[Met_help] [rt.rap.ucar.edu #45409] History for MODE tool

Fri Apr 8 11:33:59 MDT 2011

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

MET-Help,

I hope this email finds you well.  We're beginning to explore the MODE
Tool for verifying precipitation forecasts.  Since our company forecasts
globally, we are wondering if you know of anyone using MODE outside of
the U.S.?  And if so, what they are using for the observation grids?  In
particular, we're wondering about places where radar images are not
publicly accessible and metars are sparse.  

One thing we have considered is creating a grid out of lightning strike
data to make a pseudo-radar image.  Of course, this would work well in
some areas (like the tropics) and not so well in other areas.  We are
particularly concerned with verifying our forecasts in Indonesia, so
that is why the lightning idea was proposed.

Hoping to get your thoughts!  

Thanks,

Holly

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

Subject: Re: [rt.rap.ucar.edu #45409] MODE tool
From: John Halley Gotway
Time: Wed Mar 16 10:21:36 2011

Holly,

FYI, I created a met_help ticket for this so we can track the user
support we provide.

Yes, gathering high quality observations are always an issue for
verification.  And over CONUS, we're lucky to have good observation
sources.  It seems to me that your options would include:

   (1) Compare your model output to an analysis field from a different
model.  Verifying against your own model's analysis is definitely
cheating, but comparing it against a different model's analysis
is a bit more defensible.
   (2) I do know of some folks who have verified against 3-hourly TRMM
rainfall estimates (http://trmm.gsfc.nasa.gov/).  But TRMM data is
only available over about +/- 50 degrees latitude - so not
really global.  And there'd be some data reformatting required since
MET isn't set up to read that data directly.  But TRMM would be worth
taking a look at.

Hope that helps.

Thanks,
John Halley Gotway
met_help at ucar.edu

On 03/16/2011 10:20 AM, RAL HelpDesk {for John Halley Gotway} wrote:
>
> Wed Mar 16 10:20:39 2011: Request 45409 was acted upon.
> Transaction: Ticket created by johnhg
>        Queue: met_help
>      Subject: MODE tool
>        Owner: johnhg
>   Requestors: hhassenzahl at wxc.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45409 >
>
>
> MET-Help,
>
> I hope this email finds you well.  We're beginning to explore the
MODE
> Tool for verifying precipitation forecasts.  Since our company
forecasts
> globally, we are wondering if you know of anyone using MODE outside
of
> the U.S.?  And if so, what they are using for the observation grids?
In
> particular, we're wondering about places where radar images are not
> publicly accessible and metars are sparse.
>
>
>
> One thing we have considered is creating a grid out of lightning
strike
> data to make a pseudo-radar image.  Of course, this would work well
in
> some areas (like the tropics) and not so well in other areas.  We
are
> particularly concerned with verifying our forecasts in Indonesia, so
> that is why the lightning idea was proposed.
>
>
>
> Hoping to get your thoughts!
>
> Thanks,
>
> Holly

------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #45409] MODE tool
From: hhassenzahl at wxc.com
Time: Wed Mar 16 11:33:27 2011

Thanks for your response, John.  The TRMM idea is good one...I hadn't
thought of that before.  Do you happen to know what format their data
comes in?  Also, do you have contact info for the folks who have used
this as their obs field?  I wouldn't mind picking their brains a
little
bit to further explore this option.

Thanks,
Holly

________________________________
Holly C. Hassenzahl
Science Analyst, Meteorologist
Data Products Group
Weather Central, LP
401 Charmany Drive
Madison, WI 53719
(608)274-5789 ext 288

-----Original Message-----
From: RAL HelpDesk {for John Halley Gotway} [mailto:met_help at ucar.edu]
Sent: Wednesday, March 16, 2011 11:22 AM
To: Holly Hassenzahl
Subject: Re: [rt.rap.ucar.edu #45409] MODE tool

Holly,

FYI, I created a met_help ticket for this so we can track the user
support we provide.

Yes, gathering high quality observations are always an issue for
verification.  And over CONUS, we're lucky to have good observation
sources.  It seems to me that your options would include:

   (1) Compare your model output to an analysis field from a different
model.  Verifying against your own model's analysis is definitely
cheating, but comparing it against a different model's analysis
is a bit more defensible.
   (2) I do know of some folks who have verified against 3-hourly TRMM
rainfall estimates (http://trmm.gsfc.nasa.gov/).  But TRMM data is
only
available over about +/- 50 degrees latitude - so not
really global.  And there'd be some data reformatting required since
MET
isn't set up to read that data directly.  But TRMM would be worth
taking
a look at.

Hope that helps.

Thanks,
John Halley Gotway
met_help at ucar.edu

On 03/16/2011 10:20 AM, RAL HelpDesk {for John Halley Gotway} wrote:
>
> Wed Mar 16 10:20:39 2011: Request 45409 was acted upon.
> Transaction: Ticket created by johnhg
>        Queue: met_help
>      Subject: MODE tool
>        Owner: johnhg
>   Requestors: hhassenzahl at wxc.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45409
>
>
>
> MET-Help,
>
> I hope this email finds you well.  We're beginning to explore the
MODE
> Tool for verifying precipitation forecasts.  Since our company
forecasts
> globally, we are wondering if you know of anyone using MODE outside
of
> the U.S.?  And if so, what they are using for the observation grids?
In
> particular, we're wondering about places where radar images are not
> publicly accessible and metars are sparse.
>
>
>
> One thing we have considered is creating a grid out of lightning
strike
> data to make a pseudo-radar image.  Of course, this would work well
in
> some areas (like the tropics) and not so well in other areas.  We
are
> particularly concerned with verifying our forecasts in Indonesia, so
> that is why the lightning idea was proposed.
>
>
>
> Hoping to get your thoughts!
>
> Thanks,
>
> Holly

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #45409] MODE tool
From: John Halley Gotway
Time: Wed Mar 16 12:20:03 2011

Holly,

A quick search of the met_help emails turned up Erik Noble from NASA
(enoble at giss.nasa.gov).  He's the person for whom I've answered
questions regarding TRMM data.  Actually, if you just type
"met_help trmm" into google, you'd find the archived version of that
met_help correspondence.

Regarding data formats, I believe it's available in a binary format,
but it looks like Erik was using the data in an ASCII format somehow.
Please take a look on the TRMM website:
   http://trmm.gsfc.nasa.gov/data_dir/data.html
At the bottom, there's info about the TRMM-based precip estimates.
The ftp site to which that links contains the TRMM data products as
well as some C/FORTRAN code to read it.  Looks like the 3-hourly
rainfall estimates are available at 1/4 degree resolution.  If I were
in your position, I'd probably write some code to reformat it into a
gridded NetCDF file that MET can read, using the output of
the PCP-Combine tool as a guide - either from the binary data
directly, or in a two step process, dumping the binary to ASCII and
then reformatting the ASCII to NetCDF.

Thanks,
John

On 03/16/2011 11:33 AM, RAL HelpDesk {for hhassenzahl at wxc.com} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45409 >
>
> Thanks for your response, John.  The TRMM idea is good one...I
hadn't
> thought of that before.  Do you happen to know what format their
data
> comes in?  Also, do you have contact info for the folks who have
used
> this as their obs field?  I wouldn't mind picking their brains a
little
> bit to further explore this option.
>
> Thanks,
> Holly
>
> ________________________________
> Holly C. Hassenzahl
> Science Analyst, Meteorologist
> Data Products Group
> Weather Central, LP
> 401 Charmany Drive
> Madison, WI 53719
> (608)274-5789 ext 288
>
> -----Original Message-----
> From: RAL HelpDesk {for John Halley Gotway}
[mailto:met_help at ucar.edu]
> Sent: Wednesday, March 16, 2011 11:22 AM
> To: Holly Hassenzahl
> Subject: Re: [rt.rap.ucar.edu #45409] MODE tool
>
> Holly,
>
> FYI, I created a met_help ticket for this so we can track the user
> support we provide.
>
> Yes, gathering high quality observations are always an issue for
> verification.  And over CONUS, we're lucky to have good observation
> sources.  It seems to me that your options would include:
>
>    (1) Compare your model output to an analysis field from a
different
> model.  Verifying against your own model's analysis is definitely
> cheating, but comparing it against a different model's analysis
> is a bit more defensible.
>    (2) I do know of some folks who have verified against 3-hourly
TRMM
> rainfall estimates (http://trmm.gsfc.nasa.gov/).  But TRMM data is
only
> available over about +/- 50 degrees latitude - so not
> really global.  And there'd be some data reformatting required since
MET
> isn't set up to read that data directly.  But TRMM would be worth
taking
> a look at.
>
> Hope that helps.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
> On 03/16/2011 10:20 AM, RAL HelpDesk {for John Halley Gotway} wrote:
>>
>> Wed Mar 16 10:20:39 2011: Request 45409 was acted upon.
>> Transaction: Ticket created by johnhg
>>        Queue: met_help
>>      Subject: MODE tool
>>        Owner: johnhg
>>   Requestors: hhassenzahl at wxc.com
>>       Status: new
>>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45409
>>
>>
>>
>> MET-Help,
>>
>> I hope this email finds you well.  We're beginning to explore the
MODE
>> Tool for verifying precipitation forecasts.  Since our company
> forecasts
>> globally, we are wondering if you know of anyone using MODE outside
of
>> the U.S.?  And if so, what they are using for the observation
grids?
> In
>> particular, we're wondering about places where radar images are not
>> publicly accessible and metars are sparse.
>>
>>
>>
>> One thing we have considered is creating a grid out of lightning
> strike
>> data to make a pseudo-radar image.  Of course, this would work well
in
>> some areas (like the tropics) and not so well in other areas.  We
are
>> particularly concerned with verifying our forecasts in Indonesia,
so
>> that is why the lightning idea was proposed.
>>
>>
>>
>> Hoping to get your thoughts!
>>
>> Thanks,
>>
>> Holly
>

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