[Met_help] [rt.rap.ucar.edu #41994] History for Lat/Lon Grid in MODE
RAL HelpDesk {for John Halley Gotway}
met_help at ucar.edu
Wed Feb 23 14:47:48 MST 2011
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
I contacted you previously regarding pushing NARCCAP model NetCDF output
into MODE. Now I am trying to push an observational dataset in from
University of Washington. The data is on a lat/lon grid, which I was
reading was supposed to be acceptable for MET 2.0. MODE is definitely not
reading in the grid correctly, and if I set the metadata to say "Lat/Lon"
for the projection attribute, it says it is not supported. I was wondering
if there was something I'm missing to get MODE to read in a Lat/Lon grid?
Thanks for your help again!
Casey Zoellick
czoellic at iastate.edu
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Mon Nov 08 15:57:39 2010
Casey,
I've attached two sample NetCDF files, one that's output from METv2.0
and
one that's output from METv3.0. Please choose the example to use
based on
the version of MET that you're using.
Specifically, I'm guessing you should be specifying the projection as
"LatLon" instead of "Lat/Lon".
Please run "ncdump -h" on these NetCDF files and take a look at the
global
attributes where the grids are defined.
Give that a look, and let me know if you continue to experience
problems.
Thanks,
John
>
> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> Transaction: Ticket created by czoellic at iastate.edu
> Queue: met_help
> Subject: Lat/Lon Grid in MODE
> Owner: Nobody
> Requestors: czoellic at iastate.edu
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
>
> I contacted you previously regarding pushing NARCCAP model NetCDF
output
> into MODE. Now I am trying to push an observational dataset in from
> University of Washington. The data is on a lat/lon grid, which I
was
> reading was supposed to be acceptable for MET 2.0. MODE is
definitely not
> reading in the grid correctly, and if I set the metadata to say
"Lat/Lon"
> for the projection attribute, it says it is not supported. I was
> wondering
> if there was something I'm missing to get MODE to read in a Lat/Lon
grid?
>
> Thanks for your help again!
> Casey Zoellick
> czoellic at iastate.edu
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Thu Nov 11 11:14:27 2010
John,
I adjusted my global attributes, and now MODE does indeed read in the
LatLon
grid and the domain I'm using is what MODE recognizes as well.
However, it
keeps plotting the precip field the same for everyday I run it over a
small
bullet of abnormally high values within my domain. The results are
the same
for everyday. I have checked, and I know the files I'm feeding to
MODE are
indeed different for each day and that the values for the precip are
reasonable in the file. I guess I don't know why it reads in my
domain, but
not the precip values.
Thanks,
Casey
On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Casey,
>
> I've attached two sample NetCDF files, one that's output from
METv2.0 and
> one that's output from METv3.0. Please choose the example to use
based on
> the version of MET that you're using.
>
> Specifically, I'm guessing you should be specifying the projection
as
> "LatLon" instead of "Lat/Lon".
>
> Please run "ncdump -h" on these NetCDF files and take a look at the
global
> attributes where the grids are defined.
>
> Give that a look, and let me know if you continue to experience
problems.
>
> Thanks,
> John
>
>
> >
> > Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> > Transaction: Ticket created by czoellic at iastate.edu
> > Queue: met_help
> > Subject: Lat/Lon Grid in MODE
> > Owner: Nobody
> > Requestors: czoellic at iastate.edu
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> >
> > I contacted you previously regarding pushing NARCCAP model NetCDF
output
> > into MODE. Now I am trying to push an observational dataset in
from
> > University of Washington. The data is on a lat/lon grid, which I
was
> > reading was supposed to be acceptable for MET 2.0. MODE is
definitely
> not
> > reading in the grid correctly, and if I set the metadata to say
"Lat/Lon"
> > for the projection attribute, it says it is not supported. I was
> > wondering
> > if there was something I'm missing to get MODE to read in a
Lat/Lon grid?
> >
> > Thanks for your help again!
> > Casey Zoellick
> > czoellic at iastate.edu
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #41994] Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Thu Nov 11 11:23:49 2010
Casey,
Do you have multiple output times in a single NetCDF file?
If so, you may just be retrieving the first time from that file each
time you run it. It'd probably be easiest for you to just send me
your NetCDF file(s) and the MODE configuration file(s) you're
using. I'll run it here and try to figure out what's going on. If
you have multiple times in your NetCDF file, you'll need to vary the
string that defines which 2D slice you want to extract from
that NetCDF file each time you run MODE.
You can post data to our anonymous ftp site using these instructions:
http://www.dtcenter.org/met/users/support/met_help.php#ftp
Thanks,
John
On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
> John,
>
> I adjusted my global attributes, and now MODE does indeed read in
the LatLon
> grid and the domain I'm using is what MODE recognizes as well.
However, it
> keeps plotting the precip field the same for everyday I run it over
a small
> bullet of abnormally high values within my domain. The results are
the same
> for everyday. I have checked, and I know the files I'm feeding to
MODE are
> indeed different for each day and that the values for the precip are
> reasonable in the file. I guess I don't know why it reads in my
domain, but
> not the precip values.
>
> Thanks,
> Casey
>
>
> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Casey,
>>
>> I've attached two sample NetCDF files, one that's output from
METv2.0 and
>> one that's output from METv3.0. Please choose the example to use
based on
>> the version of MET that you're using.
>>
>> Specifically, I'm guessing you should be specifying the projection
as
>> "LatLon" instead of "Lat/Lon".
>>
>> Please run "ncdump -h" on these NetCDF files and take a look at the
global
>> attributes where the grids are defined.
>>
>> Give that a look, and let me know if you continue to experience
problems.
>>
>> Thanks,
>> John
>>
>>
>>>
>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
>>> Transaction: Ticket created by czoellic at iastate.edu
>>> Queue: met_help
>>> Subject: Lat/Lon Grid in MODE
>>> Owner: Nobody
>>> Requestors: czoellic at iastate.edu
>>> Status: new
>>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>
>>>
>>> I contacted you previously regarding pushing NARCCAP model NetCDF
output
>>> into MODE. Now I am trying to push an observational dataset in
from
>>> University of Washington. The data is on a lat/lon grid, which I
was
>>> reading was supposed to be acceptable for MET 2.0. MODE is
definitely
>> not
>>> reading in the grid correctly, and if I set the metadata to say
"Lat/Lon"
>>> for the projection attribute, it says it is not supported. I was
>>> wondering
>>> if there was something I'm missing to get MODE to read in a
Lat/Lon grid?
>>>
>>> Thanks for your help again!
>>> Casey Zoellick
>>> czoellic at iastate.edu
>>>
>>
>>
>
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Thu Nov 11 11:35:16 2010
John,
There should be only one timestep in each file. I posted a NetCDF
file
before I converted it and one after I'm done converting it as well as
the
MODE config file on the ftp site. I should note that part of the
conversion
process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
grid.
Thanks again,
Casey
On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway} <
met_help at ucar.edu> wrote:
> Casey,
>
> Do you have multiple output times in a single NetCDF file?
>
> If so, you may just be retrieving the first time from that file each
time
> you run it. It'd probably be easiest for you to just send me your
NetCDF
> file(s) and the MODE configuration file(s) you're
> using. I'll run it here and try to figure out what's going on. If
you
> have multiple times in your NetCDF file, you'll need to vary the
string that
> defines which 2D slice you want to extract from
> that NetCDF file each time you run MODE.
>
> You can post data to our anonymous ftp site using these
instructions:
> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>
> Thanks,
> John
>
>
> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> > John,
> >
> > I adjusted my global attributes, and now MODE does indeed read in
the
> LatLon
> > grid and the domain I'm using is what MODE recognizes as well.
However,
> it
> > keeps plotting the precip field the same for everyday I run it
over a
> small
> > bullet of abnormally high values within my domain. The results
are the
> same
> > for everyday. I have checked, and I know the files I'm feeding to
MODE
> are
> > indeed different for each day and that the values for the precip
are
> > reasonable in the file. I guess I don't know why it reads in my
domain,
> but
> > not the precip values.
> >
> > Thanks,
> > Casey
> >
> >
> > On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Casey,
> >>
> >> I've attached two sample NetCDF files, one that's output from
METv2.0
> and
> >> one that's output from METv3.0. Please choose the example to use
based
> on
> >> the version of MET that you're using.
> >>
> >> Specifically, I'm guessing you should be specifying the
projection as
> >> "LatLon" instead of "Lat/Lon".
> >>
> >> Please run "ncdump -h" on these NetCDF files and take a look at
the
> global
> >> attributes where the grids are defined.
> >>
> >> Give that a look, and let me know if you continue to experience
> problems.
> >>
> >> Thanks,
> >> John
> >>
> >>
> >>>
> >>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> >>> Transaction: Ticket created by czoellic at iastate.edu
> >>> Queue: met_help
> >>> Subject: Lat/Lon Grid in MODE
> >>> Owner: Nobody
> >>> Requestors: czoellic at iastate.edu
> >>> Status: new
> >>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
> >>>
> >>>
> >>> I contacted you previously regarding pushing NARCCAP model
NetCDF
> output
> >>> into MODE. Now I am trying to push an observational dataset in
from
> >>> University of Washington. The data is on a lat/lon grid, which
I was
> >>> reading was supposed to be acceptable for MET 2.0. MODE is
definitely
> >> not
> >>> reading in the grid correctly, and if I set the metadata to say
> "Lat/Lon"
> >>> for the projection attribute, it says it is not supported. I
was
> >>> wondering
> >>> if there was something I'm missing to get MODE to read in a
Lat/Lon
> grid?
> >>>
> >>> Thanks for your help again!
> >>> Casey Zoellick
> >>> czoellic at iastate.edu
> >>>
> >>
> >>
> >
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #41994] Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Thu Nov 11 11:42:57 2010
Casey,
I suspect that you made a mistake when posting the data to our ftp
site. The file size and structure of the "temp_1.nc" and
"1982_06_01.nc" files are the same, and neither can be read by MODE.
John
On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
> John,
>
> There should be only one timestep in each file. I posted a NetCDF
file
> before I converted it and one after I'm done converting it as well
as the
> MODE config file on the ftp site. I should note that part of the
conversion
> process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
> grid.
>
> Thanks again,
> Casey
>
> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Casey,
>>
>> Do you have multiple output times in a single NetCDF file?
>>
>> If so, you may just be retrieving the first time from that file
each time
>> you run it. It'd probably be easiest for you to just send me your
NetCDF
>> file(s) and the MODE configuration file(s) you're
>> using. I'll run it here and try to figure out what's going on. If
you
>> have multiple times in your NetCDF file, you'll need to vary the
string that
>> defines which 2D slice you want to extract from
>> that NetCDF file each time you run MODE.
>>
>> You can post data to our anonymous ftp site using these
instructions:
>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>
>> Thanks,
>> John
>>
>>
>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>
>>> John,
>>>
>>> I adjusted my global attributes, and now MODE does indeed read in
the
>> LatLon
>>> grid and the domain I'm using is what MODE recognizes as well.
However,
>> it
>>> keeps plotting the precip field the same for everyday I run it
over a
>> small
>>> bullet of abnormally high values within my domain. The results
are the
>> same
>>> for everyday. I have checked, and I know the files I'm feeding to
MODE
>> are
>>> indeed different for each day and that the values for the precip
are
>>> reasonable in the file. I guess I don't know why it reads in my
domain,
>> but
>>> not the precip values.
>>>
>>> Thanks,
>>> Casey
>>>
>>>
>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Casey,
>>>>
>>>> I've attached two sample NetCDF files, one that's output from
METv2.0
>> and
>>>> one that's output from METv3.0. Please choose the example to use
based
>> on
>>>> the version of MET that you're using.
>>>>
>>>> Specifically, I'm guessing you should be specifying the
projection as
>>>> "LatLon" instead of "Lat/Lon".
>>>>
>>>> Please run "ncdump -h" on these NetCDF files and take a look at
the
>> global
>>>> attributes where the grids are defined.
>>>>
>>>> Give that a look, and let me know if you continue to experience
>> problems.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>>
>>>>>
>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
>>>>> Transaction: Ticket created by czoellic at iastate.edu
>>>>> Queue: met_help
>>>>> Subject: Lat/Lon Grid in MODE
>>>>> Owner: Nobody
>>>>> Requestors: czoellic at iastate.edu
>>>>> Status: new
>>>>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
>>>>>
>>>>>
>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
>> output
>>>>> into MODE. Now I am trying to push an observational dataset in
from
>>>>> University of Washington. The data is on a lat/lon grid, which
I was
>>>>> reading was supposed to be acceptable for MET 2.0. MODE is
definitely
>>>> not
>>>>> reading in the grid correctly, and if I set the metadata to say
>> "Lat/Lon"
>>>>> for the projection attribute, it says it is not supported. I
was
>>>>> wondering
>>>>> if there was something I'm missing to get MODE to read in a
Lat/Lon
>> grid?
>>>>>
>>>>> Thanks for your help again!
>>>>> Casey Zoellick
>>>>> czoellic at iastate.edu
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Thu Nov 11 11:50:02 2010
John,
I accidently copied the 1982_06_01.nc file twice and it got renamed as
temp_1.nc. The 1982_06_01.nc file should be correct and MODE
shouldn't be
able to read it. I just put a file called "new_temp_1.nc" which
should be
the correct converted file. Sorry for the confusion.
-Casey
On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway} <
met_help at ucar.edu> wrote:
> Casey,
>
> I suspect that you made a mistake when posting the data to our ftp
site.
> The file size and structure of the "temp_1.nc" and "1982_06_01.nc"
files
> are the same, and neither can be read by MODE.
>
> John
>
> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> > John,
> >
> > There should be only one timestep in each file. I posted a NetCDF
file
> > before I converted it and one after I'm done converting it as well
as the
> > MODE config file on the ftp site. I should note that part of the
> conversion
> > process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
> > grid.
> >
> > Thanks again,
> > Casey
> >
> > On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Casey,
> >>
> >> Do you have multiple output times in a single NetCDF file?
> >>
> >> If so, you may just be retrieving the first time from that file
each
> time
> >> you run it. It'd probably be easiest for you to just send me
your
> NetCDF
> >> file(s) and the MODE configuration file(s) you're
> >> using. I'll run it here and try to figure out what's going on.
If you
> >> have multiple times in your NetCDF file, you'll need to vary the
string
> that
> >> defines which 2D slice you want to extract from
> >> that NetCDF file each time you run MODE.
> >>
> >> You can post data to our anonymous ftp site using these
instructions:
> >> http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>
> >> Thanks,
> >> John
> >>
> >>
> >> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >>>
> >>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>
> >>> John,
> >>>
> >>> I adjusted my global attributes, and now MODE does indeed read
in the
> >> LatLon
> >>> grid and the domain I'm using is what MODE recognizes as well.
> However,
> >> it
> >>> keeps plotting the precip field the same for everyday I run it
over a
> >> small
> >>> bullet of abnormally high values within my domain. The results
are the
> >> same
> >>> for everyday. I have checked, and I know the files I'm feeding
to MODE
> >> are
> >>> indeed different for each day and that the values for the precip
are
> >>> reasonable in the file. I guess I don't know why it reads in my
> domain,
> >> but
> >>> not the precip values.
> >>>
> >>> Thanks,
> >>> Casey
> >>>
> >>>
> >>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway} <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Casey,
> >>>>
> >>>> I've attached two sample NetCDF files, one that's output from
METv2.0
> >> and
> >>>> one that's output from METv3.0. Please choose the example to
use
> based
> >> on
> >>>> the version of MET that you're using.
> >>>>
> >>>> Specifically, I'm guessing you should be specifying the
projection as
> >>>> "LatLon" instead of "Lat/Lon".
> >>>>
> >>>> Please run "ncdump -h" on these NetCDF files and take a look at
the
> >> global
> >>>> attributes where the grids are defined.
> >>>>
> >>>> Give that a look, and let me know if you continue to experience
> >> problems.
> >>>>
> >>>> Thanks,
> >>>> John
> >>>>
> >>>>
> >>>>>
> >>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> >>>>> Transaction: Ticket created by czoellic at iastate.edu
> >>>>> Queue: met_help
> >>>>> Subject: Lat/Lon Grid in MODE
> >>>>> Owner: Nobody
> >>>>> Requestors: czoellic at iastate.edu
> >>>>> Status: new
> >>>>> Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
> >>>>>
> >>>>>
> >>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
> >> output
> >>>>> into MODE. Now I am trying to push an observational dataset
in from
> >>>>> University of Washington. The data is on a lat/lon grid,
which I was
> >>>>> reading was supposed to be acceptable for MET 2.0. MODE is
> definitely
> >>>> not
> >>>>> reading in the grid correctly, and if I set the metadata to
say
> >> "Lat/Lon"
> >>>>> for the projection attribute, it says it is not supported. I
was
> >>>>> wondering
> >>>>> if there was something I'm missing to get MODE to read in a
Lat/Lon
> >> grid?
> >>>>>
> >>>>> Thanks for your help again!
> >>>>> Casey Zoellick
> >>>>> czoellic at iastate.edu
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #41994] Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Thu Nov 11 14:25:43 2010
Casey,
I think I know what's causing the problem. Unfortunately, our support
for NetCDF is limited in METv2.0, and we're not handling your choice
of fill value appropriately:
APCP_24:_FillValue = 1.e+20f ;
The PCP-Combine tool uses a fill value of "-9999", as shown below, and
I believe the MET code is expecting that.
APCP_12:_FillValue = -9999.f ;
Please try resetting the fill values from "1.0+20f" to "-9999" and try
rerunning. Our support for NetCDF is limited, and we're slowly
working toward adding more general support. Also, in your config
file, I'd suggest setting the raw thresholds to 0.0:
fcst_raw_thresh = "ge0.0";
obs_raw_thresh = "ge0.0";
The convolution radius and threshold (fcst_conv_rad and
fcst_conv_thresh) are what's used to define objects. Most users
should not need to use the raw thresholds unless there's significant
differences in how their forecast and observation fields are defined.
Thanks,
John
On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
> John,
>
> I accidently copied the 1982_06_01.nc file twice and it got renamed
as
> temp_1.nc. The 1982_06_01.nc file should be correct and MODE
shouldn't be
> able to read it. I just put a file called "new_temp_1.nc" which
should be
> the correct converted file. Sorry for the confusion.
>
> -Casey
>
>
> On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Casey,
>>
>> I suspect that you made a mistake when posting the data to our ftp
site.
>> The file size and structure of the "temp_1.nc" and "1982_06_01.nc"
files
>> are the same, and neither can be read by MODE.
>>
>> John
>>
>> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>
>>> John,
>>>
>>> There should be only one timestep in each file. I posted a NetCDF
file
>>> before I converted it and one after I'm done converting it as well
as the
>>> MODE config file on the ftp site. I should note that part of the
>> conversion
>>> process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
>>> grid.
>>>
>>> Thanks again,
>>> Casey
>>>
>>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Casey,
>>>>
>>>> Do you have multiple output times in a single NetCDF file?
>>>>
>>>> If so, you may just be retrieving the first time from that file
each
>> time
>>>> you run it. It'd probably be easiest for you to just send me
your
>> NetCDF
>>>> file(s) and the MODE configuration file(s) you're
>>>> using. I'll run it here and try to figure out what's going on.
If you
>>>> have multiple times in your NetCDF file, you'll need to vary the
string
>> that
>>>> defines which 2D slice you want to extract from
>>>> that NetCDF file each time you run MODE.
>>>>
>>>> You can post data to our anonymous ftp site using these
instructions:
>>>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>>
>>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>>>
>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>>>
>>>>> John,
>>>>>
>>>>> I adjusted my global attributes, and now MODE does indeed read
in the
>>>> LatLon
>>>>> grid and the domain I'm using is what MODE recognizes as well.
>> However,
>>>> it
>>>>> keeps plotting the precip field the same for everyday I run it
over a
>>>> small
>>>>> bullet of abnormally high values within my domain. The results
are the
>>>> same
>>>>> for everyday. I have checked, and I know the files I'm feeding
to MODE
>>>> are
>>>>> indeed different for each day and that the values for the precip
are
>>>>> reasonable in the file. I guess I don't know why it reads in my
>> domain,
>>>> but
>>>>> not the precip values.
>>>>>
>>>>> Thanks,
>>>>> Casey
>>>>>
>>>>>
>>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway} <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Casey,
>>>>>>
>>>>>> I've attached two sample NetCDF files, one that's output from
METv2.0
>>>> and
>>>>>> one that's output from METv3.0. Please choose the example to
use
>> based
>>>> on
>>>>>> the version of MET that you're using.
>>>>>>
>>>>>> Specifically, I'm guessing you should be specifying the
projection as
>>>>>> "LatLon" instead of "Lat/Lon".
>>>>>>
>>>>>> Please run "ncdump -h" on these NetCDF files and take a look at
the
>>>> global
>>>>>> attributes where the grids are defined.
>>>>>>
>>>>>> Give that a look, and let me know if you continue to experience
>>>> problems.
>>>>>>
>>>>>> Thanks,
>>>>>> John
>>>>>>
>>>>>>
>>>>>>>
>>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
>>>>>>> Transaction: Ticket created by czoellic at iastate.edu
>>>>>>> Queue: met_help
>>>>>>> Subject: Lat/Lon Grid in MODE
>>>>>>> Owner: Nobody
>>>>>>> Requestors: czoellic at iastate.edu
>>>>>>> Status: new
>>>>>>> Ticket <URL:
>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
>>>>>>>
>>>>>>>
>>>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
>>>> output
>>>>>>> into MODE. Now I am trying to push an observational dataset
in from
>>>>>>> University of Washington. The data is on a lat/lon grid,
which I was
>>>>>>> reading was supposed to be acceptable for MET 2.0. MODE is
>> definitely
>>>>>> not
>>>>>>> reading in the grid correctly, and if I set the metadata to
say
>>>> "Lat/Lon"
>>>>>>> for the projection attribute, it says it is not supported. I
was
>>>>>>> wondering
>>>>>>> if there was something I'm missing to get MODE to read in a
Lat/Lon
>>>> grid?
>>>>>>>
>>>>>>> Thanks for your help again!
>>>>>>> Casey Zoellick
>>>>>>> czoellic at iastate.edu
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Mon Nov 15 14:02:54 2010
John,
I made the changes to the raw thresholds and the FillValue, and I
still seem
to have the same problem. I guess I don't understand why MODE won't
plot
the precip from one dataset, but will for my forecast dataset when the
process I'm using for both to convert the files is the same.
Thanks,
Casey
On Thu, Nov 11, 2010 at 3:25 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Casey,
>
> I think I know what's causing the problem. Unfortunately, our
support for
> NetCDF is limited in METv2.0, and we're not handling your choice of
fill
> value appropriately:
> APCP_24:_FillValue = 1.e+20f ;
>
> The PCP-Combine tool uses a fill value of "-9999", as shown below,
and I
> believe the MET code is expecting that.
> APCP_12:_FillValue = -9999.f ;
>
> Please try resetting the fill values from "1.0+20f" to "-9999" and
try
> rerunning. Our support for NetCDF is limited, and we're slowly
working
> toward adding more general support. Also, in your config
> file, I'd suggest setting the raw thresholds to 0.0:
> fcst_raw_thresh = "ge0.0";
> obs_raw_thresh = "ge0.0";
>
> The convolution radius and threshold (fcst_conv_rad and
fcst_conv_thresh)
> are what's used to define objects. Most users should not need to
use the
> raw thresholds unless there's significant
> differences in how their forecast and observation fields are
defined.
>
> Thanks,
> John
>
> On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> > John,
> >
> > I accidently copied the 1982_06_01.nc file twice and it got
renamed as
> > temp_1.nc. The 1982_06_01.nc file should be correct and MODE
shouldn't
> be
> > able to read it. I just put a file called "new_temp_1.nc" which
should
> be
> > the correct converted file. Sorry for the confusion.
> >
> > -Casey
> >
> >
> > On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Casey,
> >>
> >> I suspect that you made a mistake when posting the data to our
ftp site.
> >> The file size and structure of the "temp_1.nc" and
"1982_06_01.nc"
> files
> >> are the same, and neither can be read by MODE.
> >>
> >> John
> >>
> >> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >>>
> >>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>
> >>> John,
> >>>
> >>> There should be only one timestep in each file. I posted a
NetCDF file
> >>> before I converted it and one after I'm done converting it as
well as
> the
> >>> MODE config file on the ftp site. I should note that part of
the
> >> conversion
> >>> process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
> >>> grid.
> >>>
> >>> Thanks again,
> >>> Casey
> >>>
> >>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway}
> <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Casey,
> >>>>
> >>>> Do you have multiple output times in a single NetCDF file?
> >>>>
> >>>> If so, you may just be retrieving the first time from that file
each
> >> time
> >>>> you run it. It'd probably be easiest for you to just send me
your
> >> NetCDF
> >>>> file(s) and the MODE configuration file(s) you're
> >>>> using. I'll run it here and try to figure out what's going on.
If
> you
> >>>> have multiple times in your NetCDF file, you'll need to vary
the
> string
> >> that
> >>>> defines which 2D slice you want to extract from
> >>>> that NetCDF file each time you run MODE.
> >>>>
> >>>> You can post data to our anonymous ftp site using these
instructions:
> >>>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>>>
> >>>> Thanks,
> >>>> John
> >>>>
> >>>>
> >>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>
> >>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994
>
> >>>>>
> >>>>> John,
> >>>>>
> >>>>> I adjusted my global attributes, and now MODE does indeed read
in the
> >>>> LatLon
> >>>>> grid and the domain I'm using is what MODE recognizes as well.
> >> However,
> >>>> it
> >>>>> keeps plotting the precip field the same for everyday I run it
over a
> >>>> small
> >>>>> bullet of abnormally high values within my domain. The
results are
> the
> >>>> same
> >>>>> for everyday. I have checked, and I know the files I'm
feeding to
> MODE
> >>>> are
> >>>>> indeed different for each day and that the values for the
precip are
> >>>>> reasonable in the file. I guess I don't know why it reads in
my
> >> domain,
> >>>> but
> >>>>> not the precip values.
> >>>>>
> >>>>> Thanks,
> >>>>> Casey
> >>>>>
> >>>>>
> >>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway}
> <
> >>>>> met_help at ucar.edu> wrote:
> >>>>>
> >>>>>> Casey,
> >>>>>>
> >>>>>> I've attached two sample NetCDF files, one that's output from
> METv2.0
> >>>> and
> >>>>>> one that's output from METv3.0. Please choose the example to
use
> >> based
> >>>> on
> >>>>>> the version of MET that you're using.
> >>>>>>
> >>>>>> Specifically, I'm guessing you should be specifying the
projection
> as
> >>>>>> "LatLon" instead of "Lat/Lon".
> >>>>>>
> >>>>>> Please run "ncdump -h" on these NetCDF files and take a look
at the
> >>>> global
> >>>>>> attributes where the grids are defined.
> >>>>>>
> >>>>>> Give that a look, and let me know if you continue to
experience
> >>>> problems.
> >>>>>>
> >>>>>> Thanks,
> >>>>>> John
> >>>>>>
> >>>>>>
> >>>>>>>
> >>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> >>>>>>> Transaction: Ticket created by czoellic at iastate.edu
> >>>>>>> Queue: met_help
> >>>>>>> Subject: Lat/Lon Grid in MODE
> >>>>>>> Owner: Nobody
> >>>>>>> Requestors: czoellic at iastate.edu
> >>>>>>> Status: new
> >>>>>>> Ticket <URL:
> >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
> >>>>>>>
> >>>>>>>
> >>>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
> >>>> output
> >>>>>>> into MODE. Now I am trying to push an observational dataset
in
> from
> >>>>>>> University of Washington. The data is on a lat/lon grid,
which I
> was
> >>>>>>> reading was supposed to be acceptable for MET 2.0. MODE is
> >> definitely
> >>>>>> not
> >>>>>>> reading in the grid correctly, and if I set the metadata to
say
> >>>> "Lat/Lon"
> >>>>>>> for the projection attribute, it says it is not supported.
I was
> >>>>>>> wondering
> >>>>>>> if there was something I'm missing to get MODE to read in a
Lat/Lon
> >>>> grid?
> >>>>>>>
> >>>>>>> Thanks for your help again!
> >>>>>>> Casey Zoellick
> >>>>>>> czoellic at iastate.edu
> >>>>>>>
> >>>>>>
> >>>>>>
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #41994] Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Mon Nov 15 14:04:50 2010
Casey,
Please send me the forecast and observation files you're using, along
with your current MODE config file. I'll run it here and see if I can
figure it out.
Follow the instructions here:
http://www.dtcenter.org/met/users/support/met_help.php#ftp
John
On 11/15/2010 02:02 PM, RAL HelpDesk {for Casey Zoellick} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
> John,
>
> I made the changes to the raw thresholds and the FillValue, and I
still seem
> to have the same problem. I guess I don't understand why MODE won't
plot
> the precip from one dataset, but will for my forecast dataset when
the
> process I'm using for both to convert the files is the same.
>
> Thanks,
> Casey
>
>
> On Thu, Nov 11, 2010 at 3:25 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Casey,
>>
>> I think I know what's causing the problem. Unfortunately, our
support for
>> NetCDF is limited in METv2.0, and we're not handling your choice of
fill
>> value appropriately:
>> APCP_24:_FillValue = 1.e+20f ;
>>
>> The PCP-Combine tool uses a fill value of "-9999", as shown below,
and I
>> believe the MET code is expecting that.
>> APCP_12:_FillValue = -9999.f ;
>>
>> Please try resetting the fill values from "1.0+20f" to "-9999" and
try
>> rerunning. Our support for NetCDF is limited, and we're slowly
working
>> toward adding more general support. Also, in your config
>> file, I'd suggest setting the raw thresholds to 0.0:
>> fcst_raw_thresh = "ge0.0";
>> obs_raw_thresh = "ge0.0";
>>
>> The convolution radius and threshold (fcst_conv_rad and
fcst_conv_thresh)
>> are what's used to define objects. Most users should not need to
use the
>> raw thresholds unless there's significant
>> differences in how their forecast and observation fields are
defined.
>>
>> Thanks,
>> John
>>
>> On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>
>>> John,
>>>
>>> I accidently copied the 1982_06_01.nc file twice and it got
renamed as
>>> temp_1.nc. The 1982_06_01.nc file should be correct and MODE
shouldn't
>> be
>>> able to read it. I just put a file called "new_temp_1.nc" which
should
>> be
>>> the correct converted file. Sorry for the confusion.
>>>
>>> -Casey
>>>
>>>
>>> On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Casey,
>>>>
>>>> I suspect that you made a mistake when posting the data to our
ftp site.
>>>> The file size and structure of the "temp_1.nc" and
"1982_06_01.nc"
>> files
>>>> are the same, and neither can be read by MODE.
>>>>
>>>> John
>>>>
>>>> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>>>
>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>>>
>>>>> John,
>>>>>
>>>>> There should be only one timestep in each file. I posted a
NetCDF file
>>>>> before I converted it and one after I'm done converting it as
well as
>> the
>>>>> MODE config file on the ftp site. I should note that part of
the
>>>> conversion
>>>>> process is sampling data on a 1/8 degree grid to put it on a 1/2
degree
>>>>> grid.
>>>>>
>>>>> Thanks again,
>>>>> Casey
>>>>>
>>>>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John Halley
Gotway}
>> <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Casey,
>>>>>>
>>>>>> Do you have multiple output times in a single NetCDF file?
>>>>>>
>>>>>> If so, you may just be retrieving the first time from that file
each
>>>> time
>>>>>> you run it. It'd probably be easiest for you to just send me
your
>>>> NetCDF
>>>>>> file(s) and the MODE configuration file(s) you're
>>>>>> using. I'll run it here and try to figure out what's going on.
If
>> you
>>>>>> have multiple times in your NetCDF file, you'll need to vary
the
>> string
>>>> that
>>>>>> defines which 2D slice you want to extract from
>>>>>> that NetCDF file each time you run MODE.
>>>>>>
>>>>>> You can post data to our anonymous ftp site using these
instructions:
>>>>>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>>>>>
>>>>>> Thanks,
>>>>>> John
>>>>>>
>>>>>>
>>>>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
>>>>>>>
>>>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994
>
>>>>>>>
>>>>>>> John,
>>>>>>>
>>>>>>> I adjusted my global attributes, and now MODE does indeed read
in the
>>>>>> LatLon
>>>>>>> grid and the domain I'm using is what MODE recognizes as well.
>>>> However,
>>>>>> it
>>>>>>> keeps plotting the precip field the same for everyday I run it
over a
>>>>>> small
>>>>>>> bullet of abnormally high values within my domain. The
results are
>> the
>>>>>> same
>>>>>>> for everyday. I have checked, and I know the files I'm
feeding to
>> MODE
>>>>>> are
>>>>>>> indeed different for each day and that the values for the
precip are
>>>>>>> reasonable in the file. I guess I don't know why it reads in
my
>>>> domain,
>>>>>> but
>>>>>>> not the precip values.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Casey
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John Halley
Gotway}
>> <
>>>>>>> met_help at ucar.edu> wrote:
>>>>>>>
>>>>>>>> Casey,
>>>>>>>>
>>>>>>>> I've attached two sample NetCDF files, one that's output from
>> METv2.0
>>>>>> and
>>>>>>>> one that's output from METv3.0. Please choose the example to
use
>>>> based
>>>>>> on
>>>>>>>> the version of MET that you're using.
>>>>>>>>
>>>>>>>> Specifically, I'm guessing you should be specifying the
projection
>> as
>>>>>>>> "LatLon" instead of "Lat/Lon".
>>>>>>>>
>>>>>>>> Please run "ncdump -h" on these NetCDF files and take a look
at the
>>>>>> global
>>>>>>>> attributes where the grids are defined.
>>>>>>>>
>>>>>>>> Give that a look, and let me know if you continue to
experience
>>>>>> problems.
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> John
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
>>>>>>>>> Transaction: Ticket created by czoellic at iastate.edu
>>>>>>>>> Queue: met_help
>>>>>>>>> Subject: Lat/Lon Grid in MODE
>>>>>>>>> Owner: Nobody
>>>>>>>>> Requestors: czoellic at iastate.edu
>>>>>>>>> Status: new
>>>>>>>>> Ticket <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
>>>>>> output
>>>>>>>>> into MODE. Now I am trying to push an observational dataset
in
>> from
>>>>>>>>> University of Washington. The data is on a lat/lon grid,
which I
>> was
>>>>>>>>> reading was supposed to be acceptable for MET 2.0. MODE is
>>>> definitely
>>>>>>>> not
>>>>>>>>> reading in the grid correctly, and if I set the metadata to
say
>>>>>> "Lat/Lon"
>>>>>>>>> for the projection attribute, it says it is not supported.
I was
>>>>>>>>> wondering
>>>>>>>>> if there was something I'm missing to get MODE to read in a
Lat/Lon
>>>>>> grid?
>>>>>>>>>
>>>>>>>>> Thanks for your help again!
>>>>>>>>> Casey Zoellick
>>>>>>>>> czoellic at iastate.edu
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Tue Nov 16 01:17:54 2010
John,
I sent the most recent file I use for forecast/observation (
example_fcst_obs.nc) and the most recent configfile. When I run MODE
I pass
into it the same file as the forecast and the observed since I'm only
interested in the objects and not the matching.
Thanks,
Casey
On Mon, Nov 15, 2010 at 3:04 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Casey,
>
> Please send me the forecast and observation files you're using,
along with
> your current MODE config file. I'll run it here and see if I can
figure it
> out.
>
> Follow the instructions here:
> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>
> John
>
> On 11/15/2010 02:02 PM, RAL HelpDesk {for Casey Zoellick} wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> > John,
> >
> > I made the changes to the raw thresholds and the FillValue, and I
still
> seem
> > to have the same problem. I guess I don't understand why MODE
won't plot
> > the precip from one dataset, but will for my forecast dataset when
the
> > process I'm using for both to convert the files is the same.
> >
> > Thanks,
> > Casey
> >
> >
> > On Thu, Nov 11, 2010 at 3:25 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Casey,
> >>
> >> I think I know what's causing the problem. Unfortunately, our
support
> for
> >> NetCDF is limited in METv2.0, and we're not handling your choice
of fill
> >> value appropriately:
> >> APCP_24:_FillValue = 1.e+20f ;
> >>
> >> The PCP-Combine tool uses a fill value of "-9999", as shown
below, and I
> >> believe the MET code is expecting that.
> >> APCP_12:_FillValue = -9999.f ;
> >>
> >> Please try resetting the fill values from "1.0+20f" to "-9999"
and try
> >> rerunning. Our support for NetCDF is limited, and we're slowly
working
> >> toward adding more general support. Also, in your config
> >> file, I'd suggest setting the raw thresholds to 0.0:
> >> fcst_raw_thresh = "ge0.0";
> >> obs_raw_thresh = "ge0.0";
> >>
> >> The convolution radius and threshold (fcst_conv_rad and
> fcst_conv_thresh)
> >> are what's used to define objects. Most users should not need to
use
> the
> >> raw thresholds unless there's significant
> >> differences in how their forecast and observation fields are
defined.
> >>
> >> Thanks,
> >> John
> >>
> >> On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >>>
> >>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>
> >>> John,
> >>>
> >>> I accidently copied the 1982_06_01.nc file twice and it got
renamed as
> >>> temp_1.nc. The 1982_06_01.nc file should be correct and MODE
> shouldn't
> >> be
> >>> able to read it. I just put a file called "new_temp_1.nc" which
> should
> >> be
> >>> the correct converted file. Sorry for the confusion.
> >>>
> >>> -Casey
> >>>
> >>>
> >>> On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway}
> <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Casey,
> >>>>
> >>>> I suspect that you made a mistake when posting the data to our
ftp
> site.
> >>>> The file size and structure of the "temp_1.nc" and
"1982_06_01.nc"
> >> files
> >>>> are the same, and neither can be read by MODE.
> >>>>
> >>>> John
> >>>>
> >>>> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>
> >>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994
>
> >>>>>
> >>>>> John,
> >>>>>
> >>>>> There should be only one timestep in each file. I posted a
NetCDF
> file
> >>>>> before I converted it and one after I'm done converting it as
well as
> >> the
> >>>>> MODE config file on the ftp site. I should note that part of
the
> >>>> conversion
> >>>>> process is sampling data on a 1/8 degree grid to put it on a
1/2
> degree
> >>>>> grid.
> >>>>>
> >>>>> Thanks again,
> >>>>> Casey
> >>>>>
> >>>>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John
Halley
> Gotway}
> >> <
> >>>>> met_help at ucar.edu> wrote:
> >>>>>
> >>>>>> Casey,
> >>>>>>
> >>>>>> Do you have multiple output times in a single NetCDF file?
> >>>>>>
> >>>>>> If so, you may just be retrieving the first time from that
file each
> >>>> time
> >>>>>> you run it. It'd probably be easiest for you to just send me
your
> >>>> NetCDF
> >>>>>> file(s) and the MODE configuration file(s) you're
> >>>>>> using. I'll run it here and try to figure out what's going
on. If
> >> you
> >>>>>> have multiple times in your NetCDF file, you'll need to vary
the
> >> string
> >>>> that
> >>>>>> defines which 2D slice you want to extract from
> >>>>>> that NetCDF file each time you run MODE.
> >>>>>>
> >>>>>> You can post data to our anonymous ftp site using these
> instructions:
> >>>>>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>>>>>
> >>>>>> Thanks,
> >>>>>> John
> >>>>>>
> >>>>>>
> >>>>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>>>
> >>>>>>> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>>>>>
> >>>>>>> John,
> >>>>>>>
> >>>>>>> I adjusted my global attributes, and now MODE does indeed
read in
> the
> >>>>>> LatLon
> >>>>>>> grid and the domain I'm using is what MODE recognizes as
well.
> >>>> However,
> >>>>>> it
> >>>>>>> keeps plotting the precip field the same for everyday I run
it over
> a
> >>>>>> small
> >>>>>>> bullet of abnormally high values within my domain. The
results are
> >> the
> >>>>>> same
> >>>>>>> for everyday. I have checked, and I know the files I'm
feeding to
> >> MODE
> >>>>>> are
> >>>>>>> indeed different for each day and that the values for the
precip
> are
> >>>>>>> reasonable in the file. I guess I don't know why it reads
in my
> >>>> domain,
> >>>>>> but
> >>>>>>> not the precip values.
> >>>>>>>
> >>>>>>> Thanks,
> >>>>>>> Casey
> >>>>>>>
> >>>>>>>
> >>>>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John
Halley
> Gotway}
> >> <
> >>>>>>> met_help at ucar.edu> wrote:
> >>>>>>>
> >>>>>>>> Casey,
> >>>>>>>>
> >>>>>>>> I've attached two sample NetCDF files, one that's output
from
> >> METv2.0
> >>>>>> and
> >>>>>>>> one that's output from METv3.0. Please choose the example
to use
> >>>> based
> >>>>>> on
> >>>>>>>> the version of MET that you're using.
> >>>>>>>>
> >>>>>>>> Specifically, I'm guessing you should be specifying the
projection
> >> as
> >>>>>>>> "LatLon" instead of "Lat/Lon".
> >>>>>>>>
> >>>>>>>> Please run "ncdump -h" on these NetCDF files and take a
look at
> the
> >>>>>> global
> >>>>>>>> attributes where the grids are defined.
> >>>>>>>>
> >>>>>>>> Give that a look, and let me know if you continue to
experience
> >>>>>> problems.
> >>>>>>>>
> >>>>>>>> Thanks,
> >>>>>>>> John
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> >>>>>>>>> Transaction: Ticket created by czoellic at iastate.edu
> >>>>>>>>> Queue: met_help
> >>>>>>>>> Subject: Lat/Lon Grid in MODE
> >>>>>>>>> Owner: Nobody
> >>>>>>>>> Requestors: czoellic at iastate.edu
> >>>>>>>>> Status: new
> >>>>>>>>> Ticket <URL:
> >>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
> >>>>>> output
> >>>>>>>>> into MODE. Now I am trying to push an observational
dataset in
> >> from
> >>>>>>>>> University of Washington. The data is on a lat/lon grid,
which I
> >> was
> >>>>>>>>> reading was supposed to be acceptable for MET 2.0. MODE
is
> >>>> definitely
> >>>>>>>> not
> >>>>>>>>> reading in the grid correctly, and if I set the metadata
to say
> >>>>>> "Lat/Lon"
> >>>>>>>>> for the projection attribute, it says it is not supported.
I was
> >>>>>>>>> wondering
> >>>>>>>>> if there was something I'm missing to get MODE to read in
a
> Lat/Lon
> >>>>>> grid?
> >>>>>>>>>
> >>>>>>>>> Thanks for your help again!
> >>>>>>>>> Casey Zoellick
> >>>>>>>>> czoellic at iastate.edu
> >>>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>>
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: John Halley Gotway
Time: Tue Nov 16 09:17:05 2010
Casey,
It looks like there were still a handful of 1.0e20 values hanging
around in that file you sent. There were 9 to be exact. I created a
copy of your file, modified it to reset those values from 1.0e20
to -9999, and reran it through MODE. The resulting image is attached
as well as the modified NetCDF file.
Here's what's going on. When the MET tools read a gridded data field,
they try to be a little to fancy. Rather than storing each data value
as a double, we're storing it as an integer with a scale
and offset value. The scale and offset value are determined by the
min/max data values in the field. Since MODE found data values from 0
all the way up to 1.0e20, it kind of blew that logic and
screwed up how the values were stored internally - leading to the
funny plots you saw. For the next release of MET, we'll be doing an
overhaul, simplification, and cleanup of the library code. And
we'll modify this logic to simply store the raw data values as is,
rather than using a scale and offset.
For now, you just need to be sure to use "-9999" as your bad data
value, and make sure you don't have any 1.0e20 value hanging around!
Thanks,
John
On 11/16/2010 01:17 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>
> John,
>
> I sent the most recent file I use for forecast/observation (
> example_fcst_obs.nc) and the most recent configfile. When I run
MODE I pass
> into it the same file as the forecast and the observed since I'm
only
> interested in the objects and not the matching.
>
> Thanks,
> Casey
>
> On Mon, Nov 15, 2010 at 3:04 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Casey,
>>
>> Please send me the forecast and observation files you're using,
along with
>> your current MODE config file. I'll run it here and see if I can
figure it
>> out.
>>
>> Follow the instructions here:
>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>
>> John
>>
>> On 11/15/2010 02:02 PM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>
>>> John,
>>>
>>> I made the changes to the raw thresholds and the FillValue, and I
still
>> seem
>>> to have the same problem. I guess I don't understand why MODE
won't plot
>>> the precip from one dataset, but will for my forecast dataset when
the
>>> process I'm using for both to convert the files is the same.
>>>
>>> Thanks,
>>> Casey
>>>
>>>
>>> On Thu, Nov 11, 2010 at 3:25 PM, RAL HelpDesk {for John Halley
Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Casey,
>>>>
>>>> I think I know what's causing the problem. Unfortunately, our
support
>> for
>>>> NetCDF is limited in METv2.0, and we're not handling your choice
of fill
>>>> value appropriately:
>>>> APCP_24:_FillValue = 1.e+20f ;
>>>>
>>>> The PCP-Combine tool uses a fill value of "-9999", as shown
below, and I
>>>> believe the MET code is expecting that.
>>>> APCP_12:_FillValue = -9999.f ;
>>>>
>>>> Please try resetting the fill values from "1.0+20f" to "-9999"
and try
>>>> rerunning. Our support for NetCDF is limited, and we're slowly
working
>>>> toward adding more general support. Also, in your config
>>>> file, I'd suggest setting the raw thresholds to 0.0:
>>>> fcst_raw_thresh = "ge0.0";
>>>> obs_raw_thresh = "ge0.0";
>>>>
>>>> The convolution radius and threshold (fcst_conv_rad and
>> fcst_conv_thresh)
>>>> are what's used to define objects. Most users should not need to
use
>> the
>>>> raw thresholds unless there's significant
>>>> differences in how their forecast and observation fields are
defined.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick} wrote:
>>>>>
>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>>>
>>>>> John,
>>>>>
>>>>> I accidently copied the 1982_06_01.nc file twice and it got
renamed as
>>>>> temp_1.nc. The 1982_06_01.nc file should be correct and MODE
>> shouldn't
>>>> be
>>>>> able to read it. I just put a file called "new_temp_1.nc" which
>> should
>>>> be
>>>>> the correct converted file. Sorry for the confusion.
>>>>>
>>>>> -Casey
>>>>>
>>>>>
>>>>> On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John Halley
Gotway}
>> <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Casey,
>>>>>>
>>>>>> I suspect that you made a mistake when posting the data to our
ftp
>> site.
>>>>>> The file size and structure of the "temp_1.nc" and
"1982_06_01.nc"
>>>> files
>>>>>> are the same, and neither can be read by MODE.
>>>>>>
>>>>>> John
>>>>>>
>>>>>> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
>>>>>>>
>>>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994
>
>>>>>>>
>>>>>>> John,
>>>>>>>
>>>>>>> There should be only one timestep in each file. I posted a
NetCDF
>> file
>>>>>>> before I converted it and one after I'm done converting it as
well as
>>>> the
>>>>>>> MODE config file on the ftp site. I should note that part of
the
>>>>>> conversion
>>>>>>> process is sampling data on a 1/8 degree grid to put it on a
1/2
>> degree
>>>>>>> grid.
>>>>>>>
>>>>>>> Thanks again,
>>>>>>> Casey
>>>>>>>
>>>>>>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John
Halley
>> Gotway}
>>>> <
>>>>>>> met_help at ucar.edu> wrote:
>>>>>>>
>>>>>>>> Casey,
>>>>>>>>
>>>>>>>> Do you have multiple output times in a single NetCDF file?
>>>>>>>>
>>>>>>>> If so, you may just be retrieving the first time from that
file each
>>>>>> time
>>>>>>>> you run it. It'd probably be easiest for you to just send me
your
>>>>>> NetCDF
>>>>>>>> file(s) and the MODE configuration file(s) you're
>>>>>>>> using. I'll run it here and try to figure out what's going
on. If
>>>> you
>>>>>>>> have multiple times in your NetCDF file, you'll need to vary
the
>>>> string
>>>>>> that
>>>>>>>> defines which 2D slice you want to extract from
>>>>>>>> that NetCDF file each time you run MODE.
>>>>>>>>
>>>>>>>> You can post data to our anonymous ftp site using these
>> instructions:
>>>>>>>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> John
>>>>>>>>
>>>>>>>>
>>>>>>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
>>>>>>>>>
>>>>>>>>> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
>>>>>>>>>
>>>>>>>>> John,
>>>>>>>>>
>>>>>>>>> I adjusted my global attributes, and now MODE does indeed
read in
>> the
>>>>>>>> LatLon
>>>>>>>>> grid and the domain I'm using is what MODE recognizes as
well.
>>>>>> However,
>>>>>>>> it
>>>>>>>>> keeps plotting the precip field the same for everyday I run
it over
>> a
>>>>>>>> small
>>>>>>>>> bullet of abnormally high values within my domain. The
results are
>>>> the
>>>>>>>> same
>>>>>>>>> for everyday. I have checked, and I know the files I'm
feeding to
>>>> MODE
>>>>>>>> are
>>>>>>>>> indeed different for each day and that the values for the
precip
>> are
>>>>>>>>> reasonable in the file. I guess I don't know why it reads
in my
>>>>>> domain,
>>>>>>>> but
>>>>>>>>> not the precip values.
>>>>>>>>>
>>>>>>>>> Thanks,
>>>>>>>>> Casey
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John
Halley
>> Gotway}
>>>> <
>>>>>>>>> met_help at ucar.edu> wrote:
>>>>>>>>>
>>>>>>>>>> Casey,
>>>>>>>>>>
>>>>>>>>>> I've attached two sample NetCDF files, one that's output
from
>>>> METv2.0
>>>>>>>> and
>>>>>>>>>> one that's output from METv3.0. Please choose the example
to use
>>>>>> based
>>>>>>>> on
>>>>>>>>>> the version of MET that you're using.
>>>>>>>>>>
>>>>>>>>>> Specifically, I'm guessing you should be specifying the
projection
>>>> as
>>>>>>>>>> "LatLon" instead of "Lat/Lon".
>>>>>>>>>>
>>>>>>>>>> Please run "ncdump -h" on these NetCDF files and take a
look at
>> the
>>>>>>>> global
>>>>>>>>>> attributes where the grids are defined.
>>>>>>>>>>
>>>>>>>>>> Give that a look, and let me know if you continue to
experience
>>>>>>>> problems.
>>>>>>>>>>
>>>>>>>>>> Thanks,
>>>>>>>>>> John
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
>>>>>>>>>>> Transaction: Ticket created by czoellic at iastate.edu
>>>>>>>>>>> Queue: met_help
>>>>>>>>>>> Subject: Lat/Lon Grid in MODE
>>>>>>>>>>> Owner: Nobody
>>>>>>>>>>> Requestors: czoellic at iastate.edu
>>>>>>>>>>> Status: new
>>>>>>>>>>> Ticket <URL:
>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> I contacted you previously regarding pushing NARCCAP model
NetCDF
>>>>>>>> output
>>>>>>>>>>> into MODE. Now I am trying to push an observational
dataset in
>>>> from
>>>>>>>>>>> University of Washington. The data is on a lat/lon grid,
which I
>>>> was
>>>>>>>>>>> reading was supposed to be acceptable for MET 2.0. MODE
is
>>>>>> definitely
>>>>>>>>>> not
>>>>>>>>>>> reading in the grid correctly, and if I set the metadata
to say
>>>>>>>> "Lat/Lon"
>>>>>>>>>>> for the projection attribute, it says it is not supported.
I was
>>>>>>>>>>> wondering
>>>>>>>>>>> if there was something I'm missing to get MODE to read in
a
>> Lat/Lon
>>>>>>>> grid?
>>>>>>>>>>>
>>>>>>>>>>> Thanks for your help again!
>>>>>>>>>>> Casey Zoellick
>>>>>>>>>>> czoellic at iastate.edu
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Lat/Lon Grid in MODE
From: Casey Zoellick
Time: Wed Nov 17 11:53:11 2010
John,
Sounds good! That would make sense since that's how MODE determines
the
scale. Right now I'm just trying to figure out how to replace those
bad
1.0e20 values and I should be running ok.
Thanks for all your help!
Casey
On Tue, Nov 16, 2010 at 10:17 AM, RAL HelpDesk {for John Halley
Gotway} <
met_help at ucar.edu> wrote:
> Casey,
>
> It looks like there were still a handful of 1.0e20 values hanging
around in
> that file you sent. There were 9 to be exact. I created a copy of
your
> file, modified it to reset those values from 1.0e20
> to -9999, and reran it through MODE. The resulting image is
attached as
> well as the modified NetCDF file.
>
> Here's what's going on. When the MET tools read a gridded data
field, they
> try to be a little to fancy. Rather than storing each data value as
a
> double, we're storing it as an integer with a scale
> and offset value. The scale and offset value are determined by the
min/max
> data values in the field. Since MODE found data values from 0 all
the way
> up to 1.0e20, it kind of blew that logic and
> screwed up how the values were stored internally - leading to the
funny
> plots you saw. For the next release of MET, we'll be doing an
overhaul,
> simplification, and cleanup of the library code. And
> we'll modify this logic to simply store the raw data values as is,
rather
> than using a scale and offset.
>
> For now, you just need to be sure to use "-9999" as your bad data
value,
> and make sure you don't have any 1.0e20 value hanging around!
>
> Thanks,
> John
>
> On 11/16/2010 01:17 AM, RAL HelpDesk {for Casey Zoellick} wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >
> > John,
> >
> > I sent the most recent file I use for forecast/observation (
> > example_fcst_obs.nc) and the most recent configfile. When I run
MODE I
> pass
> > into it the same file as the forecast and the observed since I'm
only
> > interested in the objects and not the matching.
> >
> > Thanks,
> > Casey
> >
> > On Mon, Nov 15, 2010 at 3:04 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Casey,
> >>
> >> Please send me the forecast and observation files you're using,
along
> with
> >> your current MODE config file. I'll run it here and see if I can
figure
> it
> >> out.
> >>
> >> Follow the instructions here:
> >> http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>
> >> John
> >>
> >> On 11/15/2010 02:02 PM, RAL HelpDesk {for Casey Zoellick} wrote:
> >>>
> >>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>
> >>> John,
> >>>
> >>> I made the changes to the raw thresholds and the FillValue, and
I still
> >> seem
> >>> to have the same problem. I guess I don't understand why MODE
won't
> plot
> >>> the precip from one dataset, but will for my forecast dataset
when the
> >>> process I'm using for both to convert the files is the same.
> >>>
> >>> Thanks,
> >>> Casey
> >>>
> >>>
> >>> On Thu, Nov 11, 2010 at 3:25 PM, RAL HelpDesk {for John Halley
Gotway}
> <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Casey,
> >>>>
> >>>> I think I know what's causing the problem. Unfortunately, our
support
> >> for
> >>>> NetCDF is limited in METv2.0, and we're not handling your
choice of
> fill
> >>>> value appropriately:
> >>>> APCP_24:_FillValue = 1.e+20f ;
> >>>>
> >>>> The PCP-Combine tool uses a fill value of "-9999", as shown
below, and
> I
> >>>> believe the MET code is expecting that.
> >>>> APCP_12:_FillValue = -9999.f ;
> >>>>
> >>>> Please try resetting the fill values from "1.0+20f" to "-9999"
and try
> >>>> rerunning. Our support for NetCDF is limited, and we're slowly
> working
> >>>> toward adding more general support. Also, in your config
> >>>> file, I'd suggest setting the raw thresholds to 0.0:
> >>>> fcst_raw_thresh = "ge0.0";
> >>>> obs_raw_thresh = "ge0.0";
> >>>>
> >>>> The convolution radius and threshold (fcst_conv_rad and
> >> fcst_conv_thresh)
> >>>> are what's used to define objects. Most users should not need
to use
> >> the
> >>>> raw thresholds unless there's significant
> >>>> differences in how their forecast and observation fields are
defined.
> >>>>
> >>>> Thanks,
> >>>> John
> >>>>
> >>>> On 11/11/2010 11:50 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>
> >>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994
>
> >>>>>
> >>>>> John,
> >>>>>
> >>>>> I accidently copied the 1982_06_01.nc file twice and it got
renamed
> as
> >>>>> temp_1.nc. The 1982_06_01.nc file should be correct and MODE
> >> shouldn't
> >>>> be
> >>>>> able to read it. I just put a file called "new_temp_1.nc"
which
> >> should
> >>>> be
> >>>>> the correct converted file. Sorry for the confusion.
> >>>>>
> >>>>> -Casey
> >>>>>
> >>>>>
> >>>>> On Thu, Nov 11, 2010 at 12:42 PM, RAL HelpDesk {for John
Halley
> Gotway}
> >> <
> >>>>> met_help at ucar.edu> wrote:
> >>>>>
> >>>>>> Casey,
> >>>>>>
> >>>>>> I suspect that you made a mistake when posting the data to
our ftp
> >> site.
> >>>>>> The file size and structure of the "temp_1.nc" and
"1982_06_01.nc"
> >>>> files
> >>>>>> are the same, and neither can be read by MODE.
> >>>>>>
> >>>>>> John
> >>>>>>
> >>>>>> On 11/11/2010 11:35 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>>>
> >>>>>>> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>>>>>
> >>>>>>> John,
> >>>>>>>
> >>>>>>> There should be only one timestep in each file. I posted a
NetCDF
> >> file
> >>>>>>> before I converted it and one after I'm done converting it
as well
> as
> >>>> the
> >>>>>>> MODE config file on the ftp site. I should note that part
of the
> >>>>>> conversion
> >>>>>>> process is sampling data on a 1/8 degree grid to put it on a
1/2
> >> degree
> >>>>>>> grid.
> >>>>>>>
> >>>>>>> Thanks again,
> >>>>>>> Casey
> >>>>>>>
> >>>>>>> On Thu, Nov 11, 2010 at 12:23 PM, RAL HelpDesk {for John
Halley
> >> Gotway}
> >>>> <
> >>>>>>> met_help at ucar.edu> wrote:
> >>>>>>>
> >>>>>>>> Casey,
> >>>>>>>>
> >>>>>>>> Do you have multiple output times in a single NetCDF file?
> >>>>>>>>
> >>>>>>>> If so, you may just be retrieving the first time from that
file
> each
> >>>>>> time
> >>>>>>>> you run it. It'd probably be easiest for you to just send
me your
> >>>>>> NetCDF
> >>>>>>>> file(s) and the MODE configuration file(s) you're
> >>>>>>>> using. I'll run it here and try to figure out what's going
on.
> If
> >>>> you
> >>>>>>>> have multiple times in your NetCDF file, you'll need to
vary the
> >>>> string
> >>>>>> that
> >>>>>>>> defines which 2D slice you want to extract from
> >>>>>>>> that NetCDF file each time you run MODE.
> >>>>>>>>
> >>>>>>>> You can post data to our anonymous ftp site using these
> >> instructions:
> >>>>>>>>
http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>>>>>>>
> >>>>>>>> Thanks,
> >>>>>>>> John
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> On 11/11/2010 11:14 AM, RAL HelpDesk {for Casey Zoellick}
wrote:
> >>>>>>>>>
> >>>>>>>>> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994 >
> >>>>>>>>>
> >>>>>>>>> John,
> >>>>>>>>>
> >>>>>>>>> I adjusted my global attributes, and now MODE does indeed
read in
> >> the
> >>>>>>>> LatLon
> >>>>>>>>> grid and the domain I'm using is what MODE recognizes as
well.
> >>>>>> However,
> >>>>>>>> it
> >>>>>>>>> keeps plotting the precip field the same for everyday I
run it
> over
> >> a
> >>>>>>>> small
> >>>>>>>>> bullet of abnormally high values within my domain. The
results
> are
> >>>> the
> >>>>>>>> same
> >>>>>>>>> for everyday. I have checked, and I know the files I'm
feeding
> to
> >>>> MODE
> >>>>>>>> are
> >>>>>>>>> indeed different for each day and that the values for the
precip
> >> are
> >>>>>>>>> reasonable in the file. I guess I don't know why it reads
in my
> >>>>>> domain,
> >>>>>>>> but
> >>>>>>>>> not the precip values.
> >>>>>>>>>
> >>>>>>>>> Thanks,
> >>>>>>>>> Casey
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> On Mon, Nov 8, 2010 at 4:57 PM, RAL HelpDesk {for John
Halley
> >> Gotway}
> >>>> <
> >>>>>>>>> met_help at ucar.edu> wrote:
> >>>>>>>>>
> >>>>>>>>>> Casey,
> >>>>>>>>>>
> >>>>>>>>>> I've attached two sample NetCDF files, one that's output
from
> >>>> METv2.0
> >>>>>>>> and
> >>>>>>>>>> one that's output from METv3.0. Please choose the
example to
> use
> >>>>>> based
> >>>>>>>> on
> >>>>>>>>>> the version of MET that you're using.
> >>>>>>>>>>
> >>>>>>>>>> Specifically, I'm guessing you should be specifying the
> projection
> >>>> as
> >>>>>>>>>> "LatLon" instead of "Lat/Lon".
> >>>>>>>>>>
> >>>>>>>>>> Please run "ncdump -h" on these NetCDF files and take a
look at
> >> the
> >>>>>>>> global
> >>>>>>>>>> attributes where the grids are defined.
> >>>>>>>>>>
> >>>>>>>>>> Give that a look, and let me know if you continue to
experience
> >>>>>>>> problems.
> >>>>>>>>>>
> >>>>>>>>>> Thanks,
> >>>>>>>>>> John
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> Mon Nov 08 13:58:41 2010: Request 41994 was acted upon.
> >>>>>>>>>>> Transaction: Ticket created by czoellic at iastate.edu
> >>>>>>>>>>> Queue: met_help
> >>>>>>>>>>> Subject: Lat/Lon Grid in MODE
> >>>>>>>>>>> Owner: Nobody
> >>>>>>>>>>> Requestors: czoellic at iastate.edu
> >>>>>>>>>>> Status: new
> >>>>>>>>>>> Ticket <URL:
> >>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=41994>
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> I contacted you previously regarding pushing NARCCAP
model
> NetCDF
> >>>>>>>> output
> >>>>>>>>>>> into MODE. Now I am trying to push an observational
dataset in
> >>>> from
> >>>>>>>>>>> University of Washington. The data is on a lat/lon
grid, which
> I
> >>>> was
> >>>>>>>>>>> reading was supposed to be acceptable for MET 2.0. MODE
is
> >>>>>> definitely
> >>>>>>>>>> not
> >>>>>>>>>>> reading in the grid correctly, and if I set the metadata
to say
> >>>>>>>> "Lat/Lon"
> >>>>>>>>>>> for the projection attribute, it says it is not
supported. I
> was
> >>>>>>>>>>> wondering
> >>>>>>>>>>> if there was something I'm missing to get MODE to read
in a
> >> Lat/Lon
> >>>>>>>> grid?
> >>>>>>>>>>>
> >>>>>>>>>>> Thanks for your help again!
> >>>>>>>>>>> Casey Zoellick
> >>>>>>>>>>> czoellic at iastate.edu
> >>>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>>
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
>
>
--
CASEY L. ZOELLICK, C/Maj, AFROTC
Mission Support Group Deputy Commander
Detachment 250
630-890-2911
czoellic at iastate.edu
------------------------------------------------
More information about the Met_help
mailing list