[Met_help] [rt.rap.ucar.edu #79394] History for help with file format for MODE v 5.2

John Halley Gotway via RT met_help at ucar.edu
Wed Feb 15 14:06:30 MST 2017


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

I am using MODE to verify precipitation forecasts. I have an observation
dataset (it's basically MRMS, but processed from the raw stuff) in netCDF
format, but MODE doesn't seem to like it when I use it for MODE.

[jdduda at schooner1 MODE]$ ./mode
/scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
/condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc MODEConfig
DEBUG 1: Default Config File:
/home/jdduda/tool/met/share/met/config/MODEConfig_default
DEBUG 1: Match Config File: MODEConfig
DEBUG 1: Merge Config File: MODEConfig
ERROR  :
ERROR  : Trouble reading observation file "/condo/map/jdduda/MRMS_obs/rain/
1hr_precip_20160525_000000.nc"
ERROR  :

Here is the dump for that file:

netcdf \1hr_precip_20160525_000000 {
dimensions:
        x = 1568 ;
        y = 1120 ;
        z = 1 ;
variables:
        float hsr1hr(z, y, x) ;
                hsr1hr:standard_name = "hsr1hr" ;
                hsr1hr:long_name = "1-h precipitation" ;
                hsr1hr:units = "mm" ;
}

It is my understanding that this may not be "CF compliant", whatever that
means. If not, what can I do to make it CF compliant, or is it possible I'm
mis-specifying the part of the config file where this is important. Here is
the config file part:

obs = {
   field = {
      name = "hsr1hr";
      level = "(*,*,*)";

   raw_thresh        = NA;
   conv_radius       = 50.0/grid_res; // in grid squares
   conv_thresh       = >=0.5;
   vld_thresh        = 0.5;
   area_thresh       = NA;
   inten_perc_value  = 100;
   inten_perc_thresh = NA;
   merge_thresh      = >=1.25;
   merge_flag        = ENGINE;

   }
}

I tried (0,*,*) and (1,*,*) for the level, but I got the same error for
each.

Thanks.

Jeff Duda

-- 
Jeff Duda
Post-doctoral research fellow
University of Oklahoma School of Meteorology


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

Subject: help with file format for MODE v 5.2
From: John Halley Gotway
Time: Wed Feb 08 16:41:46 2017

Jeff,

Yes, that file doesn't follow the NetCDF Climate Forecast conventions
which
are described in this document:
   http://cfconventions.org/latest.html

NetCDF is a very wide-open format, and it isn't really realistic for
software to be able to read them in general without agreeing to some
ground
rules... and the CF conventions are an attempt to establish them.  In
particular, MET is looking for a definition of the grid on which this
data
resides and timing information.  And there are CF-compliant ways of
specifying those.

Alternatively, you could make your NetCDF file look like the NetCDF
output
files that MET generates (i.e. like pcp_combine or gen_vx_mask).  MET
knows
how to read it's own output.  And you may find that simplistic format
easier to follow than the CF-conventions.

We do have some ideas about how to make this easier in the future.
Ideally
MET would be able to read data from your file as-is... but then
specify the
grid and timing information directly in the MODE config file.
Unfortunately, the software is not set up to do that yet.

Just let us know if you get stuck making your NetCDF file look like
the
output of pcp_combine.

Thanks,
John Halley Gotway
met_help at ucar.edu


On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <met_help at ucar.edu>
wrote:

>
> Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> Transaction: Ticket created by jeffduda319 at gmail.com
>        Queue: met_help
>      Subject: help with file format for MODE v 5.2
>        Owner: Nobody
>   Requestors: jeffduda319 at gmail.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
>
>
> I am using MODE to verify precipitation forecasts. I have an
observation
> dataset (it's basically MRMS, but processed from the raw stuff) in
netCDF
> format, but MODE doesn't seem to like it when I use it for MODE.
>
> [jdduda at schooner1 MODE]$ ./mode
> /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
MODEConfig
> DEBUG 1: Default Config File:
> /home/jdduda/tool/met/share/met/config/MODEConfig_default
> DEBUG 1: Match Config File: MODEConfig
> DEBUG 1: Merge Config File: MODEConfig
> ERROR  :
> ERROR  : Trouble reading observation file
"/condo/map/jdduda/MRMS_obs/
> rain/
> 1hr_precip_20160525_000000.nc"
> ERROR  :
>
> Here is the dump for that file:
>
> netcdf \1hr_precip_20160525_000000 {
> dimensions:
>         x = 1568 ;
>         y = 1120 ;
>         z = 1 ;
> variables:
>         float hsr1hr(z, y, x) ;
>                 hsr1hr:standard_name = "hsr1hr" ;
>                 hsr1hr:long_name = "1-h precipitation" ;
>                 hsr1hr:units = "mm" ;
> }
>
> It is my understanding that this may not be "CF compliant", whatever
that
> means. If not, what can I do to make it CF compliant, or is it
possible I'm
> mis-specifying the part of the config file where this is important.
Here is
> the config file part:
>
> obs = {
>    field = {
>       name = "hsr1hr";
>       level = "(*,*,*)";
>
>    raw_thresh        = NA;
>    conv_radius       = 50.0/grid_res; // in grid squares
>    conv_thresh       = >=0.5;
>    vld_thresh        = 0.5;
>    area_thresh       = NA;
>    inten_perc_value  = 100;
>    inten_perc_thresh = NA;
>    merge_thresh      = >=1.25;
>    merge_flag        = ENGINE;
>
>    }
> }
>
> I tried (0,*,*) and (1,*,*) for the level, but I got the same error
for
> each.
>
> Thanks.
>
> Jeff Duda
>
> --
> Jeff Duda
> Post-doctoral research fellow
> University of Oklahoma School of Meteorology
>
>

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: Jeff Duda
Time: Thu Feb 09 15:41:39 2017

Thanks for the reply, John. I made some changes to my system but am
still
having problems. I looked at Tables 5.1 and 5.2 in the User's Guide to
see
what the format from the output of pcp_combine would be, and I
reorganized
my observation files to match that format. I think i did it, but I get
the
same error.

Here is the ncdump for my obs file now:
netcdf interp_compref_2016050101 {
dimensions:
        lat = 1152 ;
        lon = 1680 ;
variables:
        float lat(lat, lon) ;
        float lon(lat, lon) ;
        float data(lat, lon) ;
}

And the MODE run command:
[jdduda at schooner1 MODE]$ ./mode
/scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
/scratch/jdduda/refl_201605/interp_compref_2016050101.nc MODEConfig
DEBUG 1: Default Config File:
/home/jdduda/tool/met/share/met/config/MODEConfig_default
DEBUG 1: Match Config File: MODEConfig
DEBUG 1: Merge Config File: MODEConfig
ERROR  :
ERROR  : Trouble reading observation file
"/scratch/jdduda/refl_201605/
interp_compref_2016050101.nc"
ERROR  :

And the section of the config file:
obs = {
   field = {
      name = "data";
      level = "(*,*)";

   raw_thresh        = NA;
   conv_radius       = 50.0/grid_res; // in grid squares
   conv_thresh       = >=0.5;
   vld_thresh        = 0.5;
   area_thresh       = NA;
   inten_perc_value  = 100;
   inten_perc_thresh = NA;
   merge_thresh      = >=1.25;
   merge_flag        = ENGINE;

   }
}

Did I do this right yet?

Jeff

On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT
<met_help at ucar.edu
> wrote:

> Jeff,
>
> Yes, that file doesn't follow the NetCDF Climate Forecast
conventions which
> are described in this document:
>    http://cfconventions.org/latest.html
>
> NetCDF is a very wide-open format, and it isn't really realistic for
> software to be able to read them in general without agreeing to some
ground
> rules... and the CF conventions are an attempt to establish them.
In
> particular, MET is looking for a definition of the grid on which
this data
> resides and timing information.  And there are CF-compliant ways of
> specifying those.
>
> Alternatively, you could make your NetCDF file look like the NetCDF
output
> files that MET generates (i.e. like pcp_combine or gen_vx_mask).
MET knows
> how to read it's own output.  And you may find that simplistic
format
> easier to follow than the CF-conventions.
>
> We do have some ideas about how to make this easier in the future.
Ideally
> MET would be able to read data from your file as-is... but then
specify the
> grid and timing information directly in the MODE config file.
> Unfortunately, the software is not set up to do that yet.
>
> Just let us know if you get stuck making your NetCDF file look like
the
> output of pcp_combine.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
>
> On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <met_help at ucar.edu>
> wrote:
>
> >
> > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > Transaction: Ticket created by jeffduda319 at gmail.com
> >        Queue: met_help
> >      Subject: help with file format for MODE v 5.2
> >        Owner: Nobody
> >   Requestors: jeffduda319 at gmail.com
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> >
> >
> > I am using MODE to verify precipitation forecasts. I have an
observation
> > dataset (it's basically MRMS, but processed from the raw stuff) in
netCDF
> > format, but MODE doesn't seem to like it when I use it for MODE.
> >
> > [jdduda at schooner1 MODE]$ ./mode
> > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
MODEConfig
> > DEBUG 1: Default Config File:
> > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > DEBUG 1: Match Config File: MODEConfig
> > DEBUG 1: Merge Config File: MODEConfig
> > ERROR  :
> > ERROR  : Trouble reading observation file
"/condo/map/jdduda/MRMS_obs/
> > rain/
> > 1hr_precip_20160525_000000.nc"
> > ERROR  :
> >
> > Here is the dump for that file:
> >
> > netcdf \1hr_precip_20160525_000000 {
> > dimensions:
> >         x = 1568 ;
> >         y = 1120 ;
> >         z = 1 ;
> > variables:
> >         float hsr1hr(z, y, x) ;
> >                 hsr1hr:standard_name = "hsr1hr" ;
> >                 hsr1hr:long_name = "1-h precipitation" ;
> >                 hsr1hr:units = "mm" ;
> > }
> >
> > It is my understanding that this may not be "CF compliant",
whatever that
> > means. If not, what can I do to make it CF compliant, or is it
possible
> I'm
> > mis-specifying the part of the config file where this is
important. Here
> is
> > the config file part:
> >
> > obs = {
> >    field = {
> >       name = "hsr1hr";
> >       level = "(*,*,*)";
> >
> >    raw_thresh        = NA;
> >    conv_radius       = 50.0/grid_res; // in grid squares
> >    conv_thresh       = >=0.5;
> >    vld_thresh        = 0.5;
> >    area_thresh       = NA;
> >    inten_perc_value  = 100;
> >    inten_perc_thresh = NA;
> >    merge_thresh      = >=1.25;
> >    merge_flag        = ENGINE;
> >
> >    }
> > }
> >
> > I tried (0,*,*) and (1,*,*) for the level, but I got the same
error for
> > each.
> >
> > Thanks.
> >
> > Jeff Duda
> >
> > --
> > Jeff Duda
> > Post-doctoral research fellow
> > University of Oklahoma School of Meteorology
> >
> >
>
>


--
Jeff Duda
Post-doctoral research fellow
University of Oklahoma School of Meteorology

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: John Halley Gotway
Time: Thu Feb 09 16:02:22 2017

Jeff,

Can you please post a sample NetCDF file to our anonymous ftp site
following these instructions:
   http://www.dtcenter.org/met/users/support/met_help.php#ftp

And I can take a closer look.  Please email me once it's posted so I
know
to go take a look.

Thanks,
John


On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT <met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
>
> Thanks for the reply, John. I made some changes to my system but am
still
> having problems. I looked at Tables 5.1 and 5.2 in the User's Guide
to see
> what the format from the output of pcp_combine would be, and I
reorganized
> my observation files to match that format. I think i did it, but I
get the
> same error.
>
> Here is the ncdump for my obs file now:
> netcdf interp_compref_2016050101 {
> dimensions:
>         lat = 1152 ;
>         lon = 1680 ;
> variables:
>         float lat(lat, lon) ;
>         float lon(lat, lon) ;
>         float data(lat, lon) ;
> }
>
> And the MODE run command:
> [jdduda at schooner1 MODE]$ ./mode
> /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> /scratch/jdduda/refl_201605/interp_compref_2016050101.nc MODEConfig
> DEBUG 1: Default Config File:
> /home/jdduda/tool/met/share/met/config/MODEConfig_default
> DEBUG 1: Match Config File: MODEConfig
> DEBUG 1: Merge Config File: MODEConfig
> ERROR  :
> ERROR  : Trouble reading observation file
"/scratch/jdduda/refl_201605/
> interp_compref_2016050101.nc"
> ERROR  :
>
> And the section of the config file:
> obs = {
>    field = {
>       name = "data";
>       level = "(*,*)";
>
>    raw_thresh        = NA;
>    conv_radius       = 50.0/grid_res; // in grid squares
>    conv_thresh       = >=0.5;
>    vld_thresh        = 0.5;
>    area_thresh       = NA;
>    inten_perc_value  = 100;
>    inten_perc_thresh = NA;
>    merge_thresh      = >=1.25;
>    merge_flag        = ENGINE;
>
>    }
> }
>
> Did I do this right yet?
>
> Jeff
>
> On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT <
> met_help at ucar.edu
> > wrote:
>
> > Jeff,
> >
> > Yes, that file doesn't follow the NetCDF Climate Forecast
conventions
> which
> > are described in this document:
> >    http://cfconventions.org/latest.html
> >
> > NetCDF is a very wide-open format, and it isn't really realistic
for
> > software to be able to read them in general without agreeing to
some
> ground
> > rules... and the CF conventions are an attempt to establish them.
In
> > particular, MET is looking for a definition of the grid on which
this
> data
> > resides and timing information.  And there are CF-compliant ways
of
> > specifying those.
> >
> > Alternatively, you could make your NetCDF file look like the
NetCDF
> output
> > files that MET generates (i.e. like pcp_combine or gen_vx_mask).
MET
> knows
> > how to read it's own output.  And you may find that simplistic
format
> > easier to follow than the CF-conventions.
> >
> > We do have some ideas about how to make this easier in the future.
> Ideally
> > MET would be able to read data from your file as-is... but then
specify
> the
> > grid and timing information directly in the MODE config file.
> > Unfortunately, the software is not set up to do that yet.
> >
> > Just let us know if you get stuck making your NetCDF file look
like the
> > output of pcp_combine.
> >
> > Thanks,
> > John Halley Gotway
> > met_help at ucar.edu
> >
> >
> > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > > Transaction: Ticket created by jeffduda319 at gmail.com
> > >        Queue: met_help
> > >      Subject: help with file format for MODE v 5.2
> > >        Owner: Nobody
> > >   Requestors: jeffduda319 at gmail.com
> > >       Status: new
> > >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394
> >
> > >
> > >
> > > I am using MODE to verify precipitation forecasts. I have an
> observation
> > > dataset (it's basically MRMS, but processed from the raw stuff)
in
> netCDF
> > > format, but MODE doesn't seem to like it when I use it for MODE.
> > >
> > > [jdduda at schooner1 MODE]$ ./mode
> > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
> MODEConfig
> > > DEBUG 1: Default Config File:
> > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > DEBUG 1: Match Config File: MODEConfig
> > > DEBUG 1: Merge Config File: MODEConfig
> > > ERROR  :
> > > ERROR  : Trouble reading observation file
"/condo/map/jdduda/MRMS_obs/
> > > rain/
> > > 1hr_precip_20160525_000000.nc"
> > > ERROR  :
> > >
> > > Here is the dump for that file:
> > >
> > > netcdf \1hr_precip_20160525_000000 {
> > > dimensions:
> > >         x = 1568 ;
> > >         y = 1120 ;
> > >         z = 1 ;
> > > variables:
> > >         float hsr1hr(z, y, x) ;
> > >                 hsr1hr:standard_name = "hsr1hr" ;
> > >                 hsr1hr:long_name = "1-h precipitation" ;
> > >                 hsr1hr:units = "mm" ;
> > > }
> > >
> > > It is my understanding that this may not be "CF compliant",
whatever
> that
> > > means. If not, what can I do to make it CF compliant, or is it
possible
> > I'm
> > > mis-specifying the part of the config file where this is
important.
> Here
> > is
> > > the config file part:
> > >
> > > obs = {
> > >    field = {
> > >       name = "hsr1hr";
> > >       level = "(*,*,*)";
> > >
> > >    raw_thresh        = NA;
> > >    conv_radius       = 50.0/grid_res; // in grid squares
> > >    conv_thresh       = >=0.5;
> > >    vld_thresh        = 0.5;
> > >    area_thresh       = NA;
> > >    inten_perc_value  = 100;
> > >    inten_perc_thresh = NA;
> > >    merge_thresh      = >=1.25;
> > >    merge_flag        = ENGINE;
> > >
> > >    }
> > > }
> > >
> > > I tried (0,*,*) and (1,*,*) for the level, but I got the same
error for
> > > each.
> > >
> > > Thanks.
> > >
> > > Jeff Duda
> > >
> > > --
> > > Jeff Duda
> > > Post-doctoral research fellow
> > > University of Oklahoma School of Meteorology
> > >
> > >
> >
> >
>
>
> --
> Jeff Duda
> Post-doctoral research fellow
> University of Oklahoma School of Meteorology
>
>

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: Jeff Duda
Time: Thu Feb 09 17:10:03 2017

John,
Okay, it's there.

Jeff

On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT
<met_help at ucar.edu
> wrote:

> Jeff,
>
> Can you please post a sample NetCDF file to our anonymous ftp site
> following these instructions:
>    http://www.dtcenter.org/met/users/support/met_help.php#ftp
>
> And I can take a closer look.  Please email me once it's posted so I
know
> to go take a look.
>
> Thanks,
> John
>
>
> On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> >
> > Thanks for the reply, John. I made some changes to my system but
am still
> > having problems. I looked at Tables 5.1 and 5.2 in the User's
Guide to
> see
> > what the format from the output of pcp_combine would be, and I
> reorganized
> > my observation files to match that format. I think i did it, but I
get
> the
> > same error.
> >
> > Here is the ncdump for my obs file now:
> > netcdf interp_compref_2016050101 {
> > dimensions:
> >         lat = 1152 ;
> >         lon = 1680 ;
> > variables:
> >         float lat(lat, lon) ;
> >         float lon(lat, lon) ;
> >         float data(lat, lon) ;
> > }
> >
> > And the MODE run command:
> > [jdduda at schooner1 MODE]$ ./mode
> > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
MODEConfig
> > DEBUG 1: Default Config File:
> > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > DEBUG 1: Match Config File: MODEConfig
> > DEBUG 1: Merge Config File: MODEConfig
> > ERROR  :
> > ERROR  : Trouble reading observation file
"/scratch/jdduda/refl_201605/
> > interp_compref_2016050101.nc"
> > ERROR  :
> >
> > And the section of the config file:
> > obs = {
> >    field = {
> >       name = "data";
> >       level = "(*,*)";
> >
> >    raw_thresh        = NA;
> >    conv_radius       = 50.0/grid_res; // in grid squares
> >    conv_thresh       = >=0.5;
> >    vld_thresh        = 0.5;
> >    area_thresh       = NA;
> >    inten_perc_value  = 100;
> >    inten_perc_thresh = NA;
> >    merge_thresh      = >=1.25;
> >    merge_flag        = ENGINE;
> >
> >    }
> > }
> >
> > Did I do this right yet?
> >
> > Jeff
> >
> > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT <
> > met_help at ucar.edu
> > > wrote:
> >
> > > Jeff,
> > >
> > > Yes, that file doesn't follow the NetCDF Climate Forecast
conventions
> > which
> > > are described in this document:
> > >    http://cfconventions.org/latest.html
> > >
> > > NetCDF is a very wide-open format, and it isn't really realistic
for
> > > software to be able to read them in general without agreeing to
some
> > ground
> > > rules... and the CF conventions are an attempt to establish
them.  In
> > > particular, MET is looking for a definition of the grid on which
this
> > data
> > > resides and timing information.  And there are CF-compliant ways
of
> > > specifying those.
> > >
> > > Alternatively, you could make your NetCDF file look like the
NetCDF
> > output
> > > files that MET generates (i.e. like pcp_combine or gen_vx_mask).
MET
> > knows
> > > how to read it's own output.  And you may find that simplistic
format
> > > easier to follow than the CF-conventions.
> > >
> > > We do have some ideas about how to make this easier in the
future.
> > Ideally
> > > MET would be able to read data from your file as-is... but then
specify
> > the
> > > grid and timing information directly in the MODE config file.
> > > Unfortunately, the software is not set up to do that yet.
> > >
> > > Just let us know if you get stuck making your NetCDF file look
like the
> > > output of pcp_combine.
> > >
> > > Thanks,
> > > John Halley Gotway
> > > met_help at ucar.edu
> > >
> > >
> > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > >        Queue: met_help
> > > >      Subject: help with file format for MODE v 5.2
> > > >        Owner: Nobody
> > > >   Requestors: jeffduda319 at gmail.com
> > > >       Status: new
> > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> Ticket/Display.html?id=79394
> > >
> > > >
> > > >
> > > > I am using MODE to verify precipitation forecasts. I have an
> > observation
> > > > dataset (it's basically MRMS, but processed from the raw
stuff) in
> > netCDF
> > > > format, but MODE doesn't seem to like it when I use it for
MODE.
> > > >
> > > > [jdduda at schooner1 MODE]$ ./mode
> > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
> > MODEConfig
> > > > DEBUG 1: Default Config File:
> > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > DEBUG 1: Match Config File: MODEConfig
> > > > DEBUG 1: Merge Config File: MODEConfig
> > > > ERROR  :
> > > > ERROR  : Trouble reading observation file
> "/condo/map/jdduda/MRMS_obs/
> > > > rain/
> > > > 1hr_precip_20160525_000000.nc"
> > > > ERROR  :
> > > >
> > > > Here is the dump for that file:
> > > >
> > > > netcdf \1hr_precip_20160525_000000 {
> > > > dimensions:
> > > >         x = 1568 ;
> > > >         y = 1120 ;
> > > >         z = 1 ;
> > > > variables:
> > > >         float hsr1hr(z, y, x) ;
> > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > >                 hsr1hr:long_name = "1-h precipitation" ;
> > > >                 hsr1hr:units = "mm" ;
> > > > }
> > > >
> > > > It is my understanding that this may not be "CF compliant",
whatever
> > that
> > > > means. If not, what can I do to make it CF compliant, or is it
> possible
> > > I'm
> > > > mis-specifying the part of the config file where this is
important.
> > Here
> > > is
> > > > the config file part:
> > > >
> > > > obs = {
> > > >    field = {
> > > >       name = "hsr1hr";
> > > >       level = "(*,*,*)";
> > > >
> > > >    raw_thresh        = NA;
> > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > >    conv_thresh       = >=0.5;
> > > >    vld_thresh        = 0.5;
> > > >    area_thresh       = NA;
> > > >    inten_perc_value  = 100;
> > > >    inten_perc_thresh = NA;
> > > >    merge_thresh      = >=1.25;
> > > >    merge_flag        = ENGINE;
> > > >
> > > >    }
> > > > }
> > > >
> > > > I tried (0,*,*) and (1,*,*) for the level, but I got the same
error
> for
> > > > each.
> > > >
> > > > Thanks.
> > > >
> > > > Jeff Duda
> > > >
> > > > --
> > > > Jeff Duda
> > > > Post-doctoral research fellow
> > > > University of Oklahoma School of Meteorology
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Jeff Duda
> > Post-doctoral research fellow
> > University of Oklahoma School of Meteorology
> >
> >
>
>


--
Jeff Duda
Post-doctoral research fellow
University of Oklahoma School of Meteorology

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: John Halley Gotway
Time: Fri Feb 10 16:25:30 2017

Jeff,

Thanks for sending the sample data file.  Looking at the latitude and
longitude values, I see that they're not evenly spaced.  And that
tells me
this is not on a simple lat/lon grid.  For regular lat/lon grids, MET
can
use the lat/lon values to infer the underlying grid information.
However,
for other projection types, like lambert conformal, polar
sterographic, and
mercator, there is no good way of processing lat/lon values to back
out the
underlying grid/projection information.

Instead, we need to add the projection info in the NetCDF file.  I see
that
the data has dimension 1152 x 1680.

Looking back in the message history I see a directory named
"MRMS_obs".
But I checked the MRMS data from the NCEP website and see that it uses
a
7000x3500 lat/lon grid.  If you can point me to the grid/projection
information, I should be able to help you out.  If you have any GRIB1
or
GRIB2 files on the same grid, I could easily extract the grid
information
from those.

Thanks,
John



On Thu, Feb 9, 2017 at 5:10 PM, Jeff Duda via RT <met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
>
> John,
> Okay, it's there.
>
> Jeff
>
> On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT <
> met_help at ucar.edu
> > wrote:
>
> > Jeff,
> >
> > Can you please post a sample NetCDF file to our anonymous ftp site
> > following these instructions:
> >    http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >
> > And I can take a closer look.  Please email me once it's posted so
I know
> > to go take a look.
> >
> > Thanks,
> > John
> >
> >
> > On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > >
> > > Thanks for the reply, John. I made some changes to my system but
am
> still
> > > having problems. I looked at Tables 5.1 and 5.2 in the User's
Guide to
> > see
> > > what the format from the output of pcp_combine would be, and I
> > reorganized
> > > my observation files to match that format. I think i did it, but
I get
> > the
> > > same error.
> > >
> > > Here is the ncdump for my obs file now:
> > > netcdf interp_compref_2016050101 {
> > > dimensions:
> > >         lat = 1152 ;
> > >         lon = 1680 ;
> > > variables:
> > >         float lat(lat, lon) ;
> > >         float lon(lat, lon) ;
> > >         float data(lat, lon) ;
> > > }
> > >
> > > And the MODE run command:
> > > [jdduda at schooner1 MODE]$ ./mode
> > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
MODEConfig
> > > DEBUG 1: Default Config File:
> > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > DEBUG 1: Match Config File: MODEConfig
> > > DEBUG 1: Merge Config File: MODEConfig
> > > ERROR  :
> > > ERROR  : Trouble reading observation file
"/scratch/jdduda/refl_201605/
> > > interp_compref_2016050101.nc"
> > > ERROR  :
> > >
> > > And the section of the config file:
> > > obs = {
> > >    field = {
> > >       name = "data";
> > >       level = "(*,*)";
> > >
> > >    raw_thresh        = NA;
> > >    conv_radius       = 50.0/grid_res; // in grid squares
> > >    conv_thresh       = >=0.5;
> > >    vld_thresh        = 0.5;
> > >    area_thresh       = NA;
> > >    inten_perc_value  = 100;
> > >    inten_perc_thresh = NA;
> > >    merge_thresh      = >=1.25;
> > >    merge_flag        = ENGINE;
> > >
> > >    }
> > > }
> > >
> > > Did I do this right yet?
> > >
> > > Jeff
> > >
> > > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu
> > > > wrote:
> > >
> > > > Jeff,
> > > >
> > > > Yes, that file doesn't follow the NetCDF Climate Forecast
conventions
> > > which
> > > > are described in this document:
> > > >    http://cfconventions.org/latest.html
> > > >
> > > > NetCDF is a very wide-open format, and it isn't really
realistic for
> > > > software to be able to read them in general without agreeing
to some
> > > ground
> > > > rules... and the CF conventions are an attempt to establish
them.  In
> > > > particular, MET is looking for a definition of the grid on
which this
> > > data
> > > > resides and timing information.  And there are CF-compliant
ways of
> > > > specifying those.
> > > >
> > > > Alternatively, you could make your NetCDF file look like the
NetCDF
> > > output
> > > > files that MET generates (i.e. like pcp_combine or
gen_vx_mask).  MET
> > > knows
> > > > how to read it's own output.  And you may find that simplistic
format
> > > > easier to follow than the CF-conventions.
> > > >
> > > > We do have some ideas about how to make this easier in the
future.
> > > Ideally
> > > > MET would be able to read data from your file as-is... but
then
> specify
> > > the
> > > > grid and timing information directly in the MODE config file.
> > > > Unfortunately, the software is not set up to do that yet.
> > > >
> > > > Just let us know if you get stuck making your NetCDF file look
like
> the
> > > > output of pcp_combine.
> > > >
> > > > Thanks,
> > > > John Halley Gotway
> > > > met_help at ucar.edu
> > > >
> > > >
> > > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > > wrote:
> > > >
> > > > >
> > > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > > >        Queue: met_help
> > > > >      Subject: help with file format for MODE v 5.2
> > > > >        Owner: Nobody
> > > > >   Requestors: jeffduda319 at gmail.com
> > > > >       Status: new
> > > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> > Ticket/Display.html?id=79394
> > > >
> > > > >
> > > > >
> > > > > I am using MODE to verify precipitation forecasts. I have an
> > > observation
> > > > > dataset (it's basically MRMS, but processed from the raw
stuff) in
> > > netCDF
> > > > > format, but MODE doesn't seem to like it when I use it for
MODE.
> > > > >
> > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > >
/condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
> > > MODEConfig
> > > > > DEBUG 1: Default Config File:
> > > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > ERROR  :
> > > > > ERROR  : Trouble reading observation file
> > "/condo/map/jdduda/MRMS_obs/
> > > > > rain/
> > > > > 1hr_precip_20160525_000000.nc"
> > > > > ERROR  :
> > > > >
> > > > > Here is the dump for that file:
> > > > >
> > > > > netcdf \1hr_precip_20160525_000000 {
> > > > > dimensions:
> > > > >         x = 1568 ;
> > > > >         y = 1120 ;
> > > > >         z = 1 ;
> > > > > variables:
> > > > >         float hsr1hr(z, y, x) ;
> > > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > > >                 hsr1hr:long_name = "1-h precipitation" ;
> > > > >                 hsr1hr:units = "mm" ;
> > > > > }
> > > > >
> > > > > It is my understanding that this may not be "CF compliant",
> whatever
> > > that
> > > > > means. If not, what can I do to make it CF compliant, or is
it
> > possible
> > > > I'm
> > > > > mis-specifying the part of the config file where this is
important.
> > > Here
> > > > is
> > > > > the config file part:
> > > > >
> > > > > obs = {
> > > > >    field = {
> > > > >       name = "hsr1hr";
> > > > >       level = "(*,*,*)";
> > > > >
> > > > >    raw_thresh        = NA;
> > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > >    conv_thresh       = >=0.5;
> > > > >    vld_thresh        = 0.5;
> > > > >    area_thresh       = NA;
> > > > >    inten_perc_value  = 100;
> > > > >    inten_perc_thresh = NA;
> > > > >    merge_thresh      = >=1.25;
> > > > >    merge_flag        = ENGINE;
> > > > >
> > > > >    }
> > > > > }
> > > > >
> > > > > I tried (0,*,*) and (1,*,*) for the level, but I got the
same error
> > for
> > > > > each.
> > > > >
> > > > > Thanks.
> > > > >
> > > > > Jeff Duda
> > > > >
> > > > > --
> > > > > Jeff Duda
> > > > > Post-doctoral research fellow
> > > > > University of Oklahoma School of Meteorology
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Jeff Duda
> > > Post-doctoral research fellow
> > > University of Oklahoma School of Meteorology
> > >
> > >
> >
> >
>
>
> --
> Jeff Duda
> Post-doctoral research fellow
> University of Oklahoma School of Meteorology
>
>

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: Jeff Duda
Time: Fri Feb 10 17:16:32 2017

I know the projection details already. I take it I should insert those
into
the netcdf file when writing it out? Is there a specific syntax I
should
use for the attributes?

Jeff

On Fri, Feb 10, 2017 at 5:25 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:

> Jeff,
>
> Thanks for sending the sample data file.  Looking at the latitude
and
> longitude values, I see that they're not evenly spaced.  And that
tells me
> this is not on a simple lat/lon grid.  For regular lat/lon grids,
MET can
> use the lat/lon values to infer the underlying grid information.
However,
> for other projection types, like lambert conformal, polar
sterographic, and
> mercator, there is no good way of processing lat/lon values to back
out the
> underlying grid/projection information.
>
> Instead, we need to add the projection info in the NetCDF file.  I
see that
> the data has dimension 1152 x 1680.
>
> Looking back in the message history I see a directory named
"MRMS_obs".
> But I checked the MRMS data from the NCEP website and see that it
uses a
> 7000x3500 lat/lon grid.  If you can point me to the grid/projection
> information, I should be able to help you out.  If you have any
GRIB1 or
> GRIB2 files on the same grid, I could easily extract the grid
information
> from those.
>
> Thanks,
> John
>
>
>
> On Thu, Feb 9, 2017 at 5:10 PM, Jeff Duda via RT <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> >
> > John,
> > Okay, it's there.
> >
> > Jeff
> >
> > On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT <
> > met_help at ucar.edu
> > > wrote:
> >
> > > Jeff,
> > >
> > > Can you please post a sample NetCDF file to our anonymous ftp
site
> > > following these instructions:
> > >    http://www.dtcenter.org/met/users/support/met_help.php#ftp
> > >
> > > And I can take a closer look.  Please email me once it's posted
so I
> know
> > > to go take a look.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > > On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394
>
> > > >
> > > > Thanks for the reply, John. I made some changes to my system
but am
> > still
> > > > having problems. I looked at Tables 5.1 and 5.2 in the User's
Guide
> to
> > > see
> > > > what the format from the output of pcp_combine would be, and I
> > > reorganized
> > > > my observation files to match that format. I think i did it,
but I
> get
> > > the
> > > > same error.
> > > >
> > > > Here is the ncdump for my obs file now:
> > > > netcdf interp_compref_2016050101 {
> > > > dimensions:
> > > >         lat = 1152 ;
> > > >         lon = 1680 ;
> > > > variables:
> > > >         float lat(lat, lon) ;
> > > >         float lon(lat, lon) ;
> > > >         float data(lat, lon) ;
> > > > }
> > > >
> > > > And the MODE run command:
> > > > [jdduda at schooner1 MODE]$ ./mode
> > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
MODEConfig
> > > > DEBUG 1: Default Config File:
> > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > DEBUG 1: Match Config File: MODEConfig
> > > > DEBUG 1: Merge Config File: MODEConfig
> > > > ERROR  :
> > > > ERROR  : Trouble reading observation file
> "/scratch/jdduda/refl_201605/
> > > > interp_compref_2016050101.nc"
> > > > ERROR  :
> > > >
> > > > And the section of the config file:
> > > > obs = {
> > > >    field = {
> > > >       name = "data";
> > > >       level = "(*,*)";
> > > >
> > > >    raw_thresh        = NA;
> > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > >    conv_thresh       = >=0.5;
> > > >    vld_thresh        = 0.5;
> > > >    area_thresh       = NA;
> > > >    inten_perc_value  = 100;
> > > >    inten_perc_thresh = NA;
> > > >    merge_thresh      = >=1.25;
> > > >    merge_flag        = ENGINE;
> > > >
> > > >    }
> > > > }
> > > >
> > > > Did I do this right yet?
> > > >
> > > > Jeff
> > > >
> > > > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu
> > > > > wrote:
> > > >
> > > > > Jeff,
> > > > >
> > > > > Yes, that file doesn't follow the NetCDF Climate Forecast
> conventions
> > > > which
> > > > > are described in this document:
> > > > >    http://cfconventions.org/latest.html
> > > > >
> > > > > NetCDF is a very wide-open format, and it isn't really
realistic
> for
> > > > > software to be able to read them in general without agreeing
to
> some
> > > > ground
> > > > > rules... and the CF conventions are an attempt to establish
them.
> In
> > > > > particular, MET is looking for a definition of the grid on
which
> this
> > > > data
> > > > > resides and timing information.  And there are CF-compliant
ways of
> > > > > specifying those.
> > > > >
> > > > > Alternatively, you could make your NetCDF file look like the
NetCDF
> > > > output
> > > > > files that MET generates (i.e. like pcp_combine or
gen_vx_mask).
> MET
> > > > knows
> > > > > how to read it's own output.  And you may find that
simplistic
> format
> > > > > easier to follow than the CF-conventions.
> > > > >
> > > > > We do have some ideas about how to make this easier in the
future.
> > > > Ideally
> > > > > MET would be able to read data from your file as-is... but
then
> > specify
> > > > the
> > > > > grid and timing information directly in the MODE config
file.
> > > > > Unfortunately, the software is not set up to do that yet.
> > > > >
> > > > > Just let us know if you get stuck making your NetCDF file
look like
> > the
> > > > > output of pcp_combine.
> > > > >
> > > > > Thanks,
> > > > > John Halley Gotway
> > > > > met_help at ucar.edu
> > > > >
> > > > >
> > > > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <
> met_help at ucar.edu>
> > > > > wrote:
> > > > >
> > > > > >
> > > > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > > > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > > > >        Queue: met_help
> > > > > >      Subject: help with file format for MODE v 5.2
> > > > > >        Owner: Nobody
> > > > > >   Requestors: jeffduda319 at gmail.com
> > > > > >       Status: new
> > > > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > Ticket/Display.html?id=79394
> > > > >
> > > > > >
> > > > > >
> > > > > > I am using MODE to verify precipitation forecasts. I have
an
> > > > observation
> > > > > > dataset (it's basically MRMS, but processed from the raw
stuff)
> in
> > > > netCDF
> > > > > > format, but MODE doesn't seem to like it when I use it for
MODE.
> > > > > >
> > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > >
/condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
> > > > MODEConfig
> > > > > > DEBUG 1: Default Config File:
> > > > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > ERROR  :
> > > > > > ERROR  : Trouble reading observation file
> > > "/condo/map/jdduda/MRMS_obs/
> > > > > > rain/
> > > > > > 1hr_precip_20160525_000000.nc"
> > > > > > ERROR  :
> > > > > >
> > > > > > Here is the dump for that file:
> > > > > >
> > > > > > netcdf \1hr_precip_20160525_000000 {
> > > > > > dimensions:
> > > > > >         x = 1568 ;
> > > > > >         y = 1120 ;
> > > > > >         z = 1 ;
> > > > > > variables:
> > > > > >         float hsr1hr(z, y, x) ;
> > > > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > > > >                 hsr1hr:long_name = "1-h precipitation" ;
> > > > > >                 hsr1hr:units = "mm" ;
> > > > > > }
> > > > > >
> > > > > > It is my understanding that this may not be "CF
compliant",
> > whatever
> > > > that
> > > > > > means. If not, what can I do to make it CF compliant, or
is it
> > > possible
> > > > > I'm
> > > > > > mis-specifying the part of the config file where this is
> important.
> > > > Here
> > > > > is
> > > > > > the config file part:
> > > > > >
> > > > > > obs = {
> > > > > >    field = {
> > > > > >       name = "hsr1hr";
> > > > > >       level = "(*,*,*)";
> > > > > >
> > > > > >    raw_thresh        = NA;
> > > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > > >    conv_thresh       = >=0.5;
> > > > > >    vld_thresh        = 0.5;
> > > > > >    area_thresh       = NA;
> > > > > >    inten_perc_value  = 100;
> > > > > >    inten_perc_thresh = NA;
> > > > > >    merge_thresh      = >=1.25;
> > > > > >    merge_flag        = ENGINE;
> > > > > >
> > > > > >    }
> > > > > > }
> > > > > >
> > > > > > I tried (0,*,*) and (1,*,*) for the level, but I got the
same
> error
> > > for
> > > > > > each.
> > > > > >
> > > > > > Thanks.
> > > > > >
> > > > > > Jeff Duda
> > > > > >
> > > > > > --
> > > > > > Jeff Duda
> > > > > > Post-doctoral research fellow
> > > > > > University of Oklahoma School of Meteorology
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Jeff Duda
> > > > Post-doctoral research fellow
> > > > University of Oklahoma School of Meteorology
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Jeff Duda
> > Post-doctoral research fellow
> > University of Oklahoma School of Meteorology
> >
> >
>
>


--
Jeff Duda
Post-doctoral research fellow
University of Oklahoma School of Meteorology

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: John Halley Gotway
Time: Fri Feb 10 19:26:45 2017

Jeff,

Yes, you can either follow the cf-convention way of doing it using a
grid_mapping attribute or make it look like the NetCDF output of the
pcp_combine tool. For the latter, pass a GRIB file on that projection
through pcp_combine and that'll describe exactly what you need.

John


On Fri, Feb 10, 2017 at 5:16 PM Jeff Duda via RT <met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
>
> I know the projection details already. I take it I should insert
those into
> the netcdf file when writing it out? Is there a specific syntax I
should
> use for the attributes?
>
> Jeff
>
> On Fri, Feb 10, 2017 at 5:25 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Jeff,
> >
> > Thanks for sending the sample data file.  Looking at the latitude
and
> > longitude values, I see that they're not evenly spaced.  And that
tells
> me
> > this is not on a simple lat/lon grid.  For regular lat/lon grids,
MET can
> > use the lat/lon values to infer the underlying grid information.
> However,
> > for other projection types, like lambert conformal, polar
sterographic,
> and
> > mercator, there is no good way of processing lat/lon values to
back out
> the
> > underlying grid/projection information.
> >
> > Instead, we need to add the projection info in the NetCDF file.  I
see
> that
> > the data has dimension 1152 x 1680.
> >
> > Looking back in the message history I see a directory named
"MRMS_obs".
> > But I checked the MRMS data from the NCEP website and see that it
uses a
> > 7000x3500 lat/lon grid.  If you can point me to the
grid/projection
> > information, I should be able to help you out.  If you have any
GRIB1 or
> > GRIB2 files on the same grid, I could easily extract the grid
information
> > from those.
> >
> > Thanks,
> > John
> >
> >
> >
> > On Thu, Feb 9, 2017 at 5:10 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > >
> > > John,
> > > Okay, it's there.
> > >
> > > Jeff
> > >
> > > On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu
> > > > wrote:
> > >
> > > > Jeff,
> > > >
> > > > Can you please post a sample NetCDF file to our anonymous ftp
site
> > > > following these instructions:
> > > >    http://www.dtcenter.org/met/users/support/met_help.php#ftp
> > > >
> > > > And I can take a closer look.  Please email me once it's
posted so I
> > know
> > > > to go take a look.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > > On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > > wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > > > >
> > > > > Thanks for the reply, John. I made some changes to my system
but am
> > > still
> > > > > having problems. I looked at Tables 5.1 and 5.2 in the
User's Guide
> > to
> > > > see
> > > > > what the format from the output of pcp_combine would be, and
I
> > > > reorganized
> > > > > my observation files to match that format. I think i did it,
but I
> > get
> > > > the
> > > > > same error.
> > > > >
> > > > > Here is the ncdump for my obs file now:
> > > > > netcdf interp_compref_2016050101 {
> > > > > dimensions:
> > > > >         lat = 1152 ;
> > > > >         lon = 1680 ;
> > > > > variables:
> > > > >         float lat(lat, lon) ;
> > > > >         float lon(lat, lon) ;
> > > > >         float data(lat, lon) ;
> > > > > }
> > > > >
> > > > > And the MODE run command:
> > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
> MODEConfig
> > > > > DEBUG 1: Default Config File:
> > > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > ERROR  :
> > > > > ERROR  : Trouble reading observation file
> > "/scratch/jdduda/refl_201605/
> > > > > interp_compref_2016050101.nc"
> > > > > ERROR  :
> > > > >
> > > > > And the section of the config file:
> > > > > obs = {
> > > > >    field = {
> > > > >       name = "data";
> > > > >       level = "(*,*)";
> > > > >
> > > > >    raw_thresh        = NA;
> > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > >    conv_thresh       = >=0.5;
> > > > >    vld_thresh        = 0.5;
> > > > >    area_thresh       = NA;
> > > > >    inten_perc_value  = 100;
> > > > >    inten_perc_thresh = NA;
> > > > >    merge_thresh      = >=1.25;
> > > > >    merge_flag        = ENGINE;
> > > > >
> > > > >    }
> > > > > }
> > > > >
> > > > > Did I do this right yet?
> > > > >
> > > > > Jeff
> > > > >
> > > > > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT <
> > > > > met_help at ucar.edu
> > > > > > wrote:
> > > > >
> > > > > > Jeff,
> > > > > >
> > > > > > Yes, that file doesn't follow the NetCDF Climate Forecast
> > conventions
> > > > > which
> > > > > > are described in this document:
> > > > > >    http://cfconventions.org/latest.html
> > > > > >
> > > > > > NetCDF is a very wide-open format, and it isn't really
realistic
> > for
> > > > > > software to be able to read them in general without
agreeing to
> > some
> > > > > ground
> > > > > > rules... and the CF conventions are an attempt to
establish them.
> > In
> > > > > > particular, MET is looking for a definition of the grid on
which
> > this
> > > > > data
> > > > > > resides and timing information.  And there are CF-
compliant ways
> of
> > > > > > specifying those.
> > > > > >
> > > > > > Alternatively, you could make your NetCDF file look like
the
> NetCDF
> > > > > output
> > > > > > files that MET generates (i.e. like pcp_combine or
gen_vx_mask).
> > MET
> > > > > knows
> > > > > > how to read it's own output.  And you may find that
simplistic
> > format
> > > > > > easier to follow than the CF-conventions.
> > > > > >
> > > > > > We do have some ideas about how to make this easier in the
> future.
> > > > > Ideally
> > > > > > MET would be able to read data from your file as-is... but
then
> > > specify
> > > > > the
> > > > > > grid and timing information directly in the MODE config
file.
> > > > > > Unfortunately, the software is not set up to do that yet.
> > > > > >
> > > > > > Just let us know if you get stuck making your NetCDF file
look
> like
> > > the
> > > > > > output of pcp_combine.
> > > > > >
> > > > > > Thanks,
> > > > > > John Halley Gotway
> > > > > > met_help at ucar.edu
> > > > > >
> > > > > >
> > > > > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <
> > met_help at ucar.edu>
> > > > > > wrote:
> > > > > >
> > > > > > >
> > > > > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted upon.
> > > > > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > > > > >        Queue: met_help
> > > > > > >      Subject: help with file format for MODE v 5.2
> > > > > > >        Owner: Nobody
> > > > > > >   Requestors: jeffduda319 at gmail.com
> > > > > > >       Status: new
> > > > > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > > Ticket/Display.html?id=79394
> > > > > >
> > > > > > >
> > > > > > >
> > > > > > > I am using MODE to verify precipitation forecasts. I
have an
> > > > > observation
> > > > > > > dataset (it's basically MRMS, but processed from the raw
stuff)
> > in
> > > > > netCDF
> > > > > > > format, but MODE doesn't seem to like it when I use it
for
> MODE.
> > > > > > >
> > > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > > >
/condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_000000.nc
> > > > > MODEConfig
> > > > > > > DEBUG 1: Default Config File:
> > > > > > >
/home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > > ERROR  :
> > > > > > > ERROR  : Trouble reading observation file
> > > > "/condo/map/jdduda/MRMS_obs/
> > > > > > > rain/
> > > > > > > 1hr_precip_20160525_000000.nc"
> > > > > > > ERROR  :
> > > > > > >
> > > > > > > Here is the dump for that file:
> > > > > > >
> > > > > > > netcdf \1hr_precip_20160525_000000 {
> > > > > > > dimensions:
> > > > > > >         x = 1568 ;
> > > > > > >         y = 1120 ;
> > > > > > >         z = 1 ;
> > > > > > > variables:
> > > > > > >         float hsr1hr(z, y, x) ;
> > > > > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > > > > >                 hsr1hr:long_name = "1-h precipitation" ;
> > > > > > >                 hsr1hr:units = "mm" ;
> > > > > > > }
> > > > > > >
> > > > > > > It is my understanding that this may not be "CF
compliant",
> > > whatever
> > > > > that
> > > > > > > means. If not, what can I do to make it CF compliant, or
is it
> > > > possible
> > > > > > I'm
> > > > > > > mis-specifying the part of the config file where this is
> > important.
> > > > > Here
> > > > > > is
> > > > > > > the config file part:
> > > > > > >
> > > > > > > obs = {
> > > > > > >    field = {
> > > > > > >       name = "hsr1hr";
> > > > > > >       level = "(*,*,*)";
> > > > > > >
> > > > > > >    raw_thresh        = NA;
> > > > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > > > >    conv_thresh       = >=0.5;
> > > > > > >    vld_thresh        = 0.5;
> > > > > > >    area_thresh       = NA;
> > > > > > >    inten_perc_value  = 100;
> > > > > > >    inten_perc_thresh = NA;
> > > > > > >    merge_thresh      = >=1.25;
> > > > > > >    merge_flag        = ENGINE;
> > > > > > >
> > > > > > >    }
> > > > > > > }
> > > > > > >
> > > > > > > I tried (0,*,*) and (1,*,*) for the level, but I got the
same
> > error
> > > > for
> > > > > > > each.
> > > > > > >
> > > > > > > Thanks.
> > > > > > >
> > > > > > > Jeff Duda
> > > > > > >
> > > > > > > --
> > > > > > > Jeff Duda
> > > > > > > Post-doctoral research fellow
> > > > > > > University of Oklahoma School of Meteorology
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > > --
> > > > > Jeff Duda
> > > > > Post-doctoral research fellow
> > > > > University of Oklahoma School of Meteorology
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Jeff Duda
> > > Post-doctoral research fellow
> > > University of Oklahoma School of Meteorology
> > >
> > >
> >
> >
>
>
> --
> Jeff Duda
> Post-doctoral research fellow
> University of Oklahoma School of Meteorology
>
>

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: Jeff Duda
Time: Wed Feb 15 09:40:33 2017

John,
Got it! Wasn't sure exactly which attributes I had to add, so I added
almost all of them. But it works now.

Thanks!

Jeff

On Fri, Feb 10, 2017 at 8:26 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:

> Jeff,
>
> Yes, you can either follow the cf-convention way of doing it using a
> grid_mapping attribute or make it look like the NetCDF output of the
> pcp_combine tool. For the latter, pass a GRIB file on that
projection
> through pcp_combine and that'll describe exactly what you need.
>
> John
>
>
> On Fri, Feb 10, 2017 at 5:16 PM Jeff Duda via RT <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> >
> > I know the projection details already. I take it I should insert
those
> into
> > the netcdf file when writing it out? Is there a specific syntax I
should
> > use for the attributes?
> >
> > Jeff
> >
> > On Fri, Feb 10, 2017 at 5:25 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Jeff,
> > >
> > > Thanks for sending the sample data file.  Looking at the
latitude and
> > > longitude values, I see that they're not evenly spaced.  And
that tells
> > me
> > > this is not on a simple lat/lon grid.  For regular lat/lon
grids, MET
> can
> > > use the lat/lon values to infer the underlying grid information.
> > However,
> > > for other projection types, like lambert conformal, polar
sterographic,
> > and
> > > mercator, there is no good way of processing lat/lon values to
back out
> > the
> > > underlying grid/projection information.
> > >
> > > Instead, we need to add the projection info in the NetCDF file.
I see
> > that
> > > the data has dimension 1152 x 1680.
> > >
> > > Looking back in the message history I see a directory named
"MRMS_obs".
> > > But I checked the MRMS data from the NCEP website and see that
it uses
> a
> > > 7000x3500 lat/lon grid.  If you can point me to the
grid/projection
> > > information, I should be able to help you out.  If you have any
GRIB1
> or
> > > GRIB2 files on the same grid, I could easily extract the grid
> information
> > > from those.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > >
> > > On Thu, Feb 9, 2017 at 5:10 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394
>
> > > >
> > > > John,
> > > > Okay, it's there.
> > > >
> > > > Jeff
> > > >
> > > > On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu
> > > > > wrote:
> > > >
> > > > > Jeff,
> > > > >
> > > > > Can you please post a sample NetCDF file to our anonymous
ftp site
> > > > > following these instructions:
> > > > >
http://www.dtcenter.org/met/users/support/met_help.php#ftp
> > > > >
> > > > > And I can take a closer look.  Please email me once it's
posted so
> I
> > > know
> > > > > to go take a look.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > >
> > > > > On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT <
> met_help at ucar.edu>
> > > > > wrote:
> > > > >
> > > > > >
> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > > > > >
> > > > > > Thanks for the reply, John. I made some changes to my
system but
> am
> > > > still
> > > > > > having problems. I looked at Tables 5.1 and 5.2 in the
User's
> Guide
> > > to
> > > > > see
> > > > > > what the format from the output of pcp_combine would be,
and I
> > > > > reorganized
> > > > > > my observation files to match that format. I think i did
it, but
> I
> > > get
> > > > > the
> > > > > > same error.
> > > > > >
> > > > > > Here is the ncdump for my obs file now:
> > > > > > netcdf interp_compref_2016050101 {
> > > > > > dimensions:
> > > > > >         lat = 1152 ;
> > > > > >         lon = 1680 ;
> > > > > > variables:
> > > > > >         float lat(lat, lon) ;
> > > > > >         float lon(lat, lon) ;
> > > > > >         float data(lat, lon) ;
> > > > > > }
> > > > > >
> > > > > > And the MODE run command:
> > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
> > MODEConfig
> > > > > > DEBUG 1: Default Config File:
> > > > > > /home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > ERROR  :
> > > > > > ERROR  : Trouble reading observation file
> > > "/scratch/jdduda/refl_201605/
> > > > > > interp_compref_2016050101.nc"
> > > > > > ERROR  :
> > > > > >
> > > > > > And the section of the config file:
> > > > > > obs = {
> > > > > >    field = {
> > > > > >       name = "data";
> > > > > >       level = "(*,*)";
> > > > > >
> > > > > >    raw_thresh        = NA;
> > > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > > >    conv_thresh       = >=0.5;
> > > > > >    vld_thresh        = 0.5;
> > > > > >    area_thresh       = NA;
> > > > > >    inten_perc_value  = 100;
> > > > > >    inten_perc_thresh = NA;
> > > > > >    merge_thresh      = >=1.25;
> > > > > >    merge_flag        = ENGINE;
> > > > > >
> > > > > >    }
> > > > > > }
> > > > > >
> > > > > > Did I do this right yet?
> > > > > >
> > > > > > Jeff
> > > > > >
> > > > > > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via RT
<
> > > > > > met_help at ucar.edu
> > > > > > > wrote:
> > > > > >
> > > > > > > Jeff,
> > > > > > >
> > > > > > > Yes, that file doesn't follow the NetCDF Climate
Forecast
> > > conventions
> > > > > > which
> > > > > > > are described in this document:
> > > > > > >    http://cfconventions.org/latest.html
> > > > > > >
> > > > > > > NetCDF is a very wide-open format, and it isn't really
> realistic
> > > for
> > > > > > > software to be able to read them in general without
agreeing to
> > > some
> > > > > > ground
> > > > > > > rules... and the CF conventions are an attempt to
establish
> them.
> > > In
> > > > > > > particular, MET is looking for a definition of the grid
on
> which
> > > this
> > > > > > data
> > > > > > > resides and timing information.  And there are CF-
compliant
> ways
> > of
> > > > > > > specifying those.
> > > > > > >
> > > > > > > Alternatively, you could make your NetCDF file look like
the
> > NetCDF
> > > > > > output
> > > > > > > files that MET generates (i.e. like pcp_combine or
> gen_vx_mask).
> > > MET
> > > > > > knows
> > > > > > > how to read it's own output.  And you may find that
simplistic
> > > format
> > > > > > > easier to follow than the CF-conventions.
> > > > > > >
> > > > > > > We do have some ideas about how to make this easier in
the
> > future.
> > > > > > Ideally
> > > > > > > MET would be able to read data from your file as-is...
but then
> > > > specify
> > > > > > the
> > > > > > > grid and timing information directly in the MODE config
file.
> > > > > > > Unfortunately, the software is not set up to do that
yet.
> > > > > > >
> > > > > > > Just let us know if you get stuck making your NetCDF
file look
> > like
> > > > the
> > > > > > > output of pcp_combine.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John Halley Gotway
> > > > > > > met_help at ucar.edu
> > > > > > >
> > > > > > >
> > > > > > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <
> > > met_help at ucar.edu>
> > > > > > > wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted
upon.
> > > > > > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > > > > > >        Queue: met_help
> > > > > > > >      Subject: help with file format for MODE v 5.2
> > > > > > > >        Owner: Nobody
> > > > > > > >   Requestors: jeffduda319 at gmail.com
> > > > > > > >       Status: new
> > > > > > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > > > Ticket/Display.html?id=79394
> > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > I am using MODE to verify precipitation forecasts. I
have an
> > > > > > observation
> > > > > > > > dataset (it's basically MRMS, but processed from the
raw
> stuff)
> > > in
> > > > > > netCDF
> > > > > > > > format, but MODE doesn't seem to like it when I use it
for
> > MODE.
> > > > > > > >
> > > > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > > >
/scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > > > > /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_
> 000000.nc
> > > > > > MODEConfig
> > > > > > > > DEBUG 1: Default Config File:
> > > > > > > >
/home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > > > ERROR  :
> > > > > > > > ERROR  : Trouble reading observation file
> > > > > "/condo/map/jdduda/MRMS_obs/
> > > > > > > > rain/
> > > > > > > > 1hr_precip_20160525_000000.nc"
> > > > > > > > ERROR  :
> > > > > > > >
> > > > > > > > Here is the dump for that file:
> > > > > > > >
> > > > > > > > netcdf \1hr_precip_20160525_000000 {
> > > > > > > > dimensions:
> > > > > > > >         x = 1568 ;
> > > > > > > >         y = 1120 ;
> > > > > > > >         z = 1 ;
> > > > > > > > variables:
> > > > > > > >         float hsr1hr(z, y, x) ;
> > > > > > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > > > > > >                 hsr1hr:long_name = "1-h precipitation"
;
> > > > > > > >                 hsr1hr:units = "mm" ;
> > > > > > > > }
> > > > > > > >
> > > > > > > > It is my understanding that this may not be "CF
compliant",
> > > > whatever
> > > > > > that
> > > > > > > > means. If not, what can I do to make it CF compliant,
or is
> it
> > > > > possible
> > > > > > > I'm
> > > > > > > > mis-specifying the part of the config file where this
is
> > > important.
> > > > > > Here
> > > > > > > is
> > > > > > > > the config file part:
> > > > > > > >
> > > > > > > > obs = {
> > > > > > > >    field = {
> > > > > > > >       name = "hsr1hr";
> > > > > > > >       level = "(*,*,*)";
> > > > > > > >
> > > > > > > >    raw_thresh        = NA;
> > > > > > > >    conv_radius       = 50.0/grid_res; // in grid
squares
> > > > > > > >    conv_thresh       = >=0.5;
> > > > > > > >    vld_thresh        = 0.5;
> > > > > > > >    area_thresh       = NA;
> > > > > > > >    inten_perc_value  = 100;
> > > > > > > >    inten_perc_thresh = NA;
> > > > > > > >    merge_thresh      = >=1.25;
> > > > > > > >    merge_flag        = ENGINE;
> > > > > > > >
> > > > > > > >    }
> > > > > > > > }
> > > > > > > >
> > > > > > > > I tried (0,*,*) and (1,*,*) for the level, but I got
the same
> > > error
> > > > > for
> > > > > > > > each.
> > > > > > > >
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > > Jeff Duda
> > > > > > > >
> > > > > > > > --
> > > > > > > > Jeff Duda
> > > > > > > > Post-doctoral research fellow
> > > > > > > > University of Oklahoma School of Meteorology
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > Jeff Duda
> > > > > > Post-doctoral research fellow
> > > > > > University of Oklahoma School of Meteorology
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Jeff Duda
> > > > Post-doctoral research fellow
> > > > University of Oklahoma School of Meteorology
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Jeff Duda
> > Post-doctoral research fellow
> > University of Oklahoma School of Meteorology
> >
> >
>
>


--
Jeff Duda
Post-doctoral research fellow
University of Oklahoma School of Meteorology

------------------------------------------------
Subject: help with file format for MODE v 5.2
From: John Halley Gotway
Time: Wed Feb 15 11:34:43 2017

Jeff,

Great, I'm glad you were able to get it up and running.  I'll resolve
this
ticket, but just let us know what issues or questions arise in your
use of
MET.

Thanks,
John

On Wed, Feb 15, 2017 at 9:40 AM, Jeff Duda via RT <met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
>
> John,
> Got it! Wasn't sure exactly which attributes I had to add, so I
added
> almost all of them. But it works now.
>
> Thanks!
>
> Jeff
>
> On Fri, Feb 10, 2017 at 8:26 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Jeff,
> >
> > Yes, you can either follow the cf-convention way of doing it using
a
> > grid_mapping attribute or make it look like the NetCDF output of
the
> > pcp_combine tool. For the latter, pass a GRIB file on that
projection
> > through pcp_combine and that'll describe exactly what you need.
> >
> > John
> >
> >
> > On Fri, Feb 10, 2017 at 5:16 PM Jeff Duda via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > >
> > > I know the projection details already. I take it I should insert
those
> > into
> > > the netcdf file when writing it out? Is there a specific syntax
I
> should
> > > use for the attributes?
> > >
> > > Jeff
> > >
> > > On Fri, Feb 10, 2017 at 5:25 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Jeff,
> > > >
> > > > Thanks for sending the sample data file.  Looking at the
latitude and
> > > > longitude values, I see that they're not evenly spaced.  And
that
> tells
> > > me
> > > > this is not on a simple lat/lon grid.  For regular lat/lon
grids, MET
> > can
> > > > use the lat/lon values to infer the underlying grid
information.
> > > However,
> > > > for other projection types, like lambert conformal, polar
> sterographic,
> > > and
> > > > mercator, there is no good way of processing lat/lon values to
back
> out
> > > the
> > > > underlying grid/projection information.
> > > >
> > > > Instead, we need to add the projection info in the NetCDF
file.  I
> see
> > > that
> > > > the data has dimension 1152 x 1680.
> > > >
> > > > Looking back in the message history I see a directory named
> "MRMS_obs".
> > > > But I checked the MRMS data from the NCEP website and see that
it
> uses
> > a
> > > > 7000x3500 lat/lon grid.  If you can point me to the
grid/projection
> > > > information, I should be able to help you out.  If you have
any GRIB1
> > or
> > > > GRIB2 files on the same grid, I could easily extract the grid
> > information
> > > > from those.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > >
> > > > On Thu, Feb 9, 2017 at 5:10 PM, Jeff Duda via RT
<met_help at ucar.edu>
> > > > wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394 >
> > > > >
> > > > > John,
> > > > > Okay, it's there.
> > > > >
> > > > > Jeff
> > > > >
> > > > > On Thu, Feb 9, 2017 at 5:02 PM, John Halley Gotway via RT <
> > > > > met_help at ucar.edu
> > > > > > wrote:
> > > > >
> > > > > > Jeff,
> > > > > >
> > > > > > Can you please post a sample NetCDF file to our anonymous
ftp
> site
> > > > > > following these instructions:
> > > > > >
http://www.dtcenter.org/met/users/support/met_help.php#ftp
> > > > > >
> > > > > > And I can take a closer look.  Please email me once it's
posted
> so
> > I
> > > > know
> > > > > > to go take a look.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > >
> > > > > > On Thu, Feb 9, 2017 at 3:41 PM, Jeff Duda via RT <
> > met_help at ucar.edu>
> > > > > > wrote:
> > > > > >
> > > > > > >
> > > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=79394
> >
> > > > > > >
> > > > > > > Thanks for the reply, John. I made some changes to my
system
> but
> > am
> > > > > still
> > > > > > > having problems. I looked at Tables 5.1 and 5.2 in the
User's
> > Guide
> > > > to
> > > > > > see
> > > > > > > what the format from the output of pcp_combine would be,
and I
> > > > > > reorganized
> > > > > > > my observation files to match that format. I think i did
it,
> but
> > I
> > > > get
> > > > > > the
> > > > > > > same error.
> > > > > > >
> > > > > > > Here is the ncdump for my obs file now:
> > > > > > > netcdf interp_compref_2016050101 {
> > > > > > > dimensions:
> > > > > > >         lat = 1152 ;
> > > > > > >         lon = 1680 ;
> > > > > > > variables:
> > > > > > >         float lat(lat, lon) ;
> > > > > > >         float lon(lat, lon) ;
> > > > > > >         float data(lat, lon) ;
> > > > > > > }
> > > > > > >
> > > > > > > And the MODE run command:
> > > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > > /scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > > > /scratch/jdduda/refl_201605/interp_compref_2016050101.nc
> > > MODEConfig
> > > > > > > DEBUG 1: Default Config File:
> > > > > > >
/home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > > ERROR  :
> > > > > > > ERROR  : Trouble reading observation file
> > > > "/scratch/jdduda/refl_201605/
> > > > > > > interp_compref_2016050101.nc"
> > > > > > > ERROR  :
> > > > > > >
> > > > > > > And the section of the config file:
> > > > > > > obs = {
> > > > > > >    field = {
> > > > > > >       name = "data";
> > > > > > >       level = "(*,*)";
> > > > > > >
> > > > > > >    raw_thresh        = NA;
> > > > > > >    conv_radius       = 50.0/grid_res; // in grid squares
> > > > > > >    conv_thresh       = >=0.5;
> > > > > > >    vld_thresh        = 0.5;
> > > > > > >    area_thresh       = NA;
> > > > > > >    inten_perc_value  = 100;
> > > > > > >    inten_perc_thresh = NA;
> > > > > > >    merge_thresh      = >=1.25;
> > > > > > >    merge_flag        = ENGINE;
> > > > > > >
> > > > > > >    }
> > > > > > > }
> > > > > > >
> > > > > > > Did I do this right yet?
> > > > > > >
> > > > > > > Jeff
> > > > > > >
> > > > > > > On Wed, Feb 8, 2017 at 5:41 PM, John Halley Gotway via
RT <
> > > > > > > met_help at ucar.edu
> > > > > > > > wrote:
> > > > > > >
> > > > > > > > Jeff,
> > > > > > > >
> > > > > > > > Yes, that file doesn't follow the NetCDF Climate
Forecast
> > > > conventions
> > > > > > > which
> > > > > > > > are described in this document:
> > > > > > > >    http://cfconventions.org/latest.html
> > > > > > > >
> > > > > > > > NetCDF is a very wide-open format, and it isn't really
> > realistic
> > > > for
> > > > > > > > software to be able to read them in general without
agreeing
> to
> > > > some
> > > > > > > ground
> > > > > > > > rules... and the CF conventions are an attempt to
establish
> > them.
> > > > In
> > > > > > > > particular, MET is looking for a definition of the
grid on
> > which
> > > > this
> > > > > > > data
> > > > > > > > resides and timing information.  And there are CF-
compliant
> > ways
> > > of
> > > > > > > > specifying those.
> > > > > > > >
> > > > > > > > Alternatively, you could make your NetCDF file look
like the
> > > NetCDF
> > > > > > > output
> > > > > > > > files that MET generates (i.e. like pcp_combine or
> > gen_vx_mask).
> > > > MET
> > > > > > > knows
> > > > > > > > how to read it's own output.  And you may find that
> simplistic
> > > > format
> > > > > > > > easier to follow than the CF-conventions.
> > > > > > > >
> > > > > > > > We do have some ideas about how to make this easier in
the
> > > future.
> > > > > > > Ideally
> > > > > > > > MET would be able to read data from your file as-is...
but
> then
> > > > > specify
> > > > > > > the
> > > > > > > > grid and timing information directly in the MODE
config file.
> > > > > > > > Unfortunately, the software is not set up to do that
yet.
> > > > > > > >
> > > > > > > > Just let us know if you get stuck making your NetCDF
file
> look
> > > like
> > > > > the
> > > > > > > > output of pcp_combine.
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John Halley Gotway
> > > > > > > > met_help at ucar.edu
> > > > > > > >
> > > > > > > >
> > > > > > > > On Wed, Feb 8, 2017 at 4:22 PM, Jeff Duda via RT <
> > > > met_help at ucar.edu>
> > > > > > > > wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > Wed Feb 08 16:22:45 2017: Request 79394 was acted
upon.
> > > > > > > > > Transaction: Ticket created by jeffduda319 at gmail.com
> > > > > > > > >        Queue: met_help
> > > > > > > > >      Subject: help with file format for MODE v 5.2
> > > > > > > > >        Owner: Nobody
> > > > > > > > >   Requestors: jeffduda319 at gmail.com
> > > > > > > > >       Status: new
> > > > > > > > >  Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > > > > Ticket/Display.html?id=79394
> > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > I am using MODE to verify precipitation forecasts. I
have
> an
> > > > > > > observation
> > > > > > > > > dataset (it's basically MRMS, but processed from the
raw
> > stuff)
> > > > in
> > > > > > > netCDF
> > > > > > > > > format, but MODE doesn't seem to like it when I use
it for
> > > MODE.
> > > > > > > > >
> > > > > > > > > [jdduda at schooner1 MODE]$ ./mode
> > > > > > > > >
/scratch/jdduda/NMMB_may24/NMMB_CA_00Z_free_d01_00.grib2
> > > > > > > > > /condo/map/jdduda/MRMS_obs/rain/1hr_precip_20160525_
> > 000000.nc
> > > > > > > MODEConfig
> > > > > > > > > DEBUG 1: Default Config File:
> > > > > > > > >
/home/jdduda/tool/met/share/met/config/MODEConfig_default
> > > > > > > > > DEBUG 1: Match Config File: MODEConfig
> > > > > > > > > DEBUG 1: Merge Config File: MODEConfig
> > > > > > > > > ERROR  :
> > > > > > > > > ERROR  : Trouble reading observation file
> > > > > > "/condo/map/jdduda/MRMS_obs/
> > > > > > > > > rain/
> > > > > > > > > 1hr_precip_20160525_000000.nc"
> > > > > > > > > ERROR  :
> > > > > > > > >
> > > > > > > > > Here is the dump for that file:
> > > > > > > > >
> > > > > > > > > netcdf \1hr_precip_20160525_000000 {
> > > > > > > > > dimensions:
> > > > > > > > >         x = 1568 ;
> > > > > > > > >         y = 1120 ;
> > > > > > > > >         z = 1 ;
> > > > > > > > > variables:
> > > > > > > > >         float hsr1hr(z, y, x) ;
> > > > > > > > >                 hsr1hr:standard_name = "hsr1hr" ;
> > > > > > > > >                 hsr1hr:long_name = "1-h
precipitation" ;
> > > > > > > > >                 hsr1hr:units = "mm" ;
> > > > > > > > > }
> > > > > > > > >
> > > > > > > > > It is my understanding that this may not be "CF
compliant",
> > > > > whatever
> > > > > > > that
> > > > > > > > > means. If not, what can I do to make it CF
compliant, or is
> > it
> > > > > > possible
> > > > > > > > I'm
> > > > > > > > > mis-specifying the part of the config file where
this is
> > > > important.
> > > > > > > Here
> > > > > > > > is
> > > > > > > > > the config file part:
> > > > > > > > >
> > > > > > > > > obs = {
> > > > > > > > >    field = {
> > > > > > > > >       name = "hsr1hr";
> > > > > > > > >       level = "(*,*,*)";
> > > > > > > > >
> > > > > > > > >    raw_thresh        = NA;
> > > > > > > > >    conv_radius       = 50.0/grid_res; // in grid
squares
> > > > > > > > >    conv_thresh       = >=0.5;
> > > > > > > > >    vld_thresh        = 0.5;
> > > > > > > > >    area_thresh       = NA;
> > > > > > > > >    inten_perc_value  = 100;
> > > > > > > > >    inten_perc_thresh = NA;
> > > > > > > > >    merge_thresh      = >=1.25;
> > > > > > > > >    merge_flag        = ENGINE;
> > > > > > > > >
> > > > > > > > >    }
> > > > > > > > > }
> > > > > > > > >
> > > > > > > > > I tried (0,*,*) and (1,*,*) for the level, but I got
the
> same
> > > > error
> > > > > > for
> > > > > > > > > each.
> > > > > > > > >
> > > > > > > > > Thanks.
> > > > > > > > >
> > > > > > > > > Jeff Duda
> > > > > > > > >
> > > > > > > > > --
> > > > > > > > > Jeff Duda
> > > > > > > > > Post-doctoral research fellow
> > > > > > > > > University of Oklahoma School of Meteorology
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > --
> > > > > > > Jeff Duda
> > > > > > > Post-doctoral research fellow
> > > > > > > University of Oklahoma School of Meteorology
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > > --
> > > > > Jeff Duda
> > > > > Post-doctoral research fellow
> > > > > University of Oklahoma School of Meteorology
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Jeff Duda
> > > Post-doctoral research fellow
> > > University of Oklahoma School of Meteorology
> > >
> > >
> >
> >
>
>
> --
> Jeff Duda
> Post-doctoral research fellow
> University of Oklahoma School of Meteorology
>
>

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


More information about the Met_help mailing list