_MEAN. The output
I
> > > > generated in my first test used that method since I didn't
specify
> > > > the -method. In this case, the plot I generated using
> > > > plot_data_plane shows a map with data points at random
locations
> > > > which correspond to the locations of the original point obs
from the
> > > > output of ascii2nc. My expectation was that I would see the
data
> > > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > > In theory, this NetCDF file could be used in Grid-Stat to
verify a
> > > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded
> TMP
> > > > field?
> > > >
> > > > R/
> > > > John
> > > >
> > > >
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity of
> > > > all links contained within the message prior to copying and
pasting
> > > > the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Yes, you are right. It's not the output you are expecting. We
will
> > > > improve the documentation on this (Caution-
> > > > Caution-Caution-https://github.com/NCAR/MET/issues/1413).
> > > >
> > > > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > > for the point2grid. So the difference is between MAX VS.
UW_MEAN
> > > > interpolation methods, and GAUSIAN smoothing was not involved.
> > > >
> > > > The right way for the Gaussian smoothing to the raw data will
be:
> > > > 1) regridding by using point2grid with one of interpolation
methods
> > > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > > 2) regridding by using regrid_data_plane with MAXGAUSS method
and
> > > > the same setting as the input and the target grid.
> > > >
> > > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > > > on this.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > >
> > > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Howard -
> > > > >
> > > > > Today I successfully ran point2grid but used -method
MAXGAUSS and
> > > > > it ran without error and produced output. I'm uploading the
file
> > > > > list, output, commands used, log file, and plots to the DoD
SAFE
> > > > > so you can take a look. I am also uploading the results of
the
> > > > > previous run where you suggested that I use "11" vice "TMP"
which
> > > > > fixed the problem and ran successfully.
> > > > >
> > > > > The plots are not what I was expecting somehow. The plotted
data
> > > > > look like just points where the original obs were located
rather
> > > > > than a regularly spaced grid of TMP values. What are your
thoughts
> > on
> > > > > this?
> > > > >
> > > > > Thanks
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity of
> > > > > all links contained within the message prior to copying and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Something is not right. I will take a look
> > > > >
> > > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
> > > > > 246.351 to 284.817.
> > > > > DEBUG 2: Range of regridded data (name="TMP";
level="(*,*)";) is
> > > > > -9999
> > > > > to -9999.
> > > > >
> > > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > > output is all missing value (-9999).
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > The input file has
> > > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > The next run with MAXGAUSS was successful, although I have
to
> > > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > > Does it indicate any issues with how it ran?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > > of all links contained within the message prior to copying
and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > > MAXGAUSS and GAUSSIAN method is only used for the
interpolation.
> > > > > > Please set "-method MAXGAUSE". The from_grid and to_grid
are the
> > > > > > same, there is no difference with additional MAX
interpolation.
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > >
> > > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > I tried running MET V9.0 regrid_data_plane following
your
> > > > > > > example below, but I'm getting and error:
> > > > > > > ERROR  :
> > > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid
for
> > > > > > > regridding, only interpolating.
> > > > > > >
> > > > > > > The following is the command I used to run it:
> > > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > > >
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > V2016.nc
> > > > > > >
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > V2016.nc
> > > > > > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > > P_1200UTC_11NOV2016_gaussian.nc
> > > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > > >
> > > > > > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > > ss_test
> > > > > > > -v 9
> > > > > > >
> > > > > > > Any suggestions on how to perform Gaussian smoothing on
the
> > > > > > > output TMP field from point2grid?
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > Sorry for the late response.
> > > > > > >
> > > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > > is missing.
> > > > > > > There are
> > > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > > >
> > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP";
level="(*,*)";' \
> > > > > > > -method GAUSSIAN -v 1
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > > > Just
> > > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as the
> > > > > > > > input_filename to regrid_data_plane and then you
provide the
> > > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
provide
> > > > > > > > the to_grid. Is this because it knows that the grids
of the
> > > > > > > > input and output files are the same and thus ignores
the
> > > > > > > > missing to_grid argument?
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > >
> > > > > > > > -----Original Message-----
> > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET V9.0
> > > > > > > > Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > > authenticity of all links contained within the message
prior
> > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > > too.
> > > > > > > > In
> > > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > > regriding.
> > > > > > > > The input and the target grid should be the same,
which is
> > > > > > > > the output of point2grid.
> > > > > > > >
> > > > > > > > Here is an example (assuming the point data file is
> > > > > > > > generated by pb2nc).
> > > > > > > >
> > > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > > >       -v 1
> > > > > > > >
> > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > > >       -method GAUSSIAN -v 1
> > > > > > > >
> > > > > > > > Note: override -gaussian_dx (81.271) &
-gaussian_radius
> > > > > > > > (120.0) if necessary
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
> wrote:
> > > > > > > > > Howard -
> > > > > > > > >
> > > > > > > > > Thanks for your quick response with details in
answer to
> > > > > > > > > my question.
> > > > > > > > > Right now my interest is with non-probability
fields, so
> > > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > > smoothing for -field argument is interesting.
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> > > > > > > > > V9.0 Point2Grid
> > > > > > > > > (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please
> > > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > prior to copying and pasting the address to a Web
browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for
the
> > > > > > > > > raw data
> > > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > > > probability variable. (the existence of valid values
if
> > > > > > > > > the probability threshold is not given)
> > > > > > > > >
> > > > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > > > (which
> > > > > > > > > is
> > > > > > > > > a
> > > > > > > > > default) for the layer 1 and "-method GAUSSIAN" for
the
> > > > > > > > > layer 2.
> > > > > > > > > + "-method MAXGAUSS" is processed as the "-method
MAX" for
> > > > > > > > > the
> > > > > > > > > layer
> > > > > > > > > 1
> > > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > > + Override the interpolation method of layer 1 with
> > > > > > > > > + "-method
> > > > > > > > > <METHOD>"
> > > > > > > > >
> > > > > > > > > 1. The Gaussian smoothing at point2grid is activated
and
> > > > > > > > > applied into the probability variable only if
-method
> > > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the command
line
> > > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > > argument. If you want, You can do this by using
> > > > > > > > > regrid_data_plane (passing the output of point2grid
as the
> > > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > > >
> > > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
> > > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given.
The
> > > > > > > > > default values are applied if not given.
> > > > > > > > >
> > > > > > > > > 3 -prob_cat_thresh applied regardless -method
GAUSSIAN or
> > > > > > > > > -method MAXGAUSS argument.
> > > > > > > > >
> > > > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status
> > > > > > > > > of data existence. With -prob_cat_thresh argument,
it will
> > > > > > > > > be meeting the condition or not. So the values of
the
> > > > > > > > > probability are 0 or 1.
> > > > > > > > >
> > > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN
smoothing is
> > > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > > >
> > > > > > > > > Cheers,
> > > > > > > > > Howard
> > > > > > > > >
> > > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > I'm reassigning this ticket to Howard Soh who
worked on
> > > > > > > > > > the development of the point2grid tool. He should
be
> > > > > > > > > > able to better answer these questions.
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA
CIV
> > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
url:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > > lay.html?i
> > > > > > > > > > > d=95747 >
> > > > > > > > > > >
> > > > > > > > > > > John -
> > > > > > > > > > >
> > > > > > > > > > > I ran it to see the usage and there are details
in the
> > > > > > > > > > > usage dialog and also in the XML file you
pointed me
> > > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > > There is an additional interpolation method
called
> > > > > > > > > > > GAUSSIAN which applies only when using the
probability
> > > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > > So, it
> > > > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > > > -gaussian_dx and -gaussian_radius are only used
when
> > > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Subject: Re: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > > the authenticity of all links contained within
the
> > > > > > > > > > > message prior to copying and pasting the address
to a
> > > > > > > > > > > Web browser.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > ----
> > > > > > > > > > >
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > While the "-field" command line argument
indicates
> > > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > > >
> > > > > > > > > > > If you specify "-field" more that once you'll
get this
> > > > > > > > > > > error
> > > > > > > > > > > message:
> > > > > > > > > > > ERROR  :
> > > > > > > > > > > ERROR  : process_point_file() -> Support only 1
> variable.
> > > > > > > > > > > "2"
> > > > > > > > > > > variables
> > > > > > > > > > > were requested.
> > > > > > > > > > > ERROR  :
> > > > > > > > > > >
> > > > > > > > > > > We do plan to support the processing of multiple
> > > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > > >
> > > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > > it's very simple. Keep track of the point
observations
> > > > > > > > > > > falling inside each of the target grid boxes.
Write to
> > > > > > > > > > > the output the min, max, unweighted mean, or
median of
> > > > > > > > > > > those values based on the method chosen. If you
only
> > > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > > all those methods will result in the same output
> > > > > > > > > > > value.
> > > > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > > > files, then the choice of the derivation method
would
> > > > > > > > > > > matter.
> > > > > > > > > > >
> > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA
CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > Caution-Caution-Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > > ?id=
> > > > > > > > > > > 95747 >
> > > > > > > > > > > >
> > > > > > > > > > > > Another question I have is how is -method used
in
> > > > > > > > > > > > this tool?
> > > > > > > > > > > > In the
> > > > > > > > > > > User's
> > > > > > > > > > > > Guide, -method provides options for regridding
from
> > > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > > regrid_data_plane. In the case of Point2Grid,
the
> > > > > > > > > > > > process involved is not regridding in the way
that I
> > > > > > > > > > > > understand it.
> > > > > > > > > > > > If I
> > > > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > > > and distributes them on to a user provided
grid
> > > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > > step I
> > > > > > > > > > > > think is the smoothing of the field to even
out
> > > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > > to some function which appear to be a Gaussian
> > > > > > > > > > > > function. I guess I also need to understand
what
> > > > > > > > > > > > happens when the tool is used without any of
the
> > > > > > > > > > > > optional arguments.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks.
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0
> > > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > John -
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks for your quick response which is
helpful.
> > > > > > > > > > > > Appreciate
> > > > > > > > > > > > your dedication so late into the evening! Can
I add
> > > > > > > > > > > > new variables in the
> > > > > > > > > > > config
> > > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > > interpolation methods would be appreciated,
but that
> > > > > > > > > > > > can wait until Monday.
> > > > > > > > > > > >
> > > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > All active links contained in this email were
> disabled.
> > > > > > > > > > > > Please verify
> > > > > > > > > > > the
> > > > > > > > > > > > identity of the sender, and confirm the
authenticity
> > > > > > > > > > > > of all links
> > > > > > > > > > > contained
> > > > > > > > > > > > within the message prior to copying and
pasting the
> > > > > > > > > > > > address to a Web browser.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > ----
> > > > > > > > > > > >
> > > > > > > > > > > > John,
> > > > > > > > > > > >
> > > > > > > > > > > > I see you have a question about interpolation
> > > > > > > > > > > > options in the point2grid tool. This XML file
> > > > > > > > > > > > contains the calls to the point2grid tool that
are
> > > > > > > > > > > run
> > > > > > > > > > > > for our unit tests:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > > est/xml/
> > > > > > > > > > > un
> > > > > > > > > > > it
> > > > > > > > > > > _point2grid.xml
> > > > > > > > > > > >
> > > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > > line option, a few at the end do use "-method
MAX"
> > > > > > > > > > > > to select the maximum value falling within
each grid
> > > > > > > > > > > > box.
> > > > > > > > > > > >
> > > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > > et/src/t
> > > > > > > > > > > oo
> > > > > > > > > > > ls
> > > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > > I see that the supported options for -method
are
> > > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > > >
> > > > > > > > > > > > Seems like we should add that info to the
usage
> > > > > > > > > > > > statement!
> > > > > > > > > > > > I
> > > > > > > > > > > > wrote up
> > > > > > > > > > > this
> > > > > > > > > > > > GitHub issue to make sure we remember to do so
> > > > > > > > > > > > before the
> > > > > > > > > > > > 9.1
> > > > > > > > > > > > release:
> > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
https://github.com/NCAR/MET/issu
> > > > > > > > > > > > es/1396
> > > > > > > > > > > >
> > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks,
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W
USA CIV
> > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > > > upon.
> > > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >       Status: new
> > > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > > splay.html
> > > > > > > > > > > > ?id=95747
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > >
> > > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > > User's
> > > > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > > > If
> > > > > > > > > > > > > so,
> > > > > > > > > > > > > can you
> > > > > > > > > > > > provide
> > > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > > argument?
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Stay safe.
> > > > > > > > > > > > >
> > > > > > > > > > > > > T/
> > > > > > > > > > > > > John
> > > > > > > > > > > > >
> > > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > > Range, NM
> > > > > > > > > > > > > 88002
> > > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > > Teleworking: Contact Admin Specialist at
(410)
> > > > > > > > > > > > > 278-
> > > > > > > > > > > > > 6974
> > > > > > > > > > > > > FAX
> > > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >
> > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> > >
> > >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> >
>
>
>
>

------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 20 15:47:39 2020

John -

Thanks for your assessment of URMA. I agree. NCEP/NWS developed it
with the goal of providing an analysis of record for doing model
verification. This seems like it take on the form/significance of an
accreditation. I apprecaite your tip on the use of Python embedding.
Will this work with the way I'm running MET via a container? I'm not
up on how to use Python embedding. If it is not integrated with MET
maybe running MET in a container is not an issue.

R/
John
________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Monday, July 20, 2020 3:38 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

John,

Yes, I agree that using URMA is a good idea. There are several people
working on that analysis, and generally, I'd trust their results more
than
the output of the point2grid tool. This tool is still very new, and,
as
you've illustrated, there are many other methods that could be added
to it.
I'll leave it to the scientists and project managers to determine our
development priorities within our funding constraints.

I want to point out though that Python embedding could be a very nice
option here. I would imagine that Python packages exist which support
many
methods of putting point data onto a grid. When researching potential
ways
of doing so, you could have MET execute a custom Python script and
read the
resulting gridded data into grid_stat or mode.

John

On Mon, Jul 20, 2020 at 3:20 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Thanks for sharing your thoughts and knowledge on this. Believe it
or not,
> this discussion is helpful from a research perspective. I've been
trying to
> develop ways to generated gridded obs on ever-finer grids because of
Army
> requirements. Having your sense of what works best in point2grid is
> helpful. It is valid concern about what to do in data sparse regions
where
> you don't have obs. I've been verifying the URMA gridded obs on a
2.5 km
> grid over CONUS and they are grappling with the same issues. Gridded
obs
> verify the best (against point obs) if there is a dense network of
point
> obs which are used as input. For model verification, the real
challenge is
> in areas where there are no point obs which were used as input. The
gridded
> obs in these point ob-sparse areas are as good as the method used to
> interpolate between sparse point obs.
>
> Was the intent of developing point2grid to provide gridded obs for
> verification? What happens if you use Grid-Stat and use the
point2grid
> produced gridded obs such as those I showed you for verification.
Will this
> generate erroneous results in the areas where you don't have point
obs or
> will it indicate where there were insufficient point obs and flag
it? Could
> one sort out were there are valid error stats and limit the
verification to
> just those areas where there is good coverage of point obs or will
the
> presence of large areas with poor coverage result in contaminating
the
> error statistics? Consider running point2grid for a first pass and
then
> based on that, possibly dividing up the domain into subdomains using
> masking regions where the gridded analysis(obs) is judged to be of
good
> quality?
>
> For the IDV gridded obs, I was able to specify the use of the
optimized
> Barnes Objective Analysis technique and could specify the factors
used to
> produce the analysis. The factors which can be changed are the
number of
> passes (2-4), the search radius (length scale) and the gain which
> determines the amount of reduction applied to the length scale for
the
> second pass. I did some testing and obtained the best overall
analysis with
> 4-passes, 10 grid square search radius and gain – 0.2 which is the
minimum
> gain suggested by the IDV User’s Guide. It also provides a way for
you to
> specify the output grid spacing. While I was able to generate output
on a
> 1km grid, the process is much more time consuming than running
point2grid,
> hence my interest in point2grid.
>
> Based on your thought that the issue with my point obs is their
sparseness
> and that the recommended way to workaround this is to upscale the
grid from
> 1km to 10 km, I would have to conclude that it might be better to
use the
> URMA obs on a 2.5 km grid and downscale them (regrid) to verify our
model
> output on a 1km grid.  The objective analysis technique used by
RTMA/URMA
> is more sophisticated in that they use a gridded background field
from
> Rapid Refresh (RAP) model and the High Resolution Rapid Refresh
(HRRR)
> model. Their output is blended with the North American Mesoscale
Model
> (NAM) and downscaled as a first guess background field on a 2.5 km
grid.
> The first guess field is then adjusted through a two-dimensional
> variational data assimilation technique (2dvar) to analyze point
weather
> observations from the national networks of point obs. That said, I
still
> consider the capability to allow a user to generate a gridded filed
form
> point obs to be interesting. I have access to !
>  some experimental measurements which are not being input into the
> national databases such as MADIS. If I could ingest those
measurements as
> well as the points obs we get from MADIS into point2grid, then the
quality
> of the gridded obs output might exceed that from URMA for a limited
area
> around the experimental measurements. It's worth some
experimentation.
>
> R/
> John
>
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, July 20, 2020 2:13 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> I see that you're processing 136 point observations of temperature.
And
> you're placing them onto a 204x204 grid... totalling 41,616 grid
points.
>
> Honestly, we haven't worked on the point2grid tool for many years
since
> it's so problematic. There are just so many ways a person might want
to
> interpolate point observations onto a grid, that it isn't
conceivable to
> support them all. And frankly, none of them are very good. Using 136
point
> observations to populate values at 41,616 grid points just isn't
going to
> produce meaningful output. I see that the IDV output appears to have
> structure, but it's all artifacts of the interpolation process. And
> whatever method IDV is using is not currently supported in the
point2grid
> tool. The point2grid tool works best when you have at least one
point
> observation falling into each grid cell, which is the case with
dense
> satellite data or a dense network of point stations.
>
> With your existing data, I'd recommend using a much, much coarser
target
> grid. You could try shifting from a 1-km grid to a 10-km grid, to
see if
> the results make more sense.
>
> Thanks,
> John
>
> On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-Caution-url: Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> > John -
> >
> > Welcome back. Yes, I'm still trying to understand the output
gridded file
> > which places point obs on a grid.  I'm comparing what I'm seeing
from my
> > testing of point2grid so far with an experiment I did last JAN
where I
> > used
> > IDV to produce a NetCDF file containing a set of gridded obs on a
1km
> > grid.
> > Ref ticket # 93980 for your assist in making the NetCDF file from
IDV
> > MET_compatible. The attached text file contains the point obs
which were
> > used
> > by IDV to produce the gridded obs. The result was a MET-compatible
file
> > which
> > I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> > file. If
> > you look at the plot of the IDV generated gridded obs in the
attached
> PDF,
> > it
> > looks different from that which is generated by point2grid. I'm
trying to
> > reconcile the two to gain further understanding.  See the attached
PDF of
> > the
> > point2grid output which I produced. The obs used for the
point2grid
> > testing
> > are not the same ones as those used for the IDV testing. The
NETCDF
> output
> > from point2grid is also attached. See list of files below.
> >
> > That said, I understand that Howard is working on
regrid_data_plane to
> > enable
> > the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> > wonder
> > if the Gaussian smoothing will achieve an end product which look
more
> > similar
> > to what I was expecting? Maybe the gridded obs I produced with IDV
isn't
> > ever
> > going to produce the same thing which point2grid was conceived to
produce
> > due
> > to the use of a different concept in performing an objective
analysis
> > using
> > interpolation.
> >
> > List of attachments:
> > Case1_TMP_OBS_18_UTC - point obs text file
> > 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> > 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded obs
> > 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> > 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
> >
> > R/
> > John
> >
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Monday, July 20, 2020 10:56 AM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > Point2Grid
> > (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> > contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > Hi John,
> >
> > This is John Halley Gotway. I'm covering the met-help desk for the
next
> > few
> > days and wanted to check in on this issue. Are you still having
trouble
> > understanding the output of point2grid, or does it make sense?
> >
> > Thanks,
> > John
> >
> > On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <Caution-Caution-Caution-url:
> > > Caution-Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > The plot_data_plane can not show the point observation data. It
shows
> > > the plot of the grid data.
> > > In this case, the coverage of the gridded data is 0.23% (97 out
of
> > 41616).
> > > It's the plot of the grid cells (1km by 1km) even though it
looks like
> > > the point data. It caused by few converage (0.23%).
> > >
> > > The meaning of following log mesages (from point2grid):
> > >
> > > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180 (99%)
DEBUG
> > > 2: process_point_file() ->  var_count=180, grid: 97 out of 41616
> > >
> > > - The input data has 180 TMP data
> > > - One TMP data does not belong to the target grid
> > > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 / 41616
* 100)
> > >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more TMP
data
> > >
> > > The plot has valid data for 97 cells out of 41616 (which is 204
by
> > > 204) grid cells)
> > >
> > >
> > > Two issues with regrid_data_plane (this specific case because of
0.23%
> > > coverage).
> > >
> > > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh) OR
0.008
> > > (<=
> > > 1 / 121).
> > >
> > > The default setting of "-vld_thresh" is 0.5.
> > > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The
matrix for
> > > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a
valid
> > > regrdded value for the taeget cell, it requires 61 (from 121 /
2)
> > > neighbers with the valid TMP data. Only 97 cells have the valid
data
> and
> > > they are spreaded.
> > > None of them satisfies the number of the valid neighber cells.
That's
> > > why the output of regrid_data_plae was all missing values.
> > >
> > > 2. Bug on computing the Gaussian value.
> > >
> > > I will fix this.
> > >
> > > I attached the updated NetCDF file (output of regrid_data_plane)
and
> > > the plot.
> > >
> > > Cheers,
> > > Howard
> > >
> > > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > > Thanks for working to improve the documentation. So you can
only use
> > > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in point2grid
> > > > unless you are working with a probability variable in which
case you
> > > > can use MAXGAUSS. The default of these is UW_MEAN. The output
I
> > > > generated in my first test used that method since I didn't
specify
> > > > the -method. In this case, the plot I generated using
> > > > plot_data_plane shows a map with data points at random
locations
> > > > which correspond to the locations of the original point obs
from the
> > > > output of ascii2nc. My expectation was that I would see the
data
> > > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > > In theory, this NetCDF file could be used in Grid-Stat to
verify a
> > > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded
> TMP
> > > > field?
> > > >
> > > > R/
> > > > John
> > > >
> > > >
> > > > ________________________________________
> > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> > > > verify the identity of the sender, and confirm the
authenticity of
> > > > all links contained within the message prior to copying and
pasting
> > > > the address to a Web browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Yes, you are right. It's not the output you are expecting. We
will
> > > > improve the documentation on this (Caution-
> > > > Caution-Caution-Caution-
https://github.com/NCAR/MET/issues/1413).
> > > >
> > > > Unlike regrid_data_plane, the MAXGAUSE method with point2grid
works
> > > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > > smoothing into the probability variable (TMP_prob_NA or
TMP_prob_XX).
> > > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > > for the point2grid. So the difference is between MAX VS.
UW_MEAN
> > > > interpolation methods, and GAUSIAN smoothing was not involved.
> > > >
> > > > The right way for the Gaussian smoothing to the raw data will
be:
> > > > 1) regridding by using point2grid with one of interpolation
methods
> > > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > > 2) regridding by using regrid_data_plane with MAXGAUSS method
and
> > > > the same setting as the input and the target grid.
> > > >
> > > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
working
> > > > on this.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > >
> > > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Howard -
> > > > >
> > > > > Today I successfully ran point2grid but used -method
MAXGAUSS and
> > > > > it ran without error and produced output. I'm uploading the
file
> > > > > list, output, commands used, log file, and plots to the DoD
SAFE
> > > > > so you can take a look. I am also uploading the results of
the
> > > > > previous run where you suggested that I use "11" vice "TMP"
which
> > > > > fixed the problem and ran successfully.
> > > > >
> > > > > The plots are not what I was expecting somehow. The plotted
data
> > > > > look like just points where the original obs were located
rather
> > > > > than a regularly spaced grid of TMP values. What are your
thoughts
> > on
> > > > > this?
> > > > >
> > > > > Thanks
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity of
> > > > > all links contained within the message prior to copying and
> > > > > pasting the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Something is not right. I will take a look
> > > > >
> > > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";) is
> > > > > 246.351 to 284.817.
> > > > > DEBUG 2: Range of regridded data (name="TMP";
level="(*,*)";) is
> > > > > -9999
> > > > > to -9999.
> > > > >
> > > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > > output is all missing value (-9999).
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > The input file has
> > > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > > Howard -
> > > > > >
> > > > > > The next run with MAXGAUSS was successful, although I have
to
> > > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > > Does it indicate any issues with how it ran?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > > -----Original Message-----
> > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > <john.w.raby2.civ at mail.mil>
> > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > > of all links contained within the message prior to copying
and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > > MAXGAUSS and GAUSSIAN method is only used for the
interpolation.
> > > > > > Please set "-method MAXGAUSE". The from_grid and to_grid
are the
> > > > > > same, there is no difference with additional MAX
interpolation.
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > >
> > > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > I tried running MET V9.0 regrid_data_plane following
your
> > > > > > > example below, but I'm getting and error:
> > > > > > > ERROR  :
> > > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not valid
for
> > > > > > > regridding, only interpolating.
> > > > > > >
> > > > > > > The following is the command I used to run it:
> > > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > > >
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > V2016.nc
> > > > > > >
> > >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > V2016.nc
> > > > > > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > > P_1200UTC_11NOV2016_gaussian.nc
> > > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > > >
> > > > > > >
> > >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > > ss_test
> > > > > > > -v 9
> > > > > > >
> > > > > > > Any suggestions on how to perform Gaussian smoothing on
the
> > > > > > > output TMP field from point2grid?
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
> > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > authenticity of all links contained within the message
prior
> > > > > > > to copying and pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > Sorry for the late response.
> > > > > > >
> > > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > > is missing.
> > > > > > > There are
> > > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > > >
> > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP";
level="(*,*)";' \
> > > > > > > -method GAUSSIAN -v 1
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > Thanks for tailoring the command for my particular
interests.
> > > > > > > > Just
> > > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as the
> > > > > > > > input_filename to regrid_data_plane and then you
provide the
> > > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
provide
> > > > > > > > the to_grid. Is this because it knows that the grids
of the
> > > > > > > > input and output files are the same and thus ignores
the
> > > > > > > > missing to_grid argument?
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > >
> > > > > > > > -----Original Message-----
> > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET V9.0
> > > > > > > > Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > > authenticity of all links contained within the message
prior
> > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > > too.
> > > > > > > > In
> > > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > > regriding.
> > > > > > > > The input and the target grid should be the same,
which is
> > > > > > > > the output of point2grid.
> > > > > > > >
> > > > > > > > Here is an example (assuming the point data file is
> > > > > > > > generated by pb2nc).
> > > > > > > >
> > > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > > >       -v 1
> > > > > > > >
> > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > > >       -method GAUSSIAN -v 1
> > > > > > > >
> > > > > > > > Note: override -gaussian_dx (81.271) &
-gaussian_radius
> > > > > > > > (120.0) if necessary
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > > On Mon Jun 29 14:13:44 2020, john.w.raby2.civ at mail.mil
> wrote:
> > > > > > > > > Howard -
> > > > > > > > >
> > > > > > > > > Thanks for your quick response with details in
answer to
> > > > > > > > > my question.
> > > > > > > > > Right now my interest is with non-probability
fields, so
> > > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > > smoothing for -field argument is interesting.
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > > ________________________________________
> > > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> > > > > > > > > V9.0 Point2Grid
> > > > > > > > > (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please
> > > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > prior to copying and pasting the address to a Web
browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation for
the
> > > > > > > > > raw data
> > > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for the
> > > > > > > > > probability variable. (the existence of valid values
if
> > > > > > > > > the probability threshold is not given)
> > > > > > > > >
> > > > > > > > > + "-method GAUSSIAN" is processed as the "-method
UW_MEAN"
> > > > > > > > > (which
> > > > > > > > > is
> > > > > > > > > a
> > > > > > > > > default) for the layer 1 and "-method GAUSSIAN" for
the
> > > > > > > > > layer 2.
> > > > > > > > > + "-method MAXGAUSS" is processed as the "-method
MAX" for
> > > > > > > > > the
> > > > > > > > > layer
> > > > > > > > > 1
> > > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > > + Override the interpolation method of layer 1 with
> > > > > > > > > + "-method
> > > > > > > > > <METHOD>"
> > > > > > > > >
> > > > > > > > > 1. The Gaussian smoothing at point2grid is activated
and
> > > > > > > > > applied into the probability variable only if
-method
> > > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the command
line
> > > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > > argument. If you want, You can do this by using
> > > > > > > > > regrid_data_plane (passing the output of point2grid
as the
> > > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > > >
> > > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored if
> > > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given.
The
> > > > > > > > > default values are applied if not given.
> > > > > > > > >
> > > > > > > > > 3 -prob_cat_thresh applied regardless -method
GAUSSIAN or
> > > > > > > > > -method MAXGAUSS argument.
> > > > > > > > >
> > > > > > > > > 4 Without -prob_cat_thresh argument, it will be the
status
> > > > > > > > > of data existence. With -prob_cat_thresh argument,
it will
> > > > > > > > > be meeting the condition or not. So the values of
the
> > > > > > > > > probability are 0 or 1.
> > > > > > > > >
> > > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN
smoothing is
> > > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > > >
> > > > > > > > > Cheers,
> > > > > > > > > Howard
> > > > > > > > >
> > > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > > John,
> > > > > > > > > >
> > > > > > > > > > I'm reassigning this ticket to Howard Soh who
worked on
> > > > > > > > > > the development of the point2grid tool. He should
be
> > > > > > > > > > able to better answer these questions.
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W USA
CIV
> > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > > lay.html?i
> > > > > > > > > > > d=95747 >
> > > > > > > > > > >
> > > > > > > > > > > John -
> > > > > > > > > > >
> > > > > > > > > > > I ran it to see the usage and there are details
in the
> > > > > > > > > > > usage dialog and also in the XML file you
pointed me
> > > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > > There is an additional interpolation method
called
> > > > > > > > > > > GAUSSIAN which applies only when using the
probability
> > > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > > So, it
> > > > > > > > > > > looks like the two optional smoothing arguments
> > > > > > > > > > > -gaussian_dx and -gaussian_radius are only used
when
> > > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Subject: Re: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > #95747] MET
> > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > > the authenticity of all links contained within
the
> > > > > > > > > > > message prior to copying and pasting the address
to a
> > > > > > > > > > > Web browser.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > ----
> > > > > > > > > > >
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > While the "-field" command line argument
indicates
> > > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > > >
> > > > > > > > > > > If you specify "-field" more that once you'll
get this
> > > > > > > > > > > error
> > > > > > > > > > > message:
> > > > > > > > > > > ERROR  :
> > > > > > > > > > > ERROR  : process_point_file() -> Support only 1
> variable.
> > > > > > > > > > > "2"
> > > > > > > > > > > variables
> > > > > > > > > > > were requested.
> > > > > > > > > > > ERROR  :
> > > > > > > > > > >
> > > > > > > > > > > We do plan to support the processing of multiple
> > > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > > >
> > > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > > it's very simple. Keep track of the point
observations
> > > > > > > > > > > falling inside each of the target grid boxes.
Write to
> > > > > > > > > > > the output the min, max, unweighted mean, or
median of
> > > > > > > > > > > those values based on the method chosen. If you
only
> > > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > > all those methods will result in the same output
> > > > > > > > > > > value.
> > > > > > > > > > > Whereas if you have many, as in the case of GOES
input
> > > > > > > > > > > files, then the choice of the derivation method
would
> > > > > > > > > > > matter.
> > > > > > > > > > >
> > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W USA
CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > Caution-Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > > ?id=
> > > > > > > > > > > 95747 >
> > > > > > > > > > > >
> > > > > > > > > > > > Another question I have is how is -method used
in
> > > > > > > > > > > > this tool?
> > > > > > > > > > > > In the
> > > > > > > > > > > User's
> > > > > > > > > > > > Guide, -method provides options for regridding
from
> > > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > > regrid_data_plane. In the case of Point2Grid,
the
> > > > > > > > > > > > process involved is not regridding in the way
that I
> > > > > > > > > > > > understand it.
> > > > > > > > > > > > If I
> > > > > > > > > > > > understand it correctly, it takes point
observations
> > > > > > > > > > > > and distributes them on to a user provided
grid
> > > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > > step I
> > > > > > > > > > > > think is the smoothing of the field to even
out
> > > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > > to some function which appear to be a Gaussian
> > > > > > > > > > > > function. I guess I also need to understand
what
> > > > > > > > > > > > happens when the tool is used without any of
the
> > > > > > > > > > > > optional arguments.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks.
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0
> > > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > John -
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks for your quick response which is
helpful.
> > > > > > > > > > > > Appreciate
> > > > > > > > > > > > your dedication so late into the evening! Can
I add
> > > > > > > > > > > > new variables in the
> > > > > > > > > > > config
> > > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > > interpolation methods would be appreciated,
but that
> > > > > > > > > > > > can wait until Monday.
> > > > > > > > > > > >
> > > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > All active links contained in this email were
> disabled.
> > > > > > > > > > > > Please verify
> > > > > > > > > > > the
> > > > > > > > > > > > identity of the sender, and confirm the
authenticity
> > > > > > > > > > > > of all links
> > > > > > > > > > > contained
> > > > > > > > > > > > within the message prior to copying and
pasting the
> > > > > > > > > > > > address to a Web browser.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > ----
> > > > > > > > > > > >
> > > > > > > > > > > > John,
> > > > > > > > > > > >
> > > > > > > > > > > > I see you have a question about interpolation
> > > > > > > > > > > > options in the point2grid tool. This XML file
> > > > > > > > > > > > contains the calls to the point2grid tool that
are
> > > > > > > > > > > run
> > > > > > > > > > > > for our unit tests:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> Caution-https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > > est/xml/
> > > > > > > > > > > un
> > > > > > > > > > > it
> > > > > > > > > > > _point2grid.xml
> > > > > > > > > > > >
> > > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > > line option, a few at the end do use "-method
MAX"
> > > > > > > > > > > > to select the maximum value falling within
each grid
> > > > > > > > > > > > box.
> > > > > > > > > > > >
> > > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-
> > > > > > > > > > > Caution-Caution-
> Caution-https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > > et/src/t
> > > > > > > > > > > oo
> > > > > > > > > > > ls
> > > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > > I see that the supported options for -method
are
> > > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > > >
> > > > > > > > > > > > Seems like we should add that info to the
usage
> > > > > > > > > > > > statement!
> > > > > > > > > > > > I
> > > > > > > > > > > > wrote up
> > > > > > > > > > > this
> > > > > > > > > > > > GitHub issue to make sure we remember to do so
> > > > > > > > > > > > before the
> > > > > > > > > > > > 9.1
> > > > > > > > > > > > release:
> > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-
https://github.com/NCAR/MET/issu
> > > > > > > > > > > > es/1396
> > > > > > > > > > > >
> > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks,
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W
USA CIV
> > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747 was
acted
> > > > > > > > > > > > > upon.
> > > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >       Status: new
> > > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > > splay.html
> > > > > > > > > > > > ?id=95747
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > >
> > > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > > User's
> > > > > > > > > > > > > Guide is Gaussian, but are other methods
available?
> > > > > > > > > > > > > If
> > > > > > > > > > > > > so,
> > > > > > > > > > > > > can you
> > > > > > > > > > > > provide
> > > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > > argument?
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Stay safe.
> > > > > > > > > > > > >
> > > > > > > > > > > > > T/
> > > > > > > > > > > > > John
> > > > > > > > > > > > >
> > > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > > Range, NM
> > > > > > > > > > > > > 88002
> > > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > > Teleworking: Contact Admin Specialist at
(410)
> > > > > > > > > > > > > 278-
> > > > > > > > > > > > > 6974
> > > > > > > > > > > > > FAX
> > > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > >
> > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > >
> > > > > >
> > > > > >
> > > > > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> > >
> > >
> >
> > CLASSIFICATION: UNCLASSIFIED
> >
> >
>
>
>
>



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jul 20 16:47:02 2020

I have not actually run Python embedding with MET from inside the
container, but there's no reason why it shouldn't work. Python
embedding is
enabled in the Dockerfile build of MET:
https://github.com/NCAR/MET/blob/master_v9.0/scripts/docker/Dockerfile

However, I can see how it could get tricky. The MET image includes an
instance of python version 3.6... but it includes very few python
packages.
Let's say you want a python script which loads the pygrib module.
Everytime
you instantiate that image, you'd get a container without the pygrib
module. So you'd install that package everytime you run... which would
be
inefficient. It'd be better to just build a new image on top of the
MET one
with whatever python packages you want.

So not impossible, but perhaps tricky.

John

On Mon, Jul 20, 2020 at 3:48 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> John -
>
> Thanks for your assessment of URMA. I agree. NCEP/NWS developed it
with
> the goal of providing an analysis of record for doing model
verification.
> This seems like it take on the form/significance of an
accreditation. I
> apprecaite your tip on the use of Python embedding. Will this work
with the
> way I'm running MET via a container? I'm not up on how to use Python
> embedding. If it is not integrated with MET maybe running MET in a
> container is not an issue.
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, July 20, 2020 3:38 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> Yes, I agree that using URMA is a good idea. There are several
people
> working on that analysis, and generally, I'd trust their results
more than
> the output of the point2grid tool. This tool is still very new, and,
as
> you've illustrated, there are many other methods that could be added
to it.
> I'll leave it to the scientists and project managers to determine
our
> development priorities within our funding constraints.
>
> I want to point out though that Python embedding could be a very
nice
> option here. I would imagine that Python packages exist which
support many
> methods of putting point data onto a grid. When researching
potential ways
> of doing so, you could have MET execute a custom Python script and
read the
> resulting gridded data into grid_stat or mode.
>
> John
>
> On Mon, Jul 20, 2020 at 3:20 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-url: Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > Thanks for sharing your thoughts and knowledge on this. Believe it
or
> not,
> > this discussion is helpful from a research perspective. I've been
trying
> to
> > develop ways to generated gridded obs on ever-finer grids because
of Army
> > requirements. Having your sense of what works best in point2grid
is
> > helpful. It is valid concern about what to do in data sparse
regions
> where
> > you don't have obs. I've been verifying the URMA gridded obs on a
2.5 km
> > grid over CONUS and they are grappling with the same issues.
Gridded obs
> > verify the best (against point obs) if there is a dense network of
point
> > obs which are used as input. For model verification, the real
challenge
> is
> > in areas where there are no point obs which were used as input.
The
> gridded
> > obs in these point ob-sparse areas are as good as the method used
to
> > interpolate between sparse point obs.
> >
> > Was the intent of developing point2grid to provide gridded obs for
> > verification? What happens if you use Grid-Stat and use the
point2grid
> > produced gridded obs such as those I showed you for verification.
Will
> this
> > generate erroneous results in the areas where you don't have point
obs or
> > will it indicate where there were insufficient point obs and flag
it?
> Could
> > one sort out were there are valid error stats and limit the
verification
> to
> > just those areas where there is good coverage of point obs or will
the
> > presence of large areas with poor coverage result in contaminating
the
> > error statistics? Consider running point2grid for a first pass and
then
> > based on that, possibly dividing up the domain into subdomains
using
> > masking regions where the gridded analysis(obs) is judged to be of
good
> > quality?
> >
> > For the IDV gridded obs, I was able to specify the use of the
optimized
> > Barnes Objective Analysis technique and could specify the factors
used to
> > produce the analysis. The factors which can be changed are the
number of
> > passes (2-4), the search radius (length scale) and the gain which
> > determines the amount of reduction applied to the length scale for
the
> > second pass. I did some testing and obtained the best overall
analysis
> with
> > 4-passes, 10 grid square search radius and gain – 0.2 which is the
> minimum
> > gain suggested by the IDV User’s Guide. It also provides a way for
you to
> > specify the output grid spacing. While I was able to generate
output on a
> > 1km grid, the process is much more time consuming than running
> point2grid,
> > hence my interest in point2grid.
> >
> > Based on your thought that the issue with my point obs is their
> sparseness
> > and that the recommended way to workaround this is to upscale the
grid
> from
> > 1km to 10 km, I would have to conclude that it might be better to
use the
> > URMA obs on a 2.5 km grid and downscale them (regrid) to verify
our model
> > output on a 1km grid.  The objective analysis technique used by
RTMA/URMA
> > is more sophisticated in that they use a gridded background field
from
> > Rapid Refresh (RAP) model and the High Resolution Rapid Refresh
(HRRR)
> > model. Their output is blended with the North American Mesoscale
Model
> > (NAM) and downscaled as a first guess background field on a 2.5 km
grid.
> > The first guess field is then adjusted through a two-dimensional
> > variational data assimilation technique (2dvar) to analyze point
weather
> > observations from the national networks of point obs. That said, I
still
> > consider the capability to allow a user to generate a gridded
filed form
> > point obs to be interesting. I have access to !
> >  some experimental measurements which are not being input into the
> > national databases such as MADIS. If I could ingest those
measurements as
> > well as the points obs we get from MADIS into point2grid, then the
> quality
> > of the gridded obs output might exceed that from URMA for a
limited area
> > around the experimental measurements. It's worth some
experimentation.
> >
> > R/
> > John
> >
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Monday, July 20, 2020 2:13 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > I see that you're processing 136 point observations of
temperature. And
> > you're placing them onto a 204x204 grid... totalling 41,616 grid
points.
> >
> > Honestly, we haven't worked on the point2grid tool for many years
since
> > it's so problematic. There are just so many ways a person might
want to
> > interpolate point observations onto a grid, that it isn't
conceivable to
> > support them all. And frankly, none of them are very good. Using
136
> point
> > observations to populate values at 41,616 grid points just isn't
going to
> > produce meaningful output. I see that the IDV output appears to
have
> > structure, but it's all artifacts of the interpolation process.
And
> > whatever method IDV is using is not currently supported in the
point2grid
> > tool. The point2grid tool works best when you have at least one
point
> > observation falling into each grid cell, which is the case with
dense
> > satellite data or a dense network of point stations.
> >
> > With your existing data, I'd recommend using a much, much coarser
target
> > grid. You could try shifting from a 1-km grid to a 10-km grid, to
see if
> > the results make more sense.
> >
> > Thanks,
> > John
> >
> > On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-Caution-url: Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > John -
> > >
> > > Welcome back. Yes, I'm still trying to understand the output
gridded
> file
> > > which places point obs on a grid.  I'm comparing what I'm seeing
from
> my
> > > testing of point2grid so far with an experiment I did last JAN
where I
> > > used
> > > IDV to produce a NetCDF file containing a set of gridded obs on
a 1km
> > > grid.
> > > Ref ticket # 93980 for your assist in making the NetCDF file
from IDV
> > > MET_compatible. The attached text file contains the point obs
which
> were
> > > used
> > > by IDV to produce the gridded obs. The result was a MET-
compatible file
> > > which
> > > I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> > > file. If
> > > you look at the plot of the IDV generated gridded obs in the
attached
> > PDF,
> > > it
> > > looks different from that which is generated by point2grid. I'm
trying
> to
> > > reconcile the two to gain further understanding.  See the
attached PDF
> of
> > > the
> > > point2grid output which I produced. The obs used for the
point2grid
> > > testing
> > > are not the same ones as those used for the IDV testing. The
NETCDF
> > output
> > > from point2grid is also attached. See list of files below.
> > >
> > > That said, I understand that Howard is working on
regrid_data_plane to
> > > enable
> > > the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> > > wonder
> > > if the Gaussian smoothing will achieve an end product which look
more
> > > similar
> > > to what I was expecting? Maybe the gridded obs I produced with
IDV
> isn't
> > > ever
> > > going to produce the same thing which point2grid was conceived
to
> produce
> > > due
> > > to the use of a different concept in performing an objective
analysis
> > > using
> > > interpolation.
> > >
> > > List of attachments:
> > > Case1_TMP_OBS_18_UTC - point obs text file
> > > 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> > > 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded
obs
> > > 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> > > 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
> > >
> > > R/
> > > John
> > >
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT <met_help at ucar.edu>
> > > Sent: Monday, July 20, 2020 10:56 AM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
verify
> > the
> > > identity of the sender, and confirm the authenticity of all
links
> > > contained
> > > within the message prior to copying and pasting the address to a
Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Hi John,
> > >
> > > This is John Halley Gotway. I'm covering the met-help desk for
the next
> > > few
> > > days and wanted to check in on this issue. Are you still having
trouble
> > > understanding the output of point2grid, or does it make sense?
> > >
> > > Thanks,
> > > John
> > >
> > > On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-url:
> > > > Caution-Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > >
> > > > The plot_data_plane can not show the point observation data.
It shows
> > > > the plot of the grid data.
> > > > In this case, the coverage of the gridded data is 0.23% (97
out of
> > > 41616).
> > > > It's the plot of the grid cells (1km by 1km) even though it
looks
> like
> > > > the point data. It caused by few converage (0.23%).
> > > >
> > > > The meaning of following log mesages (from point2grid):
> > > >
> > > > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180
(99%)
> DEBUG
> > > > 2: process_point_file() ->  var_count=180, grid: 97 out of
41616
> > > >
> > > > - The input data has 180 TMP data
> > > > - One TMP data does not belong to the target grid
> > > > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 /
41616 *
> 100)
> > > >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more
TMP data
> > > >
> > > > The plot has valid data for 97 cells out of 41616 (which is
204 by
> > > > 204) grid cells)
> > > >
> > > >
> > > > Two issues with regrid_data_plane (this specific case because
of
> 0.23%
> > > > coverage).
> > > >
> > > > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh)
OR 0.008
> > > > (<=
> > > > 1 / 121).
> > > >
> > > > The default setting of "-vld_thresh" is 0.5.
> > > > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The
matrix
> for
> > > > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a
valid
> > > > regrdded value for the taeget cell, it requires 61 (from 121 /
2)
> > > > neighbers with the valid TMP data. Only 97 cells have the
valid data
> > and
> > > > they are spreaded.
> > > > None of them satisfies the number of the valid neighber cells.
That's
> > > > why the output of regrid_data_plae was all missing values.
> > > >
> > > > 2. Bug on computing the Gaussian value.
> > > >
> > > > I will fix this.
> > > >
> > > > I attached the updated NetCDF file (output of
regrid_data_plane) and
> > > > the plot.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Thanks for working to improve the documentation. So you can
only
> use
> > > > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in
point2grid
> > > > > unless you are working with a probability variable in which
case
> you
> > > > > can use MAXGAUSS. The default of these is UW_MEAN. The
output I
> > > > > generated in my first test used that method since I didn't
specify
> > > > > the -method. In this case, the plot I generated using
> > > > > plot_data_plane shows a map with data points at random
locations
> > > > > which correspond to the locations of the original point obs
from
> the
> > > > > output of ascii2nc. My expectation was that I would see the
data
> > > > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > > > In theory, this NetCDF file could be used in Grid-Stat to
verify a
> > > > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded
> > TMP
> > > > > field?
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > >
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity of
> > > > > all links contained within the message prior to copying and
pasting
> > > > > the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Yes, you are right. It's not the output you are expecting.
We will
> > > > > improve the documentation on this (Caution-
> > > > > Caution-Caution-Caution-
https://github.com/NCAR/MET/issues/1413).
> > > > >
> > > > > Unlike regrid_data_plane, the MAXGAUSE method with
point2grid works
> > > > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > > > smoothing into the probability variable (TMP_prob_NA or
> TMP_prob_XX).
> > > > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > > > for the point2grid. So the difference is between MAX VS.
UW_MEAN
> > > > > interpolation methods, and GAUSIAN smoothing was not
involved.
> > > > >
> > > > > The right way for the Gaussian smoothing to the raw data
will be:
> > > > > 1) regridding by using point2grid with one of interpolation
methods
> > > > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > > > 2) regridding by using regrid_data_plane with MAXGAUSS
method and
> > > > > the same setting as the input and the target grid.
> > > > >
> > > > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
> working
> > > > > on this.
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > Howard -
> > > > > >
> > > > > > Today I successfully ran point2grid but used -method
MAXGAUSS and
> > > > > > it ran without error and produced output. I'm uploading
the file
> > > > > > list, output, commands used, log file, and plots to the
DoD SAFE
> > > > > > so you can take a look. I am also uploading the results of
the
> > > > > > previous run where you suggested that I use "11" vice
"TMP" which
> > > > > > fixed the problem and ran successfully.
> > > > > >
> > > > > > The plots are not what I was expecting somehow. The
plotted data
> > > > > > look like just points where the original obs were located
rather
> > > > > > than a regularly spaced grid of TMP values. What are your
> thoughts
> > > on
> > > > > > this?
> > > > > >
> > > > > > Thanks
> > > > > > ________________________________________
> > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> of
> > > > > > all links contained within the message prior to copying
and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Something is not right. I will take a look
> > > > > >
> > > > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";)
is
> > > > > > 246.351 to 284.817.
> > > > > > DEBUG 2: Range of regridded data (name="TMP";
level="(*,*)";) is
> > > > > > -9999
> > > > > > to -9999.
> > > > > >
> > > > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > > > output is all missing value (-9999).
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > > The input file has
> > > > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > The next run with MAXGAUSS was successful, although I
have to
> > > > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > > > Does it indicate any issues with how it ran?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
Please
> > > > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > > > of all links contained within the message prior to
copying and
> > > > > > > pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > > > MAXGAUSS and GAUSSIAN method is only used for the
> interpolation.
> > > > > > > Please set "-method MAXGAUSE". The from_grid and to_grid
are
> the
> > > > > > > same, there is no difference with additional MAX
interpolation.
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > >
> > > > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > I tried running MET V9.0 regrid_data_plane following
your
> > > > > > > > example below, but I'm getting and error:
> > > > > > > > ERROR  :
> > > > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not
valid for
> > > > > > > > regridding, only interpolating.
> > > > > > > >
> > > > > > > > The following is the command I used to run it:
> > > > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > V2016.nc
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > V2016.nc
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > > > P_1200UTC_11NOV2016_gaussian.nc
> > > > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > > > >
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > > > ss_test
> > > > > > > > -v 9
> > > > > > > >
> > > > > > > > Any suggestions on how to perform Gaussian smoothing
on the
> > > > > > > > output TMP field from point2grid?
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > >
> > > > > > > > -----Original Message-----
> > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET V9.0
> > > > > > > > Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > > authenticity of all links contained within the message
prior
> > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > Sorry for the late response.
> > > > > > > >
> > > > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > > > is missing.
> > > > > > > > There are
> > > > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > > > >
> > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP";
level="(*,*)";' \
> > > > > > > > -method GAUSSIAN -v 1
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
> wrote:
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > > Howard -
> > > > > > > > >
> > > > > > > > > Thanks for tailoring the command for my particular
> interests.
> > > > > > > > > Just
> > > > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as
the
> > > > > > > > > input_filename to regrid_data_plane and then you
provide
> the
> > > > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
> provide
> > > > > > > > > the to_grid. Is this because it knows that the grids
of the
> > > > > > > > > input and output files are the same and thus ignores
the
> > > > > > > > > missing to_grid argument?
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > -----Original Message-----
> > > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> V9.0
> > > > > > > > > Point2Grid
> > > > > > > > > (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify the identity of the sender, and
confirm the
> > > > > > > > > authenticity of all links contained within the
message
> prior
> > > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > > > too.
> > > > > > > > > In
> > > > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > > > regriding.
> > > > > > > > > The input and the target grid should be the same,
which is
> > > > > > > > > the output of point2grid.
> > > > > > > > >
> > > > > > > > > Here is an example (assuming the point data file is
> > > > > > > > > generated by pb2nc).
> > > > > > > > >
> > > > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > > > >       -v 1
> > > > > > > > >
> > > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > > > >       -method GAUSSIAN -v 1
> > > > > > > > >
> > > > > > > > > Note: override -gaussian_dx (81.271) &
-gaussian_radius
> > > > > > > > > (120.0) if necessary
> > > > > > > > >
> > > > > > > > > Cheers,
> > > > > > > > > Howard
> > > > > > > > >
> > > > > > > > > On Mon Jun 29 14:13:44 2020,
john.w.raby2.civ at mail.mil
> > wrote:
> > > > > > > > > > Howard -
> > > > > > > > > >
> > > > > > > > > > Thanks for your quick response with details in
answer to
> > > > > > > > > > my question.
> > > > > > > > > > Right now my interest is with non-probability
fields, so
> > > > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > > > smoothing for -field argument is interesting.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> > > > > > > > > > V9.0 Point2Grid
> > > > > > > > > > (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please
> > > > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > > prior to copying and pasting the address to a Web
> browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation
for the
> > > > > > > > > > raw data
> > > > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for
the
> > > > > > > > > > probability variable. (the existence of valid
values if
> > > > > > > > > > the probability threshold is not given)
> > > > > > > > > >
> > > > > > > > > > + "-method GAUSSIAN" is processed as the "-method
> UW_MEAN"
> > > > > > > > > > (which
> > > > > > > > > > is
> > > > > > > > > > a
> > > > > > > > > > default) for the layer 1 and "-method GAUSSIAN"
for the
> > > > > > > > > > layer 2.
> > > > > > > > > > + "-method MAXGAUSS" is processed as the "-method
MAX"
> for
> > > > > > > > > > the
> > > > > > > > > > layer
> > > > > > > > > > 1
> > > > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > > > + Override the interpolation method of layer 1
with
> > > > > > > > > > + "-method
> > > > > > > > > > <METHOD>"
> > > > > > > > > >
> > > > > > > > > > 1. The Gaussian smoothing at point2grid is
activated and
> > > > > > > > > > applied into the probability variable only if
-method
> > > > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the
command line
> > > > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > > > argument. If you want, You can do this by using
> > > > > > > > > > regrid_data_plane (passing the output of
point2grid as
> the
> > > > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > > > >
> > > > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored
if
> > > > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given.
The
> > > > > > > > > > default values are applied if not given.
> > > > > > > > > >
> > > > > > > > > > 3 -prob_cat_thresh applied regardless -method
GAUSSIAN or
> > > > > > > > > > -method MAXGAUSS argument.
> > > > > > > > > >
> > > > > > > > > > 4 Without -prob_cat_thresh argument, it will be
the
> status
> > > > > > > > > > of data existence. With -prob_cat_thresh argument,
it
> will
> > > > > > > > > > be meeting the condition or not. So the values of
the
> > > > > > > > > > probability are 0 or 1.
> > > > > > > > > >
> > > > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN
smoothing is
> > > > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > > > >
> > > > > > > > > > Cheers,
> > > > > > > > > > Howard
> > > > > > > > > >
> > > > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > I'm reassigning this ticket to Howard Soh who
worked on
> > > > > > > > > > > the development of the point2grid tool. He
should be
> > > > > > > > > > > able to better answer these questions.
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W
USA CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > >
> <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > > > lay.html?i
> > > > > > > > > > > > d=95747 >
> > > > > > > > > > > >
> > > > > > > > > > > > John -
> > > > > > > > > > > >
> > > > > > > > > > > > I ran it to see the usage and there are
details in
> the
> > > > > > > > > > > > usage dialog and also in the XML file you
pointed me
> > > > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > > > There is an additional interpolation method
called
> > > > > > > > > > > > GAUSSIAN which applies only when using the
> probability
> > > > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > > > So, it
> > > > > > > > > > > > looks like the two optional smoothing
arguments
> > > > > > > > > > > > -gaussian_dx and -gaussian_radius are only
used when
> > > > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Subject: Re: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > All active links contained in this email were
> disabled.
> > > > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > > > the authenticity of all links contained within
the
> > > > > > > > > > > > message prior to copying and pasting the
address to a
> > > > > > > > > > > > Web browser.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > ----
> > > > > > > > > > > >
> > > > > > > > > > > > John,
> > > > > > > > > > > >
> > > > > > > > > > > > While the "-field" command line argument
indicates
> > > > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > > > >
> > > > > > > > > > > > If you specify "-field" more that once you'll
get
> this
> > > > > > > > > > > > error
> > > > > > > > > > > > message:
> > > > > > > > > > > > ERROR  :
> > > > > > > > > > > > ERROR  : process_point_file() -> Support only
1
> > variable.
> > > > > > > > > > > > "2"
> > > > > > > > > > > > variables
> > > > > > > > > > > > were requested.
> > > > > > > > > > > > ERROR  :
> > > > > > > > > > > >
> > > > > > > > > > > > We do plan to support the processing of
multiple
> > > > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > > > >
> > > > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > > > it's very simple. Keep track of the point
> observations
> > > > > > > > > > > > falling inside each of the target grid boxes.
Write
> to
> > > > > > > > > > > > the output the min, max, unweighted mean, or
median
> of
> > > > > > > > > > > > those values based on the method chosen. If
you only
> > > > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > > > all those methods will result in the same
output
> > > > > > > > > > > > value.
> > > > > > > > > > > > Whereas if you have many, as in the case of
GOES
> input
> > > > > > > > > > > > files, then the choice of the derivation
method would
> > > > > > > > > > > > matter.
> > > > > > > > > > > >
> > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks,
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W
USA CIV
> > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > > > ?id=
> > > > > > > > > > > > 95747 >
> > > > > > > > > > > > >
> > > > > > > > > > > > > Another question I have is how is -method
used in
> > > > > > > > > > > > > this tool?
> > > > > > > > > > > > > In the
> > > > > > > > > > > > User's
> > > > > > > > > > > > > Guide, -method provides options for
regridding from
> > > > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > > > regrid_data_plane. In the case of
Point2Grid, the
> > > > > > > > > > > > > process involved is not regridding in the
way that
> I
> > > > > > > > > > > > > understand it.
> > > > > > > > > > > > > If I
> > > > > > > > > > > > > understand it correctly, it takes point
> observations
> > > > > > > > > > > > > and distributes them on to a user provided
grid
> > > > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > > > step I
> > > > > > > > > > > > > think is the smoothing of the field to even
out
> > > > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > > > to some function which appear to be a
Gaussian
> > > > > > > > > > > > > function. I guess I also need to understand
what
> > > > > > > > > > > > > happens when the tool is used without any of
the
> > > > > > > > > > > > > optional arguments.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > >
> > > > > > > > > > > > > R/
> > > > > > > > > > > > > John
> > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > V9.0
> > > > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > >
> > > > > > > > > > > > > John -
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks for your quick response which is
helpful.
> > > > > > > > > > > > > Appreciate
> > > > > > > > > > > > > your dedication so late into the evening!
Can I add
> > > > > > > > > > > > > new variables in the
> > > > > > > > > > > > config
> > > > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > > > interpolation methods would be appreciated,
but
> that
> > > > > > > > > > > > > can wait until Monday.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > > > >
> > > > > > > > > > > > > R/
> > > > > > > > > > > > > John
> > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu
> ]
> > > > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > Subject: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > >
> > > > > > > > > > > > > All active links contained in this email
were
> > disabled.
> > > > > > > > > > > > > Please verify
> > > > > > > > > > > > the
> > > > > > > > > > > > > identity of the sender, and confirm the
> authenticity
> > > > > > > > > > > > > of all links
> > > > > > > > > > > > contained
> > > > > > > > > > > > > within the message prior to copying and
pasting the
> > > > > > > > > > > > > address to a Web browser.
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > ----
> > > > > > > > > > > > >
> > > > > > > > > > > > > John,
> > > > > > > > > > > > >
> > > > > > > > > > > > > I see you have a question about
interpolation
> > > > > > > > > > > > > options in the point2grid tool. This XML
file
> > > > > > > > > > > > > contains the calls to the point2grid tool
that are
> > > > > > > > > > > > run
> > > > > > > > > > > > > for our unit tests:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > > > est/xml/
> > > > > > > > > > > > un
> > > > > > > > > > > > it
> > > > > > > > > > > > _point2grid.xml
> > > > > > > > > > > > >
> > > > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > > > line option, a few at the end do use "-
method MAX"
> > > > > > > > > > > > > to select the maximum value falling within
each
> grid
> > > > > > > > > > > > > box.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > > > et/src/t
> > > > > > > > > > > > oo
> > > > > > > > > > > > ls
> > > > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > > > I see that the supported options for -method
are
> > > > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Seems like we should add that info to the
usage
> > > > > > > > > > > > > statement!
> > > > > > > > > > > > > I
> > > > > > > > > > > > > wrote up
> > > > > > > > > > > > this
> > > > > > > > > > > > > GitHub issue to make sure we remember to do
so
> > > > > > > > > > > > > before the
> > > > > > > > > > > > > 9.1
> > > > > > > > > > > > > release:
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-
> https://github.com/NCAR/MET/issu
> > > > > > > > > > > > > es/1396
> > > > > > > > > > > > >
> > > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks,
> > > > > > > > > > > > > John
> > > > > > > > > > > > >
> > > > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W
USA
> CIV
> > > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747
was acted
> > > > > > > > > > > > > > upon.
> > > > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >       Status: new
> > > > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-Caution-Caution-
url:
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > > > splay.html
> > > > > > > > > > > > > ?id=95747
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > > > User's
> > > > > > > > > > > > > > Guide is Gaussian, but are other methods
> available?
> > > > > > > > > > > > > > If
> > > > > > > > > > > > > > so,
> > > > > > > > > > > > > > can you
> > > > > > > > > > > > > provide
> > > > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > > > argument?
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Stay safe.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > T/
> > > > > > > > > > > > > > John
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > > > Range, NM
> > > > > > > > > > > > > > 88002
> > > > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > > > Teleworking: Contact Admin Specialist at
(410)
> > > > > > > > > > > > > > 278-
> > > > > > > > > > > > > > 6974
> > > > > > > > > > > > > > FAX
> > > > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > > >
> > > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> >
> >
> >
> >
>
>
>
>

------------------------------------------------
Subject: RE: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0 Point2Grid (UNCLASSIFIED)
From: Raby, John W USA CIV
Time: Mon Jul 20 16:52:42 2020

Thanks for your input on that, John.

I'll look around on the HPC to see how they run python packages. Maybe
I can run one to produce a set of gridded obs and then  run MET Grid-
Stat so it looks for those obs.

R/
John

________________________________________
From: John Halley Gotway via RT [met_help at ucar.edu]
Sent: Monday, July 20, 2020 4:47 PM
To: Raby, John W CIV USARMY CCDC ARL (USA)
Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
Point2Grid (UNCLASSIFIED)

All active links contained in this email were disabled.  Please verify
the identity of the sender, and confirm the authenticity of all links
contained within the message prior to copying and pasting the address
to a Web browser.




----

I have not actually run Python embedding with MET from inside the
container, but there's no reason why it shouldn't work. Python
embedding is
enabled in the Dockerfile build of MET:
Caution-
https://github.com/NCAR/MET/blob/master_v9.0/scripts/docker/Dockerfile

However, I can see how it could get tricky. The MET image includes an
instance of python version 3.6... but it includes very few python
packages.
Let's say you want a python script which loads the pygrib module.
Everytime
you instantiate that image, you'd get a container without the pygrib
module. So you'd install that package everytime you run... which would
be
inefficient. It'd be better to just build a new image on top of the
MET one
with whatever python packages you want.

So not impossible, but perhaps tricky.

John

On Mon, Jul 20, 2020 at 3:48 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <Caution-url: Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> John -
>
> Thanks for your assessment of URMA. I agree. NCEP/NWS developed it
with
> the goal of providing an analysis of record for doing model
verification.
> This seems like it take on the form/significance of an
accreditation. I
> apprecaite your tip on the use of Python embedding. Will this work
with the
> way I'm running MET via a container? I'm not up on how to use Python
> embedding. If it is not integrated with MET maybe running MET in a
> container is not an issue.
>
> R/
> John
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, July 20, 2020 3:38 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> John,
>
> Yes, I agree that using URMA is a good idea. There are several
people
> working on that analysis, and generally, I'd trust their results
more than
> the output of the point2grid tool. This tool is still very new, and,
as
> you've illustrated, there are many other methods that could be added
to it.
> I'll leave it to the scientists and project managers to determine
our
> development priorities within our funding constraints.
>
> I want to point out though that Python embedding could be a very
nice
> option here. I would imagine that Python packages exist which
support many
> methods of putting point data onto a grid. When researching
potential ways
> of doing so, you could have MET execute a custom Python script and
read the
> resulting gridded data into grid_stat or mode.
>
> John
>
> On Mon, Jul 20, 2020 at 3:20 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-Caution-url: Caution-
> Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > Thanks for sharing your thoughts and knowledge on this. Believe it
or
> not,
> > this discussion is helpful from a research perspective. I've been
trying
> to
> > develop ways to generated gridded obs on ever-finer grids because
of Army
> > requirements. Having your sense of what works best in point2grid
is
> > helpful. It is valid concern about what to do in data sparse
regions
> where
> > you don't have obs. I've been verifying the URMA gridded obs on a
2.5 km
> > grid over CONUS and they are grappling with the same issues.
Gridded obs
> > verify the best (against point obs) if there is a dense network of
point
> > obs which are used as input. For model verification, the real
challenge
> is
> > in areas where there are no point obs which were used as input.
The
> gridded
> > obs in these point ob-sparse areas are as good as the method used
to
> > interpolate between sparse point obs.
> >
> > Was the intent of developing point2grid to provide gridded obs for
> > verification? What happens if you use Grid-Stat and use the
point2grid
> > produced gridded obs such as those I showed you for verification.
Will
> this
> > generate erroneous results in the areas where you don't have point
obs or
> > will it indicate where there were insufficient point obs and flag
it?
> Could
> > one sort out were there are valid error stats and limit the
verification
> to
> > just those areas where there is good coverage of point obs or will
the
> > presence of large areas with poor coverage result in contaminating
the
> > error statistics? Consider running point2grid for a first pass and
then
> > based on that, possibly dividing up the domain into subdomains
using
> > masking regions where the gridded analysis(obs) is judged to be of
good
> > quality?
> >
> > For the IDV gridded obs, I was able to specify the use of the
optimized
> > Barnes Objective Analysis technique and could specify the factors
used to
> > produce the analysis. The factors which can be changed are the
number of
> > passes (2-4), the search radius (length scale) and the gain which
> > determines the amount of reduction applied to the length scale for
the
> > second pass. I did some testing and obtained the best overall
analysis
> with
> > 4-passes, 10 grid square search radius and gain – 0.2 which is the
> minimum
> > gain suggested by the IDV User’s Guide. It also provides a way for
you to
> > specify the output grid spacing. While I was able to generate
output on a
> > 1km grid, the process is much more time consuming than running
> point2grid,
> > hence my interest in point2grid.
> >
> > Based on your thought that the issue with my point obs is their
> sparseness
> > and that the recommended way to workaround this is to upscale the
grid
> from
> > 1km to 10 km, I would have to conclude that it might be better to
use the
> > URMA obs on a 2.5 km grid and downscale them (regrid) to verify
our model
> > output on a 1km grid.  The objective analysis technique used by
RTMA/URMA
> > is more sophisticated in that they use a gridded background field
from
> > Rapid Refresh (RAP) model and the High Resolution Rapid Refresh
(HRRR)
> > model. Their output is blended with the North American Mesoscale
Model
> > (NAM) and downscaled as a first guess background field on a 2.5 km
grid.
> > The first guess field is then adjusted through a two-dimensional
> > variational data assimilation technique (2dvar) to analyze point
weather
> > observations from the national networks of point obs. That said, I
still
> > consider the capability to allow a user to generate a gridded
filed form
> > point obs to be interesting. I have access to !
> >  some experimental measurements which are not being input into the
> > national databases such as MADIS. If I could ingest those
measurements as
> > well as the points obs we get from MADIS into point2grid, then the
> quality
> > of the gridded obs output might exceed that from URMA for a
limited area
> > around the experimental measurements. It's worth some
experimentation.
> >
> > R/
> > John
> >
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Monday, July 20, 2020 2:13 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > I see that you're processing 136 point observations of
temperature. And
> > you're placing them onto a 204x204 grid... totalling 41,616 grid
points.
> >
> > Honestly, we haven't worked on the point2grid tool for many years
since
> > it's so problematic. There are just so many ways a person might
want to
> > interpolate point observations onto a grid, that it isn't
conceivable to
> > support them all. And frankly, none of them are very good. Using
136
> point
> > observations to populate values at 41,616 grid points just isn't
going to
> > produce meaningful output. I see that the IDV output appears to
have
> > structure, but it's all artifacts of the interpolation process.
And
> > whatever method IDV is using is not currently supported in the
point2grid
> > tool. The point2grid tool works best when you have at least one
point
> > observation falling into each grid cell, which is the case with
dense
> > satellite data or a dense network of point stations.
> >
> > With your existing data, I'd recommend using a much, much coarser
target
> > grid. You could try shifting from a 1-km grid to a 10-km grid, to
see if
> > the results make more sense.
> >
> > Thanks,
> > John
> >
> > On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-Caution-Caution-url: Caution-
> > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > > John -
> > >
> > > Welcome back. Yes, I'm still trying to understand the output
gridded
> file
> > > which places point obs on a grid.  I'm comparing what I'm seeing
from
> my
> > > testing of point2grid so far with an experiment I did last JAN
where I
> > > used
> > > IDV to produce a NetCDF file containing a set of gridded obs on
a 1km
> > > grid.
> > > Ref ticket # 93980 for your assist in making the NetCDF file
from IDV
> > > MET_compatible. The attached text file contains the point obs
which
> were
> > > used
> > > by IDV to produce the gridded obs. The result was a MET-
compatible file
> > > which
> > > I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> > > file. If
> > > you look at the plot of the IDV generated gridded obs in the
attached
> > PDF,
> > > it
> > > looks different from that which is generated by point2grid. I'm
trying
> to
> > > reconcile the two to gain further understanding.  See the
attached PDF
> of
> > > the
> > > point2grid output which I produced. The obs used for the
point2grid
> > > testing
> > > are not the same ones as those used for the IDV testing. The
NETCDF
> > output
> > > from point2grid is also attached. See list of files below.
> > >
> > > That said, I understand that Howard is working on
regrid_data_plane to
> > > enable
> > > the use of Gaussian smoothing (MAXGAUSS) on the output of
point2grid. I
> > > wonder
> > > if the Gaussian smoothing will achieve an end product which look
more
> > > similar
> > > to what I was expecting? Maybe the gridded obs I produced with
IDV
> isn't
> > > ever
> > > going to produce the same thing which point2grid was conceived
to
> produce
> > > due
> > > to the use of a different concept in performing an objective
analysis
> > > using
> > > interpolation.
> > >
> > > List of attachments:
> > > Case1_TMP_OBS_18_UTC - point obs text file
> > > 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> > > 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded
obs
> > > 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> > > 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
> > >
> > > R/
> > > John
> > >
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT <met_help at ucar.edu>
> > > Sent: Monday, July 20, 2020 10:56 AM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
<john.w.raby2.civ at mail.mil>
> > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> > > Point2Grid
> > > (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
verify
> > the
> > > identity of the sender, and confirm the authenticity of all
links
> > > contained
> > > within the message prior to copying and pasting the address to a
Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > Hi John,
> > >
> > > This is John Halley Gotway. I'm covering the met-help desk for
the next
> > > few
> > > days and wanted to check in on this issue. Are you still having
trouble
> > > understanding the output of point2grid, or does it make sense?
> > >
> > > Thanks,
> > > John
> > >
> > > On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-Caution-url:
> > > > Caution-Caution-
> > Caution-Caution-
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > >
> > > > The plot_data_plane can not show the point observation data.
It shows
> > > > the plot of the grid data.
> > > > In this case, the coverage of the gridded data is 0.23% (97
out of
> > > 41616).
> > > > It's the plot of the grid cells (1km by 1km) even though it
looks
> like
> > > > the point data. It caused by few converage (0.23%).
> > > >
> > > > The meaning of following log mesages (from point2grid):
> > > >
> > > > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180
(99%)
> DEBUG
> > > > 2: process_point_file() ->  var_count=180, grid: 97 out of
41616
> > > >
> > > > - The input data has 180 TMP data
> > > > - One TMP data does not belong to the target grid
> > > > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 /
41616 *
> 100)
> > > >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more
TMP data
> > > >
> > > > The plot has valid data for 97 cells out of 41616 (which is
204 by
> > > > 204) grid cells)
> > > >
> > > >
> > > > Two issues with regrid_data_plane (this specific case because
of
> 0.23%
> > > > coverage).
> > > >
> > > > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh)
OR 0.008
> > > > (<=
> > > > 1 / 121).
> > > >
> > > > The default setting of "-vld_thresh" is 0.5.
> > > > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The
matrix
> for
> > > > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a
valid
> > > > regrdded value for the taeget cell, it requires 61 (from 121 /
2)
> > > > neighbers with the valid TMP data. Only 97 cells have the
valid data
> > and
> > > > they are spreaded.
> > > > None of them satisfies the number of the valid neighber cells.
That's
> > > > why the output of regrid_data_plae was all missing values.
> > > >
> > > > 2. Bug on computing the Gaussian value.
> > > >
> > > > I will fix this.
> > > >
> > > > I attached the updated NetCDF file (output of
regrid_data_plane) and
> > > > the plot.
> > > >
> > > > Cheers,
> > > > Howard
> > > >
> > > > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil wrote:
> > > > > Thanks for working to improve the documentation. So you can
only
> use
> > > > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in
point2grid
> > > > > unless you are working with a probability variable in which
case
> you
> > > > > can use MAXGAUSS. The default of these is UW_MEAN. The
output I
> > > > > generated in my first test used that method since I didn't
specify
> > > > > the -method. In this case, the plot I generated using
> > > > > plot_data_plane shows a map with data points at random
locations
> > > > > which correspond to the locations of the original point obs
from
> the
> > > > > output of ascii2nc. My expectation was that I would see the
data
> > > > > points with TMP values in a regularly spaced grid with 1km
spacing.
> > > > > In theory, this NetCDF file could be used in Grid-Stat to
verify a
> > > > > TMP fcst on the same 1km grid. Why doesn't the plot show a
gridded
> > TMP
> > > > > field?
> > > > >
> > > > > R/
> > > > > John
> > > > >
> > > > >
> > > > > ________________________________________
> > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET V9.0
> > > > > Point2Grid
> > > > > (UNCLASSIFIED)
> > > > >
> > > > > All active links contained in this email were disabled.
Please
> > > > > verify the identity of the sender, and confirm the
authenticity of
> > > > > all links contained within the message prior to copying and
pasting
> > > > > the address to a Web browser.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > ----
> > > > >
> > > > > Yes, you are right. It's not the output you are expecting.
We will
> > > > > improve the documentation on this (Caution-
> > > > > Caution-Caution-Caution-Caution-
https://github.com/NCAR/MET/issues/1413).
> > > > >
> > > > > Unlike regrid_data_plane, the MAXGAUSE method with
point2grid works
> > > > > differently: MAX interpolation for the raw data (TMP) and
GAUSSIAN
> > > > > smoothing into the probability variable (TMP_prob_NA or
> TMP_prob_XX).
> > > > > MAXGAUSS is the same as the MAX interpolation for the TMP
variable
> > > > > for the point2grid. So the difference is between MAX VS.
UW_MEAN
> > > > > interpolation methods, and GAUSIAN smoothing was not
involved.
> > > > >
> > > > > The right way for the Gaussian smoothing to the raw data
will be:
> > > > > 1) regridding by using point2grid with one of interpolation
methods
> > > > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > > > 2) regridding by using regrid_data_plane with MAXGAUSS
method and
> > > > > the same setting as the input and the target grid.
> > > > >
> > > > > The output of regrid_data_plane (MAXGAUSS) is not right. I'm
> working
> > > > > on this.
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > >
> > > > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > Howard -
> > > > > >
> > > > > > Today I successfully ran point2grid but used -method
MAXGAUSS and
> > > > > > it ran without error and produced output. I'm uploading
the file
> > > > > > list, output, commands used, log file, and plots to the
DoD SAFE
> > > > > > so you can take a look. I am also uploading the results of
the
> > > > > > previous run where you suggested that I use "11" vice
"TMP" which
> > > > > > fixed the problem and ran successfully.
> > > > > >
> > > > > > The plots are not what I was expecting somehow. The
plotted data
> > > > > > look like just points where the original obs were located
rather
> > > > > > than a regularly spaced grid of TMP values. What are your
> thoughts
> > > on
> > > > > > this?
> > > > > >
> > > > > > Thanks
> > > > > > ________________________________________
> > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> of
> > > > > > all links contained within the message prior to copying
and
> > > > > > pasting the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Something is not right. I will take a look
> > > > > >
> > > > > > DEBUG 2: Range of input data (name="TMP"; level="(*,*)";)
is
> > > > > > 246.351 to 284.817.
> > > > > > DEBUG 2: Range of regridded data (name="TMP";
level="(*,*)";) is
> > > > > > -9999
> > > > > > to -9999.
> > > > > >
> > > > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > > > output is all missing value (-9999).
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > > The input file has
> > > > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > > Howard -
> > > > > > >
> > > > > > > The next run with MAXGAUSS was successful, although I
have to
> > > > > > > plot it to see what it looks like. I've attached the log
file.
> > > > > > > Does it indicate any issues with how it ran?
> > > > > > >
> > > > > > > R/
> > > > > > > John
> > > > > > >
> > > > > > > -----Original Message-----
> > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
Please
> > > > > > > verify the identity of the sender, and confirm the
authenticity
> > > > > > > of all links contained within the message prior to
copying and
> > > > > > > pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed to
> > > > > > > MAXGAUSS and GAUSSIAN method is only used for the
> interpolation.
> > > > > > > Please set "-method MAXGAUSE". The from_grid and to_grid
are
> the
> > > > > > > same, there is no difference with additional MAX
interpolation.
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > >
> > > > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > I tried running MET V9.0 regrid_data_plane following
your
> > > > > > > > example below, but I'm getting and error:
> > > > > > > > ERROR  :
> > > > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not
valid for
> > > > > > > > regridding, only interpolating.
> > > > > > > >
> > > > > > > > The following is the command I used to run it:
> > > > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > V2016.nc
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > V2016.nc
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > > > P_1200UTC_11NOV2016_gaussian.nc
> > > > > > > >   -field 'name="TMP"; level="(*,*)";' -method GAUSSIAN
-log
> > > > > > > >
> > > > > > > >
> > > >
>
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > > > ss_test
> > > > > > > > -v 9
> > > > > > > >
> > > > > > > > Any suggestions on how to perform Gaussian smoothing
on the
> > > > > > > > output TMP field from point2grid?
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > >
> > > > > > > > -----Original Message-----
> > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET V9.0
> > > > > > > > Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > Please verify the identity of the sender, and confirm
the
> > > > > > > > authenticity of all links contained within the message
prior
> > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > Sorry for the late response.
> > > > > > > >
> > > > > > > > The input and to_grid must be the same for Gaussian
smoothing
> > > > > > > > and regrid_data_plane shows the usage message if the
to_grid
> > > > > > > > is missing.
> > > > > > > > There are
> > > > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > > > >
> > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP";
level="(*,*)";' \
> > > > > > > > -method GAUSSIAN -v 1
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > On Tue Jun 30 06:58:02 2020, john.w.raby2.civ at mail.mil
> wrote:
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > > Howard -
> > > > > > > > >
> > > > > > > > > Thanks for tailoring the command for my particular
> interests.
> > > > > > > > > Just
> > > > > > > > > curious about your regrid_data_plane command. You
provided
> > > > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as
the
> > > > > > > > > input_filename to regrid_data_plane and then you
provide
> the
> > > > > > > > > name of the file to be output from regrid_data_plane
> > > > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you didn't
> provide
> > > > > > > > > the to_grid. Is this because it knows that the grids
of the
> > > > > > > > > input and output files are the same and thus ignores
the
> > > > > > > > > missing to_grid argument?
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > -----Original Message-----
> > > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> V9.0
> > > > > > > > > Point2Grid
> > > > > > > > > (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify the identity of the sender, and
confirm the
> > > > > > > > > authenticity of all links contained within the
message
> prior
> > > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > The Gaussian smoothing is available with
regrid_data_plane,
> > > > > > > > > too.
> > > > > > > > > In
> > > > > > > > > this case, we are interested in Gaussian smoothing
without
> > > > > > > > > regriding.
> > > > > > > > > The input and the target grid should be the same,
which is
> > > > > > > > > the output of point2grid.
> > > > > > > > >
> > > > > > > > > Here is an example (assuming the point data file is
> > > > > > > > > generated by pb2nc).
> > > > > > > > >
> > > > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > > > >       -v 1
> > > > > > > > >
> > > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > > > >       -method GAUSSIAN -v 1
> > > > > > > > >
> > > > > > > > > Note: override -gaussian_dx (81.271) &
-gaussian_radius
> > > > > > > > > (120.0) if necessary
> > > > > > > > >
> > > > > > > > > Cheers,
> > > > > > > > > Howard
> > > > > > > > >
> > > > > > > > > On Mon Jun 29 14:13:44 2020,
john.w.raby2.civ at mail.mil
> > wrote:
> > > > > > > > > > Howard -
> > > > > > > > > >
> > > > > > > > > > Thanks for your quick response with details in
answer to
> > > > > > > > > > my question.
> > > > > > > > > > Right now my interest is with non-probability
fields, so
> > > > > > > > > > your point in 1. about the possibility of using
Gaussian
> > > > > > > > > > smoothing for -field argument is interesting.
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > > ________________________________________
> > > > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> > > > > > > > > > V9.0 Point2Grid
> > > > > > > > > > (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please
> > > > > > > > > > verify the identity of the sender, and confirm the
> > > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > > prior to copying and pasting the address to a Web
> browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > Two layers for the interpolation with point2grid:
> > > > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation
for the
> > > > > > > > > > raw data
> > > > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for
the
> > > > > > > > > > probability variable. (the existence of valid
values if
> > > > > > > > > > the probability threshold is not given)
> > > > > > > > > >
> > > > > > > > > > + "-method GAUSSIAN" is processed as the "-method
> UW_MEAN"
> > > > > > > > > > (which
> > > > > > > > > > is
> > > > > > > > > > a
> > > > > > > > > > default) for the layer 1 and "-method GAUSSIAN"
for the
> > > > > > > > > > layer 2.
> > > > > > > > > > + "-method MAXGAUSS" is processed as the "-method
MAX"
> for
> > > > > > > > > > the
> > > > > > > > > > layer
> > > > > > > > > > 1
> > > > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > > > + Override the interpolation method of layer 1
with
> > > > > > > > > > + "-method
> > > > > > > > > > <METHOD>"
> > > > > > > > > >
> > > > > > > > > > 1. The Gaussian smoothing at point2grid is
activated and
> > > > > > > > > > applied into the probability variable only if
-method
> > > > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the
command line
> > > > > > > > > > argument. It's not applied into the variable at "-
field"
> > > > > > > > > > argument. If you want, You can do this by using
> > > > > > > > > > regrid_data_plane (passing the output of
point2grid as
> the
> > > > > > > > > > input of regrid_data_plane with GAUSSIAN method).
> > > > > > > > > >
> > > > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored
if
> > > > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT given.
The
> > > > > > > > > > default values are applied if not given.
> > > > > > > > > >
> > > > > > > > > > 3 -prob_cat_thresh applied regardless -method
GAUSSIAN or
> > > > > > > > > > -method MAXGAUSS argument.
> > > > > > > > > >
> > > > > > > > > > 4 Without -prob_cat_thresh argument, it will be
the
> status
> > > > > > > > > > of data existence. With -prob_cat_thresh argument,
it
> will
> > > > > > > > > > be meeting the condition or not. So the values of
the
> > > > > > > > > > probability are 0 or 1.
> > > > > > > > > >
> > > > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN
smoothing is
> > > > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > > > >
> > > > > > > > > > Cheers,
> > > > > > > > > > Howard
> > > > > > > > > >
> > > > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > > > John,
> > > > > > > > > > >
> > > > > > > > > > > I'm reassigning this ticket to Howard Soh who
worked on
> > > > > > > > > > > the development of the point2grid tool. He
should be
> > > > > > > > > > > able to better answer these questions.
> > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > Thanks,
> > > > > > > > > > > John
> > > > > > > > > > >
> > > > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W
USA CIV
> > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > >
> <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
Caution-Caution-url:
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-Caution-
> > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > > > lay.html?i
> > > > > > > > > > > > d=95747 >
> > > > > > > > > > > >
> > > > > > > > > > > > John -
> > > > > > > > > > > >
> > > > > > > > > > > > I ran it to see the usage and there are
details in
> the
> > > > > > > > > > > > usage dialog and also in the XML file you
pointed me
> > > > > > > > > > > > to which helped to answer my own question (I
think).
> > > > > > > > > > > > There is an additional interpolation method
called
> > > > > > > > > > > > GAUSSIAN which applies only when using the
> probability
> > > > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > > > So, it
> > > > > > > > > > > > looks like the two optional smoothing
arguments
> > > > > > > > > > > > -gaussian_dx and -gaussian_radius are only
used when
> > > > > > > > > > > > the method is GAUSSIAN. IS this correct? This
means
> > > > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > > > >
> > > > > > > > > > > > R/
> > > > > > > > > > > > John
> > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu]
> > > > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > Subject: Re: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > >
> > > > > > > > > > > > All active links contained in this email were
> disabled.
> > > > > > > > > > > > Please verify the identity of the sender, and
confirm
> > > > > > > > > > > > the authenticity of all links contained within
the
> > > > > > > > > > > > message prior to copying and pasting the
address to a
> > > > > > > > > > > > Web browser.
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > ----
> > > > > > > > > > > >
> > > > > > > > > > > > John,
> > > > > > > > > > > >
> > > > > > > > > > > > While the "-field" command line argument
indicates
> > > > > > > > > > > > that it can be used one or more times, in this
first
> > > > > > > > > > > > version of the tool, it can only be used once.
> > > > > > > > > > > >
> > > > > > > > > > > > If you specify "-field" more that once you'll
get
> this
> > > > > > > > > > > > error
> > > > > > > > > > > > message:
> > > > > > > > > > > > ERROR  :
> > > > > > > > > > > > ERROR  : process_point_file() -> Support only
1
> > variable.
> > > > > > > > > > > > "2"
> > > > > > > > > > > > variables
> > > > > > > > > > > > were requested.
> > > > > > > > > > > > ERROR  :
> > > > > > > > > > > >
> > > > > > > > > > > > We do plan to support the processing of
multiple
> > > > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > > > >
> > > > > > > > > > > > As for the science behind the interpolation
methods,
> > > > > > > > > > > > it's very simple. Keep track of the point
> observations
> > > > > > > > > > > > falling inside each of the target grid boxes.
Write
> to
> > > > > > > > > > > > the output the min, max, unweighted mean, or
median
> of
> > > > > > > > > > > > those values based on the method chosen. If
you only
> > > > > > > > > > > > have 1 point obs falling inside each grid box,
then
> > > > > > > > > > > > all those methods will result in the same
output
> > > > > > > > > > > > value.
> > > > > > > > > > > > Whereas if you have many, as in the case of
GOES
> input
> > > > > > > > > > > > files, then the choice of the derivation
method would
> > > > > > > > > > > > matter.
> > > > > > > > > > > >
> > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks,
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W
USA CIV
> > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > > > ?id=
> > > > > > > > > > > > 95747 >
> > > > > > > > > > > > >
> > > > > > > > > > > > > Another question I have is how is -method
used in
> > > > > > > > > > > > > this tool?
> > > > > > > > > > > > > In the
> > > > > > > > > > > > User's
> > > > > > > > > > > > > Guide, -method provides options for
regridding from
> > > > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > > > regrid_data_plane. In the case of
Point2Grid, the
> > > > > > > > > > > > > process involved is not regridding in the
way that
> I
> > > > > > > > > > > > > understand it.
> > > > > > > > > > > > > If I
> > > > > > > > > > > > > understand it correctly, it takes point
> observations
> > > > > > > > > > > > > and distributes them on to a user provided
grid
> > > > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > > > step I
> > > > > > > > > > > > > think is the smoothing of the field to even
out
> > > > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > > > etc resulting in the fitting of the observed
values
> > > > > > > > > > > > > to some function which appear to be a
Gaussian
> > > > > > > > > > > > > function. I guess I also need to understand
what
> > > > > > > > > > > > > happens when the tool is used without any of
the
> > > > > > > > > > > > > optional arguments.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > >
> > > > > > > > > > > > > R/
> > > > > > > > > > > > > John
> > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > > > Subject: RE: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > V9.0
> > > > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > >
> > > > > > > > > > > > > John -
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks for your quick response which is
helpful.
> > > > > > > > > > > > > Appreciate
> > > > > > > > > > > > > your dedication so late into the evening!
Can I add
> > > > > > > > > > > > > new variables in the
> > > > > > > > > > > > config
> > > > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > > > interpolation methods would be appreciated,
but
> that
> > > > > > > > > > > > > can wait until Monday.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > > > >
> > > > > > > > > > > > > R/
> > > > > > > > > > > > > John
> > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu
> ]
> > > > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > Subject: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > >
> > > > > > > > > > > > > All active links contained in this email
were
> > disabled.
> > > > > > > > > > > > > Please verify
> > > > > > > > > > > > the
> > > > > > > > > > > > > identity of the sender, and confirm the
> authenticity
> > > > > > > > > > > > > of all links
> > > > > > > > > > > > contained
> > > > > > > > > > > > > within the message prior to copying and
pasting the
> > > > > > > > > > > > > address to a Web browser.
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > ----
> > > > > > > > > > > > >
> > > > > > > > > > > > > John,
> > > > > > > > > > > > >
> > > > > > > > > > > > > I see you have a question about
interpolation
> > > > > > > > > > > > > options in the point2grid tool. This XML
file
> > > > > > > > > > > > > contains the calls to the point2grid tool
that are
> > > > > > > > > > > > run
> > > > > > > > > > > > > for our unit tests:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-Caution-https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > > > est/xml/
> > > > > > > > > > > > un
> > > > > > > > > > > > it
> > > > > > > > > > > > _point2grid.xml
> > > > > > > > > > > > >
> > > > > > > > > > > > > While most of them DO NOT use the "-method"
command
> > > > > > > > > > > > > line option, a few at the end do use "-
method MAX"
> > > > > > > > > > > > > to select the maximum value falling within
each
> grid
> > > > > > > > > > > > > box.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-
> > > > > > > > > > > > Caution-Caution-
> > Caution-Caution-https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > > > et/src/t
> > > > > > > > > > > > oo
> > > > > > > > > > > > ls
> > > > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > > > I see that the supported options for -method
are
> > > > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Seems like we should add that info to the
usage
> > > > > > > > > > > > > statement!
> > > > > > > > > > > > > I
> > > > > > > > > > > > > wrote up
> > > > > > > > > > > > this
> > > > > > > > > > > > > GitHub issue to make sure we remember to do
so
> > > > > > > > > > > > > before the
> > > > > > > > > > > > > 9.1
> > > > > > > > > > > > > release:
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-
> Caution-https://github.com/NCAR/MET/issu
> > > > > > > > > > > > > es/1396
> > > > > > > > > > > > >
> > > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks,
> > > > > > > > > > > > > John
> > > > > > > > > > > > >
> > > > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John W
USA
> CIV
> > > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747
was acted
> > > > > > > > > > > > > > upon.
> > > > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
(UNCLASSIFIED)
> > > > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >       Status: new
> > > > > > > > > > > > > >  Ticket <Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-Caution-Caution-
Caution-url:
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-
> > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > > > splay.html
> > > > > > > > > > > > > ?id=95747
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > > > User's
> > > > > > > > > > > > > > Guide is Gaussian, but are other methods
> available?
> > > > > > > > > > > > > > If
> > > > > > > > > > > > > > so,
> > > > > > > > > > > > > > can you
> > > > > > > > > > > > > provide
> > > > > > > > > > > > > > the names and how to specify them as an
optional
> > > > > > > > > > > > > > argument?
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Stay safe.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > T/
> > > > > > > > > > > > > > John
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > > > U.S. Army Research Laboratory White Sands
Missile
> > > > > > > > > > > > > > Range, NM
> > > > > > > > > > > > > > 88002
> > > > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > > > Teleworking: Contact Admin Specialist at
(410)
> > > > > > > > > > > > > > 278-
> > > > > > > > > > > > > > 6974
> > > > > > > > > > > > > > FAX
> > > > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > > >
> > > >
> > >
> > > CLASSIFICATION: UNCLASSIFIED
> > >
> > >
> >
> >
> >
> >
>
>
>
>



------------------------------------------------
Subject: MET V9.0 Point2Grid (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jul 20 17:15:39 2020

Sounds good. I'll go ahead and resolve this issue. Just send us a new
message if more issues arise.

Thanks,
John

On Mon, Jul 20, 2020 at 4:53 PM Raby, John W USA CIV via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
>
> Thanks for your input on that, John.
>
> I'll look around on the HPC to see how they run python packages.
Maybe I
> can run one to produce a set of gridded obs and then  run MET Grid-
Stat so
> it looks for those obs.
>
> R/
> John
>
> ________________________________________
> From: John Halley Gotway via RT [met_help at ucar.edu]
> Sent: Monday, July 20, 2020 4:47 PM
> To: Raby, John W CIV USARMY CCDC ARL (USA)
> Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET V9.0
> Point2Grid (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please
verify the
> identity of the sender, and confirm the authenticity of all links
contained
> within the message prior to copying and pasting the address to a Web
> browser.
>
>
>
>
> ----
>
> I have not actually run Python embedding with MET from inside the
> container, but there's no reason why it shouldn't work. Python
embedding is
> enabled in the Dockerfile build of MET:
> Caution-
>
https://github.com/NCAR/MET/blob/master_v9.0/scripts/docker/Dockerfile
>
> However, I can see how it could get tricky. The MET image includes
an
> instance of python version 3.6... but it includes very few python
packages.
> Let's say you want a python script which loads the pygrib module.
Everytime
> you instantiate that image, you'd get a container without the pygrib
> module. So you'd install that package everytime you run... which
would be
> inefficient. It'd be better to just build a new image on top of the
MET one
> with whatever python packages you want.
>
> So not impossible, but perhaps tricky.
>
> John
>
> On Mon, Jul 20, 2020 at 3:48 PM Raby, John W USA CIV via RT <
> met_help at ucar.edu> wrote:
>
> >
> > <Caution-url: Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> >
> > John -
> >
> > Thanks for your assessment of URMA. I agree. NCEP/NWS developed it
with
> > the goal of providing an analysis of record for doing model
verification.
> > This seems like it take on the form/significance of an
accreditation. I
> > apprecaite your tip on the use of Python embedding. Will this work
with
> the
> > way I'm running MET via a container? I'm not up on how to use
Python
> > embedding. If it is not integrated with MET maybe running MET in a
> > container is not an issue.
> >
> > R/
> > John
> > ________________________________________
> > From: John Halley Gotway via RT [met_help at ucar.edu]
> > Sent: Monday, July 20, 2020 3:38 PM
> > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > Point2Grid (UNCLASSIFIED)
> >
> > All active links contained in this email were disabled.  Please
verify
> the
> > identity of the sender, and confirm the authenticity of all links
> contained
> > within the message prior to copying and pasting the address to a
Web
> > browser.
> >
> >
> >
> >
> > ----
> >
> > John,
> >
> > Yes, I agree that using URMA is a good idea. There are several
people
> > working on that analysis, and generally, I'd trust their results
more
> than
> > the output of the point2grid tool. This tool is still very new,
and, as
> > you've illustrated, there are many other methods that could be
added to
> it.
> > I'll leave it to the scientists and project managers to determine
our
> > development priorities within our funding constraints.
> >
> > I want to point out though that Python embedding could be a very
nice
> > option here. I would imagine that Python packages exist which
support
> many
> > methods of putting point data onto a grid. When researching
potential
> ways
> > of doing so, you could have MET execute a custom Python script and
read
> the
> > resulting gridded data into grid_stat or mode.
> >
> > John
> >
> > On Mon, Jul 20, 2020 at 3:20 PM Raby, John W USA CIV via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <Caution-Caution-url: Caution-
> > Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > >
> > > Thanks for sharing your thoughts and knowledge on this. Believe
it or
> > not,
> > > this discussion is helpful from a research perspective. I've
been
> trying
> > to
> > > develop ways to generated gridded obs on ever-finer grids
because of
> Army
> > > requirements. Having your sense of what works best in point2grid
is
> > > helpful. It is valid concern about what to do in data sparse
regions
> > where
> > > you don't have obs. I've been verifying the URMA gridded obs on
a 2.5
> km
> > > grid over CONUS and they are grappling with the same issues.
Gridded
> obs
> > > verify the best (against point obs) if there is a dense network
of
> point
> > > obs which are used as input. For model verification, the real
challenge
> > is
> > > in areas where there are no point obs which were used as input.
The
> > gridded
> > > obs in these point ob-sparse areas are as good as the method
used to
> > > interpolate between sparse point obs.
> > >
> > > Was the intent of developing point2grid to provide gridded obs
for
> > > verification? What happens if you use Grid-Stat and use the
point2grid
> > > produced gridded obs such as those I showed you for
verification. Will
> > this
> > > generate erroneous results in the areas where you don't have
point obs
> or
> > > will it indicate where there were insufficient point obs and
flag it?
> > Could
> > > one sort out were there are valid error stats and limit the
> verification
> > to
> > > just those areas where there is good coverage of point obs or
will the
> > > presence of large areas with poor coverage result in
contaminating the
> > > error statistics? Consider running point2grid for a first pass
and then
> > > based on that, possibly dividing up the domain into subdomains
using
> > > masking regions where the gridded analysis(obs) is judged to be
of good
> > > quality?
> > >
> > > For the IDV gridded obs, I was able to specify the use of the
optimized
> > > Barnes Objective Analysis technique and could specify the
factors used
> to
> > > produce the analysis. The factors which can be changed are the
number
> of
> > > passes (2-4), the search radius (length scale) and the gain
which
> > > determines the amount of reduction applied to the length scale
for the
> > > second pass. I did some testing and obtained the best overall
analysis
> > with
> > > 4-passes, 10 grid square search radius and gain – 0.2 which is
the
> > minimum
> > > gain suggested by the IDV User’s Guide. It also provides a way
for you
> to
> > > specify the output grid spacing. While I was able to generate
output
> on a
> > > 1km grid, the process is much more time consuming than running
> > point2grid,
> > > hence my interest in point2grid.
> > >
> > > Based on your thought that the issue with my point obs is their
> > sparseness
> > > and that the recommended way to workaround this is to upscale
the grid
> > from
> > > 1km to 10 km, I would have to conclude that it might be better
to use
> the
> > > URMA obs on a 2.5 km grid and downscale them (regrid) to verify
our
> model
> > > output on a 1km grid.  The objective analysis technique used by
> RTMA/URMA
> > > is more sophisticated in that they use a gridded background
field from
> > > Rapid Refresh (RAP) model and the High Resolution Rapid Refresh
(HRRR)
> > > model. Their output is blended with the North American Mesoscale
Model
> > > (NAM) and downscaled as a first guess background field on a 2.5
km
> grid.
> > > The first guess field is then adjusted through a two-dimensional
> > > variational data assimilation technique (2dvar) to analyze point
> weather
> > > observations from the national networks of point obs. That said,
I
> still
> > > consider the capability to allow a user to generate a gridded
filed
> form
> > > point obs to be interesting. I have access to !
> > >  some experimental measurements which are not being input into
the
> > > national databases such as MADIS. If I could ingest those
measurements
> as
> > > well as the points obs we get from MADIS into point2grid, then
the
> > quality
> > > of the gridded obs output might exceed that from URMA for a
limited
> area
> > > around the experimental measurements. It's worth some
experimentation.
> > >
> > > R/
> > > John
> > >
> > > ________________________________________
> > > From: John Halley Gotway via RT [met_help at ucar.edu]
> > > Sent: Monday, July 20, 2020 2:13 PM
> > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > Subject: Re: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > Point2Grid (UNCLASSIFIED)
> > >
> > > All active links contained in this email were disabled.  Please
verify
> > the
> > > identity of the sender, and confirm the authenticity of all
links
> > contained
> > > within the message prior to copying and pasting the address to a
Web
> > > browser.
> > >
> > >
> > >
> > >
> > > ----
> > >
> > > John,
> > >
> > > I see that you're processing 136 point observations of
temperature. And
> > > you're placing them onto a 204x204 grid... totalling 41,616 grid
> points.
> > >
> > > Honestly, we haven't worked on the point2grid tool for many
years since
> > > it's so problematic. There are just so many ways a person might
want to
> > > interpolate point observations onto a grid, that it isn't
conceivable
> to
> > > support them all. And frankly, none of them are very good. Using
136
> > point
> > > observations to populate values at 41,616 grid points just isn't
going
> to
> > > produce meaningful output. I see that the IDV output appears to
have
> > > structure, but it's all artifacts of the interpolation process.
And
> > > whatever method IDV is using is not currently supported in the
> point2grid
> > > tool. The point2grid tool works best when you have at least one
point
> > > observation falling into each grid cell, which is the case with
dense
> > > satellite data or a dense network of point stations.
> > >
> > > With your existing data, I'd recommend using a much, much
coarser
> target
> > > grid. You could try shifting from a 1-km grid to a 10-km grid,
to see
> if
> > > the results make more sense.
> > >
> > > Thanks,
> > > John
> > >
> > > On Mon, Jul 20, 2020 at 1:22 PM Raby, John W USA CIV via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <Caution-Caution-Caution-url: Caution-
> > > Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > > John -
> > > >
> > > > Welcome back. Yes, I'm still trying to understand the output
gridded
> > file
> > > > which places point obs on a grid.  I'm comparing what I'm
seeing from
> > my
> > > > testing of point2grid so far with an experiment I did last JAN
where
> I
> > > > used
> > > > IDV to produce a NetCDF file containing a set of gridded obs
on a 1km
> > > > grid.
> > > > Ref ticket # 93980 for your assist in making the NetCDF file
from IDV
> > > > MET_compatible. The attached text file contains the point obs
which
> > were
> > > > used
> > > > by IDV to produce the gridded obs. The result was a MET-
compatible
> file
> > > > which
> > > > I was able to successfully ingest into Grid-Stat. See attached
NetCDF
> > > > file. If
> > > > you look at the plot of the IDV generated gridded obs in the
attached
> > > PDF,
> > > > it
> > > > looks different from that which is generated by point2grid.
I'm
> trying
> > to
> > > > reconcile the two to gain further understanding.  See the
attached
> PDF
> > of
> > > > the
> > > > point2grid output which I produced. The obs used for the
point2grid
> > > > testing
> > > > are not the same ones as those used for the IDV testing. The
NETCDF
> > > output
> > > > from point2grid is also attached. See list of files below.
> > > >
> > > > That said, I understand that Howard is working on
regrid_data_plane
> to
> > > > enable
> > > > the use of Gaussian smoothing (MAXGAUSS) on the output of
> point2grid. I
> > > > wonder
> > > > if the Gaussian smoothing will achieve an end product which
look more
> > > > similar
> > > > to what I was expecting? Maybe the gridded obs I produced with
IDV
> > isn't
> > > > ever
> > > > going to produce the same thing which point2grid was conceived
to
> > produce
> > > > due
> > > > to the use of a different concept in performing an objective
analysis
> > > > using
> > > > interpolation.
> > > >
> > > > List of attachments:
> > > > Case1_TMP_OBS_18_UTC - point obs text file
> > > > 1km_grid_obs_test_29JAN2020_NEW.nc - IDV-produced gridded obs
> > > > 1km_grid_obs_test_29JAN2020_NEW - plot of IDV-produced gridded
obs
> > > > 1km_TMP_1200UTC_11NOV2016.nc - point2grid output
> > > > 1km_TMP_1200UTC_11NOV2016 - plot of point2grid output
> > > >
> > > > R/
> > > > John
> > > >
> > > >
> > > > -----Original Message-----
> > > > From: John Halley Gotway via RT <met_help at ucar.edu>
> > > > Sent: Monday, July 20, 2020 10:56 AM
> > > > To: Raby, John W CIV USARMY CCDC ARL (USA) <
> john.w.raby2.civ at mail.mil>
> > > > Subject: [Non-DoD Source] Re: [rt.rap.ucar.edu #95747] MET
V9.0
> > > > Point2Grid
> > > > (UNCLASSIFIED)
> > > >
> > > > All active links contained in this email were disabled.
Please
> verify
> > > the
> > > > identity of the sender, and confirm the authenticity of all
links
> > > > contained
> > > > within the message prior to copying and pasting the address to
a Web
> > > > browser.
> > > >
> > > >
> > > >
> > > >
> > > > ----
> > > >
> > > > Hi John,
> > > >
> > > > This is John Halley Gotway. I'm covering the met-help desk for
the
> next
> > > > few
> > > > days and wanted to check in on this issue. Are you still
having
> trouble
> > > > understanding the output of point2grid, or does it make sense?
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Wed, Jul 15, 2020 at 2:55 PM Howard Soh via RT
<met_help at ucar.edu
> >
> > > > wrote:
> > > >
> > > > >
> > > > > <Caution-Caution-Caution-Caution-url:
> > > > > Caution-Caution-
> > > Caution-Caution-
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=95747 >
> > > > >
> > > > > The plot_data_plane can not show the point observation data.
It
> shows
> > > > > the plot of the grid data.
> > > > > In this case, the coverage of the gridded data is 0.23% (97
out of
> > > > 41616).
> > > > > It's the plot of the grid cells (1km by 1km) even though it
looks
> > like
> > > > > the point data. It caused by few converage (0.23%).
> > > > >
> > > > > The meaning of following log mesages (from point2grid):
> > > > >
> > > > > DEBUG 3: get_grid_mapping()  count in grid: 179 out of 180
(99%)
> > DEBUG
> > > > > 2: process_point_file() ->  var_count=180, grid: 97 out of
41616
> > > > >
> > > > > - The input data has 180 TMP data
> > > > > - One TMP data does not belong to the target grid
> > > > > - 179 TMP data were mapped to 97 grid cells (0.23% = 97 /
41616 *
> > 100)
> > > > >   ==> 82 cells (82 = 179 - 97, for TO_GRID) have two or more
TMP
> data
> > > > >
> > > > > The plot has valid data for 97 cells out of 41616 (which is
204 by
> > > > > 204) grid cells)
> > > > >
> > > > >
> > > > > Two issues with regrid_data_plane (this specific case
because of
> > 0.23%
> > > > > coverage).
> > > > >
> > > > > 1. "-vld_thresh" should be set to 0 (disable the vld_thresh)
OR
> 0.008
> > > > > (<=
> > > > > 1 / 121).
> > > > >
> > > > > The default setting of "-vld_thresh" is 0.5.
> > > > > The max raduis (MAX_R) for the Gaussuan smoothing is 5. The
matrix
> > for
> > > > > Gaussian smoothing is 11 by 11 (11 = 2 * MAX_R + 1). To be a
valid
> > > > > regrdded value for the taeget cell, it requires 61 (from 121
/ 2)
> > > > > neighbers with the valid TMP data. Only 97 cells have the
valid
> data
> > > and
> > > > > they are spreaded.
> > > > > None of them satisfies the number of the valid neighber
cells.
> That's
> > > > > why the output of regrid_data_plae was all missing values.
> > > > >
> > > > > 2. Bug on computing the Gaussian value.
> > > > >
> > > > > I will fix this.
> > > > >
> > > > > I attached the updated NetCDF file (output of
regrid_data_plane)
> and
> > > > > the plot.
> > > > >
> > > > > Cheers,
> > > > > Howard
> > > > >
> > > > > On Wed Jul 15 13:10:58 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > Thanks for working to improve the documentation. So you
can only
> > use
> > > > > > the interpolation methods MAX/MIN/UW_MEAN/MEDIAN in
point2grid
> > > > > > unless you are working with a probability variable in
which case
> > you
> > > > > > can use MAXGAUSS. The default of these is UW_MEAN. The
output I
> > > > > > generated in my first test used that method since I didn't
> specify
> > > > > > the -method. In this case, the plot I generated using
> > > > > > plot_data_plane shows a map with data points at random
locations
> > > > > > which correspond to the locations of the original point
obs from
> > the
> > > > > > output of ascii2nc. My expectation was that I would see
the data
> > > > > > points with TMP values in a regularly spaced grid with 1km
> spacing.
> > > > > > In theory, this NetCDF file could be used in Grid-Stat to
verify
> a
> > > > > > TMP fcst on the same 1km grid. Why doesn't the plot show a
> gridded
> > > TMP
> > > > > > field?
> > > > > >
> > > > > > R/
> > > > > > John
> > > > > >
> > > > > >
> > > > > > ________________________________________
> > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > Sent: Wednesday, July 15, 2020 12:17 PM
> > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > Point2Grid
> > > > > > (UNCLASSIFIED)
> > > > > >
> > > > > > All active links contained in this email were disabled.
Please
> > > > > > verify the identity of the sender, and confirm the
authenticity
> of
> > > > > > all links contained within the message prior to copying
and
> pasting
> > > > > > the address to a Web browser.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > ----
> > > > > >
> > > > > > Yes, you are right. It's not the output you are expecting.
We
> will
> > > > > > improve the documentation on this (Caution-
> > > > > > Caution-Caution-Caution-Caution-
> https://github.com/NCAR/MET/issues/1413).
> > > > > >
> > > > > > Unlike regrid_data_plane, the MAXGAUSE method with
point2grid
> works
> > > > > > differently: MAX interpolation for the raw data (TMP) and
> GAUSSIAN
> > > > > > smoothing into the probability variable (TMP_prob_NA or
> > TMP_prob_XX).
> > > > > > MAXGAUSS is the same as the MAX interpolation for the TMP
> variable
> > > > > > for the point2grid. So the difference is between MAX VS.
UW_MEAN
> > > > > > interpolation methods, and GAUSIAN smoothing was not
involved.
> > > > > >
> > > > > > The right way for the Gaussian smoothing to the raw data
will be:
> > > > > > 1) regridding by using point2grid with one of
interpolation
> methods
> > > > > > (MAX/MIN/UW_MEAN/MEDIAN)
> > > > > > 2) regridding by using regrid_data_plane with MAXGAUSS
method and
> > > > > > the same setting as the input and the target grid.
> > > > > >
> > > > > > The output of regrid_data_plane (MAXGAUSS) is not right.
I'm
> > working
> > > > > > on this.
> > > > > >
> > > > > > Cheers,
> > > > > > Howard
> > > > > >
> > > > > >
> > > > > > On Wed Jul 15 10:30:53 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > Howard -
> > > > > > >
> > > > > > > Today I successfully ran point2grid but used -method
MAXGAUSS
> and
> > > > > > > it ran without error and produced output. I'm uploading
the
> file
> > > > > > > list, output, commands used, log file, and plots to the
DoD
> SAFE
> > > > > > > so you can take a look. I am also uploading the results
of the
> > > > > > > previous run where you suggested that I use "11" vice
"TMP"
> which
> > > > > > > fixed the problem and ran successfully.
> > > > > > >
> > > > > > > The plots are not what I was expecting somehow. The
plotted
> data
> > > > > > > look like just points where the original obs were
located
> rather
> > > > > > > than a regularly spaced grid of TMP values. What are
your
> > thoughts
> > > > on
> > > > > > > this?
> > > > > > >
> > > > > > > Thanks
> > > > > > > ________________________________________
> > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > Sent: Monday, July 13, 2020 3:33 PM
> > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747] MET
V9.0
> > > > > > > Point2Grid
> > > > > > > (UNCLASSIFIED)
> > > > > > >
> > > > > > > All active links contained in this email were disabled.
Please
> > > > > > > verify the identity of the sender, and confirm the
authenticity
> > of
> > > > > > > all links contained within the message prior to copying
and
> > > > > > > pasting the address to a Web browser.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----
> > > > > > >
> > > > > > > Something is not right. I will take a look
> > > > > > >
> > > > > > > DEBUG 2: Range of input data (name="TMP";
level="(*,*)";) is
> > > > > > > 246.351 to 284.817.
> > > > > > > DEBUG 2: Range of regridded data (name="TMP";
level="(*,*)";)
> is
> > > > > > > -9999
> > > > > > > to -9999.
> > > > > > >
> > > > > > > The input is OK (between 246.351 and 284.817), but the
regrided
> > > > > > > output is all missing value (-9999).
> > > > > > >
> > > > > > > Cheers,
> > > > > > > Howard
> > > > > > >
> > > > > > > The input file has
> > > > > > > On Mon Jul 13 13:58:03 2020, john.w.raby2.civ at mail.mil
wrote:
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > > Howard -
> > > > > > > >
> > > > > > > > The next run with MAXGAUSS was successful, although I
have to
> > > > > > > > plot it to see what it looks like. I've attached the
log
> file.
> > > > > > > > Does it indicate any issues with how it ran?
> > > > > > > >
> > > > > > > > R/
> > > > > > > > John
> > > > > > > >
> > > > > > > > -----Original Message-----
> > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > Sent: Monday, July 13, 2020 1:14 PM
> > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET V9.0
> > > > > > > > Point2Grid
> > > > > > > > (UNCLASSIFIED)
> > > > > > > >
> > > > > > > > All active links contained in this email were
disabled.
> Please
> > > > > > > > verify the identity of the sender, and confirm the
> authenticity
> > > > > > > > of all links contained within the message prior to
copying
> and
> > > > > > > > pasting the address to a Web browser.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > ----
> > > > > > > >
> > > > > > > > I'm sorry. GAUSSIAN for regrid_data_plane was renamed
to
> > > > > > > > MAXGAUSS and GAUSSIAN method is only used for the
> > interpolation.
> > > > > > > > Please set "-method MAXGAUSE". The from_grid and
to_grid are
> > the
> > > > > > > > same, there is no difference with additional MAX
> interpolation.
> > > > > > > >
> > > > > > > > Cheers,
> > > > > > > > Howard
> > > > > > > >
> > > > > > > >
> > > > > > > > On Mon Jul 13 12:55:31 2020, john.w.raby2.civ at mail.mil
> wrote:
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > > Howard -
> > > > > > > > >
> > > > > > > > > I tried running MET V9.0 regrid_data_plane following
your
> > > > > > > > > example below, but I'm getting and error:
> > > > > > > > > ERROR  :
> > > > > > > > > ERROR  : RegridInfo::validate() -> "GAUSSIAN" not
valid for
> > > > > > > > > regridding, only interpolating.
> > > > > > > > >
> > > > > > > > > The following is the command I used to run it:
> > > > > > > > >  [~/MET/METV9.0]$ /usr/local/bin/regrid_data_plane
> > > > > > > > >
> > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > > V2016.nc
> > > > > > > > >
> > > > >
> >
/p/home/jraby/MET/MET_Point2Grid/results_P2G_test/1km_TMP_1200UTC_11NO
> > > > > > > > > V2016.nc
> > > > > > > > >
> > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/results_gauss_smoothing/1km_TM
> > > > > P_1200UTC_11NOV2016_gaussian.nc
> > > > > > > > >   -field 'name="TMP"; level="(*,*)";' -method
GAUSSIAN -log
> > > > > > > > >
> > > > > > > > >
> > > > >
> >
/p/home/jraby/MET/MET_Regrid_Data_Plane/logs/1km_TMP_12Z_11NOV2016_gau
> > > > > > > > > ss_test
> > > > > > > > > -v 9
> > > > > > > > >
> > > > > > > > > Any suggestions on how to perform Gaussian smoothing
on the
> > > > > > > > > output TMP field from point2grid?
> > > > > > > > >
> > > > > > > > > Thanks.
> > > > > > > > >
> > > > > > > > > R/
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > -----Original Message-----
> > > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > > Sent: Monday, July 6, 2020 11:38 PM
> > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu #95747]
MET
> V9.0
> > > > > > > > > Point2Grid
> > > > > > > > > (UNCLASSIFIED)
> > > > > > > > >
> > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > Please verify the identity of the sender, and
confirm the
> > > > > > > > > authenticity of all links contained within the
message
> prior
> > > > > > > > > to copying and pasting the address to a Web browser.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > ----
> > > > > > > > >
> > > > > > > > > Sorry for the late response.
> > > > > > > > >
> > > > > > > > > The input and to_grid must be the same for Gaussian
> smoothing
> > > > > > > > > and regrid_data_plane shows the usage message if the
> to_grid
> > > > > > > > > is missing.
> > > > > > > > > There are
> > > > > > > > > two "pb2nc_TMP.nc"s as the input NetCDF and as the
to_grid.
> > > > > > > > >
> > > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > > pb2nc_gaussian_TMP.nc -field 'name="TMP";
level="(*,*)";' \
> > > > > > > > > -method GAUSSIAN -v 1
> > > > > > > > >
> > > > > > > > > Cheers,
> > > > > > > > > Howard
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On Tue Jun 30 06:58:02 2020,
john.w.raby2.civ at mail.mil
> > wrote:
> > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > >
> > > > > > > > > > Howard -
> > > > > > > > > >
> > > > > > > > > > Thanks for tailoring the command for my particular
> > interests.
> > > > > > > > > > Just
> > > > > > > > > > curious about your regrid_data_plane command. You
> provided
> > > > > > > > > > the input file from point2grid (pb2nc_TMP.nc ) as
the
> > > > > > > > > > input_filename to regrid_data_plane and then you
provide
> > the
> > > > > > > > > > name of the file to be output from
regrid_data_plane
> > > > > > > > > > (pb2nc_TMP.nc pb2nc_gaussian_TMP.nc), but you
didn't
> > provide
> > > > > > > > > > the to_grid. Is this because it knows that the
grids of
> the
> > > > > > > > > > input and output files are the same and thus
ignores the
> > > > > > > > > > missing to_grid argument?
> > > > > > > > > >
> > > > > > > > > > R/
> > > > > > > > > > John
> > > > > > > > > >
> > > > > > > > > > -----Original Message-----
> > > > > > > > > > From: Howard Soh via RT <met_help at ucar.edu>
> > > > > > > > > > Sent: Monday, June 29, 2020 2:50 PM
> > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > <john.w.raby2.civ at mail.mil>
> > > > > > > > > >  Subject: [Non-DoD Source] [rt.rap.ucar.edu
#95747] MET
> > V9.0
> > > > > > > > > > Point2Grid
> > > > > > > > > > (UNCLASSIFIED)
> > > > > > > > > >
> > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > Please verify the identity of the sender, and
confirm the
> > > > > > > > > > authenticity of all links contained within the
message
> > prior
> > > > > > > > > > to copying and pasting the address to a Web
browser.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > ----
> > > > > > > > > >
> > > > > > > > > > The Gaussian smoothing is available with
> regrid_data_plane,
> > > > > > > > > > too.
> > > > > > > > > > In
> > > > > > > > > > this case, we are interested in Gaussian smoothing
> without
> > > > > > > > > > regriding.
> > > > > > > > > > The input and the target grid should be the same,
which
> is
> > > > > > > > > > the output of point2grid.
> > > > > > > > > >
> > > > > > > > > > Here is an example (assuming the point data file
is
> > > > > > > > > > generated by pb2nc).
> > > > > > > > > >
> > > > > > > > > > ./point2grid ndas.20120409.t12z.prepbufr.nc \
> > > > > > > > > >       G212 pb2nc_TMP.nc \
> > > > > > > > > >       -field 'name="TMP"; level="*";' \
> > > > > > > > > >       -v 1
> > > > > > > > > >
> > > > > > > > > > ./regrid_data_plane pb2nc_TMP.nc pb2nc_TMP.nc
> > > > > > > > > > pb2nc_gaussian_TMP.nc
> > > > > > > > > >       -field 'name="TMP"; level="(*,*)";' \
> > > > > > > > > >       -method GAUSSIAN -v 1
> > > > > > > > > >
> > > > > > > > > > Note: override -gaussian_dx (81.271) &
-gaussian_radius
> > > > > > > > > > (120.0) if necessary
> > > > > > > > > >
> > > > > > > > > > Cheers,
> > > > > > > > > > Howard
> > > > > > > > > >
> > > > > > > > > > On Mon Jun 29 14:13:44 2020,
john.w.raby2.civ at mail.mil
> > > wrote:
> > > > > > > > > > > Howard -
> > > > > > > > > > >
> > > > > > > > > > > Thanks for your quick response with details in
answer
> to
> > > > > > > > > > > my question.
> > > > > > > > > > > Right now my interest is with non-probability
fields,
> so
> > > > > > > > > > > your point in 1. about the possibility of using
> Gaussian
> > > > > > > > > > > smoothing for -field argument is interesting.
> > > > > > > > > > >
> > > > > > > > > > > R/
> > > > > > > > > > > John
> > > > > > > > > > > ________________________________________
> > > > > > > > > > > From: Howard Soh via RT [met_help at ucar.edu]
> > > > > > > > > > > Sent: Monday, June 29, 2020 12:50 PM
> > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > Subject: [Non-DoD Source] [rt.rap.ucar.edu
#95747] MET
> > > > > > > > > > > V9.0 Point2Grid
> > > > > > > > > > > (UNCLASSIFIED)
> > > > > > > > > > >
> > > > > > > > > > > All active links contained in this email were
disabled.
> > > > > > > > > > > Please
> > > > > > > > > > > verify the identity of the sender, and confirm
the
> > > > > > > > > > > authenticity of all links contained within the
message
> > > > > > > > > > > prior to copying and pasting the address to a
Web
> > browser.
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > >
> > > > > > > > > > > ----
> > > > > > > > > > >
> > > > > > > > > > > Two layers for the interpolation with
point2grid:
> > > > > > > > > > > - layer 1: MIN/MAX/MEDIAN/UW_MEAN interpolation
for the
> > > > > > > > > > > raw data
> > > > > > > > > > > - layer 2: Optional, the GAUSSIAN smoothing for
the
> > > > > > > > > > > probability variable. (the existence of valid
values if
> > > > > > > > > > > the probability threshold is not given)
> > > > > > > > > > >
> > > > > > > > > > > + "-method GAUSSIAN" is processed as the "-
method
> > UW_MEAN"
> > > > > > > > > > > (which
> > > > > > > > > > > is
> > > > > > > > > > > a
> > > > > > > > > > > default) for the layer 1 and "-method GAUSSIAN"
for the
> > > > > > > > > > > layer 2.
> > > > > > > > > > > + "-method MAXGAUSS" is processed as the "-
method MAX"
> > for
> > > > > > > > > > > the
> > > > > > > > > > > layer
> > > > > > > > > > > 1
> > > > > > > > > > > and "-method GAUSSIAN" for the layer 2.
> > > > > > > > > > > + Override the interpolation method of layer 1
with
> > > > > > > > > > > + "-method
> > > > > > > > > > > <METHOD>"
> > > > > > > > > > >
> > > > > > > > > > > 1. The Gaussian smoothing at point2grid is
activated
> and
> > > > > > > > > > > applied into the probability variable only if
-method
> > > > > > > > > > > GAUSSIAN or -method MAXGAUSS is given as the
command
> line
> > > > > > > > > > > argument. It's not applied into the variable at
> "-field"
> > > > > > > > > > > argument. If you want, You can do this by using
> > > > > > > > > > > regrid_data_plane (passing the output of
point2grid as
> > the
> > > > > > > > > > > input of regrid_data_plane with GAUSSIAN
method).
> > > > > > > > > > >
> > > > > > > > > > > 2. -gaussian_dx and -gaussian_radius are ignored
if
> > > > > > > > > > > -method GAUSSIAN or -method MAXGAUSS is NOT
given. The
> > > > > > > > > > > default values are applied if not given.
> > > > > > > > > > >
> > > > > > > > > > > 3 -prob_cat_thresh applied regardless -method
GAUSSIAN
> or
> > > > > > > > > > > -method MAXGAUSS argument.
> > > > > > > > > > >
> > > > > > > > > > > 4 Without -prob_cat_thresh argument, it will be
the
> > status
> > > > > > > > > > > of data existence. With -prob_cat_thresh
argument, it
> > will
> > > > > > > > > > > be meeting the condition or not. So the values
of the
> > > > > > > > > > > probability are 0 or 1.
> > > > > > > > > > >
> > > > > > > > > > > 5 Gaussian smoothing is applied if GAUSSIAN
smoothing
> is
> > > > > > > > > > > requested (between 0. and 1.0 with GAUSSIAN
smoothing).
> > > > > > > > > > >
> > > > > > > > > > > Cheers,
> > > > > > > > > > > Howard
> > > > > > > > > > >
> > > > > > > > > > > On Mon Jun 29 11:31:23 2020, johnhg wrote:
> > > > > > > > > > > > John,
> > > > > > > > > > > >
> > > > > > > > > > > > I'm reassigning this ticket to Howard Soh who
worked
> on
> > > > > > > > > > > > the development of the point2grid tool. He
should be
> > > > > > > > > > > > able to better answer these questions.
> > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks,
> > > > > > > > > > > > John
> > > > > > > > > > > >
> > > > > > > > > > > > On Mon, Jun 29, 2020 at 10:58 AM Raby, John W
USA CIV
> > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > >
> >
> <Caution-Caution-Caution-Caution-Caution-Caution-Caution-Caution-
Caution-Caution-url:
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-Caution-
> > > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Disp
> > > > > > > > > > > > > lay.html?i
> > > > > > > > > > > > > d=95747 >
> > > > > > > > > > > > >
> > > > > > > > > > > > > John -
> > > > > > > > > > > > >
> > > > > > > > > > > > > I ran it to see the usage and there are
details in
> > the
> > > > > > > > > > > > > usage dialog and also in the XML file you
pointed
> me
> > > > > > > > > > > > > to which helped to answer my own question (I
> think).
> > > > > > > > > > > > > There is an additional interpolation method
called
> > > > > > > > > > > > > GAUSSIAN which applies only when using the
> > probability
> > > > > > > > > > > > > variable (- prob_cat_thresh 280.0).
> > > > > > > > > > > > > So, it
> > > > > > > > > > > > > looks like the two optional smoothing
arguments
> > > > > > > > > > > > > -gaussian_dx and -gaussian_radius are only
used
> when
> > > > > > > > > > > > > the method is GAUSSIAN. IS this correct?
This means
> > > > > > > > > > > > > that there is no smoothing except when using
> > > > > > > > > > > > > probability variables. Is this correct?
> > > > > > > > > > > > >
> > > > > > > > > > > > > R/
> > > > > > > > > > > > > John
> > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > From: John Halley Gotway via RT
[met_help at ucar.edu
> ]
> > > > > > > > > > > > > Sent: Monday, June 29, 2020 9:41 AM
> > > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > Subject: Re: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > >
> > > > > > > > > > > > > All active links contained in this email
were
> > disabled.
> > > > > > > > > > > > > Please verify the identity of the sender,
and
> confirm
> > > > > > > > > > > > > the authenticity of all links contained
within the
> > > > > > > > > > > > > message prior to copying and pasting the
address
> to a
> > > > > > > > > > > > > Web browser.
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > ----
> > > > > > > > > > > > >
> > > > > > > > > > > > > John,
> > > > > > > > > > > > >
> > > > > > > > > > > > > While the "-field" command line argument
indicates
> > > > > > > > > > > > > that it can be used one or more times, in
this
> first
> > > > > > > > > > > > > version of the tool, it can only be used
once.
> > > > > > > > > > > > >
> > > > > > > > > > > > > If you specify "-field" more that once
you'll get
> > this
> > > > > > > > > > > > > error
> > > > > > > > > > > > > message:
> > > > > > > > > > > > > ERROR  :
> > > > > > > > > > > > > ERROR  : process_point_file() -> Support
only 1
> > > variable.
> > > > > > > > > > > > > "2"
> > > > > > > > > > > > > variables
> > > > > > > > > > > > > were requested.
> > > > > > > > > > > > > ERROR  :
> > > > > > > > > > > > >
> > > > > > > > > > > > > We do plan to support the processing of
multiple
> > > > > > > > > > > > > fields but it has not yet been implemented.
> > > > > > > > > > > > >
> > > > > > > > > > > > > As for the science behind the interpolation
> methods,
> > > > > > > > > > > > > it's very simple. Keep track of the point
> > observations
> > > > > > > > > > > > > falling inside each of the target grid
boxes. Write
> > to
> > > > > > > > > > > > > the output the min, max, unweighted mean, or
median
> > of
> > > > > > > > > > > > > those values based on the method chosen. If
you
> only
> > > > > > > > > > > > > have 1 point obs falling inside each grid
box, then
> > > > > > > > > > > > > all those methods will result in the same
output
> > > > > > > > > > > > > value.
> > > > > > > > > > > > > Whereas if you have many, as in the case of
GOES
> > input
> > > > > > > > > > > > > files, then the choice of the derivation
method
> would
> > > > > > > > > > > > > matter.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > > >
> > > > > > > > > > > > > Thanks,
> > > > > > > > > > > > > John
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > On Sat, Jun 27, 2020 at 8:49 PM Raby, John W
USA
> CIV
> > > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > <Caution-Caution-Caution-Caution-Caution-
Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-Caution-Caution-
url:
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Display.html
> > > > > > > > > > > > > ?id=
> > > > > > > > > > > > > 95747 >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Another question I have is how is -method
used in
> > > > > > > > > > > > > > this tool?
> > > > > > > > > > > > > > In the
> > > > > > > > > > > > > User's
> > > > > > > > > > > > > > Guide, -method provides options for
regridding
> from
> > > > > > > > > > > > > > one grid to another such as is used in
> > > > > > > > > > > > > > regrid_data_plane. In the case of
Point2Grid, the
> > > > > > > > > > > > > > process involved is not regridding in the
way
> that
> > I
> > > > > > > > > > > > > > understand it.
> > > > > > > > > > > > > > If I
> > > > > > > > > > > > > > understand it correctly, it takes point
> > observations
> > > > > > > > > > > > > > and distributes them on to a user provided
grid
> > > > > > > > > > > > > > using some interpolation method. The next
> > > > > > > > > > > > > step I
> > > > > > > > > > > > > > think is the smoothing of the field to
even out
> > > > > > > > > > > > > > irregularities which are caused by errors,
noise,
> > > > > > > > > > > > > > etc resulting in the fitting of the
observed
> values
> > > > > > > > > > > > > > to some function which appear to be a
Gaussian
> > > > > > > > > > > > > > function. I guess I also need to
understand what
> > > > > > > > > > > > > > happens when the tool is used without any
of the
> > > > > > > > > > > > > > optional arguments.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > R/
> > > > > > > > > > > > > > John
> > > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > > From: Raby, John W CIV USARMY CCDC ARL
(USA)
> > > > > > > > > > > > > > Sent: Saturday, June 27, 2020 7:59 AM
> > > > > > > > > > > > > > To: met_help at ucar.edu
> > > > > > > > > > > > > > Subject: RE: [Non-DoD Source] Re: [
> rt.rap.ucar.edu
> > > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > > V9.0
> > > > > > > > > > > > > > Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > John -
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Thanks for your quick response which is
helpful.
> > > > > > > > > > > > > > Appreciate
> > > > > > > > > > > > > > your dedication so late into the evening!
Can I
> add
> > > > > > > > > > > > > > new variables in the
> > > > > > > > > > > > > config
> > > > > > > > > > > > > > file? Any info on the science behind the
> > > > > > > > > > > > > > interpolation methods would be
appreciated, but
> > that
> > > > > > > > > > > > > > can wait until Monday.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Take care and stay safe.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > R/
> > > > > > > > > > > > > > John
> > > > > > > > > > > > > > ________________________________________
> > > > > > > > > > > > > > From: John Halley Gotway via RT [
> met_help at ucar.edu
> > ]
> > > > > > > > > > > > > > Sent: Friday, June 26, 2020 11:06 PM
> > > > > > > > > > > > > > To: Raby, John W CIV USARMY CCDC ARL (USA)
> > > > > > > > > > > > > > Subject: [Non-DoD Source] Re:
[rt.rap.ucar.edu
> > > > > > > > > > > > > > #95747] MET
> > > > > > > > > > > > > > V9.0 Point2Grid (UNCLASSIFIED)
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > All active links contained in this email
were
> > > disabled.
> > > > > > > > > > > > > > Please verify
> > > > > > > > > > > > > the
> > > > > > > > > > > > > > identity of the sender, and confirm the
> > authenticity
> > > > > > > > > > > > > > of all links
> > > > > > > > > > > > > contained
> > > > > > > > > > > > > > within the message prior to copying and
pasting
> the
> > > > > > > > > > > > > > address to a Web browser.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > ----
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > John,
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > I see you have a question about
interpolation
> > > > > > > > > > > > > > options in the point2grid tool. This XML
file
> > > > > > > > > > > > > > contains the calls to the point2grid tool
that
> are
> > > > > > > > > > > > > run
> > > > > > > > > > > > > > for our unit tests:
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > Caution-Caution-https://github.com/NCAR/MET/blob/master_v9.0/t
> > > > > > > > > > > > > est/xml/
> > > > > > > > > > > > > un
> > > > > > > > > > > > > it
> > > > > > > > > > > > > _point2grid.xml
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > While most of them DO NOT use the "-
method"
> command
> > > > > > > > > > > > > > line option, a few at the end do use "-
method
> MAX"
> > > > > > > > > > > > > > to select the maximum value falling within
each
> > grid
> > > > > > > > > > > > > > box.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Inspecting the code in this file:
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > Caution-Caution-
> > > Caution-Caution-https://github.com/NCAR/MET/blob/master_v9.0/m
> > > > > > > > > > > > > et/src/t
> > > > > > > > > > > > > oo
> > > > > > > > > > > > > ls
> > > > > > > > > > > > > /other/point2grid/point2grid.cc
> > > > > > > > > > > > > > I see that the supported options for
-method are
> > > > > > > > > > > > > > MIN, MAX, MEDIAN, and UW_MEAN.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Seems like we should add that info to the
usage
> > > > > > > > > > > > > > statement!
> > > > > > > > > > > > > > I
> > > > > > > > > > > > > > wrote up
> > > > > > > > > > > > > this
> > > > > > > > > > > > > > GitHub issue to make sure we remember to
do so
> > > > > > > > > > > > > > before the
> > > > > > > > > > > > > > 9.1
> > > > > > > > > > > > > > release:
> > > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-
> > Caution-https://github.com/NCAR/MET/issu
> > > > > > > > > > > > > > es/1396
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Hope that helps clarify.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Thanks,
> > > > > > > > > > > > > > John
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > On Fri, Jun 26, 2020 at 8:40 PM Raby, John
W USA
> > CIV
> > > > > > > > > > > > > > via RT < met_help at ucar.edu> wrote:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Fri Jun 26 20:40:22 2020: Request 95747
was
> acted
> > > > > > > > > > > > > > > upon.
> > > > > > > > > > > > > > > Transaction: Ticket created by
> > > > > > > > > > > > > > > john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > > >        Queue: met_help
> > > > > > > > > > > > > > >      Subject: MET V9.0 Point2Grid
> (UNCLASSIFIED)
> > > > > > > > > > > > > > >        Owner: Nobody
> > > > > > > > > > > > > > >   Requestors: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > > >       Status: new
> > > > > > > > > > > > > > >  Ticket
> <Caution-Caution-Caution-Caution-Caution-
> > > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > >
> Caution-Caution-Caution-Caution-Caution-Caution-url:
> > > > > > > > > > > > > > > Caution-
> > > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > > Caution-Caution-
> > > > > > > > > > > > > > Caution-Caution-Caution-
> > > Caution-Caution-https://rt.rap.ucar.edu/rt/Ticket/Di
> > > > > > > > > > > > > > splay.html
> > > > > > > > > > > > > > ?id=95747
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > The optional argument for smoothing or
> > > > > > > > > > > > > > > interpolation shown in the
> > > > > > > > > > > > > User's
> > > > > > > > > > > > > > > Guide is Gaussian, but are other methods
> > available?
> > > > > > > > > > > > > > > If
> > > > > > > > > > > > > > > so,
> > > > > > > > > > > > > > > can you
> > > > > > > > > > > > > > provide
> > > > > > > > > > > > > > > the names and how to specify them as an
> optional
> > > > > > > > > > > > > > > argument?
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Thanks.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Stay safe.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > T/
> > > > > > > > > > > > > > > John
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > Mr. John W. Raby
> > > > > > > > > > > > > > > U.S. Army Research Laboratory White
Sands
> Missile
> > > > > > > > > > > > > > > Range, NM
> > > > > > > > > > > > > > > 88002
> > > > > > > > > > > > > > > Office:(575) 678-2004 DSN 258-2004
> > > > > > > > > > > > > > > Teleworking: Contact Admin Specialist at
(410)
> > > > > > > > > > > > > > > 278-
> > > > > > > > > > > > > > > 6974
> > > > > > > > > > > > > > > FAX
> > > > > > > > > > > > > > > (575) 678-1230 DSN 258-1230
> > > > > > > > > > > > > > > Email: john.w.raby2.civ at mail.mil
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > CLASSIFICATION: UNCLASSIFIED
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > > CLASSIFICATION: UNCLASSIFIED
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
>

------------------------------------------------

More information about the Met_help mailing list