[Met_help] [rt.rap.ucar.edu #42668] History for queries regarding MODE component of METv3.0 software
RAL HelpDesk {for Paul Oldenburg}
met_help at ucar.edu
Tue Dec 28 12:50:53 MST 2010
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hi,
I am a Meteorologist at the India Meteorological Department, New Delhi.
I wish to use the MET software (more specifically the MODE component) to validate the output of the Nowcasting softwares that were operationally run during the Commonwealth Games, 2010 at Delhi. I have the following queries, and will be grateful if you can help in this regard.
1. My data is reflectivity values in a 1 km x 1 km domain over a part of the Indian region. Projection is lat-lon. i checked the link http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html
But am unable to decide which of the grids to use. can you please help.
2. I need an example netcdf file of a similar quantity in lat-lon projection that i can input into the MODE software. the examples files provided with the installation package are all in lambert conformal projection. I need a file in whose model i will convert my files, to input into the MODE package.
3. Any other corrections i need to make?
thank you and regards
Dr. Soma Sen Roy
Meteorologist
India Meteorological Department
Delhi 110034
India
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: queries regarding MODE component of METv3.0 software
From: John Halley Gotway
Time: Fri Dec 10 09:59:19 2010
Dr. Soma Sen Roy,
Based on your first question, it sounds like there's some
misunderstanding about what type of gridded data the MET tools can
support. You are not limited to just the grids defined in that GRIB
documentation
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html). Instead,
MET parses the grid definition information from the input GRIB or
NetCDF files and instantiates the grid those
parameters define. So really, MET can handle any regular grid defined
on a polar stereographic, lat/lon, lambert conformal, or mercator
projection.
The copygb tool is useful for regridding data in GRIB format.
However, it sounds like you're dealing with NetCDF data and would like
to set it up for the MET tools to read. In order to do that, I'd
suggest making it look like the output of the MET/PCP-Combine tool.
That way, MODE should be able to read it.
Using the commands listed below, I ran some sample data through copygb
to regrid it to a lat/lon grid, and then I ran it through pcp_combine
to get it into NetCDF format:
cd METv3.0/data/sample_obs/ST2ml
copygb -xg"255 0 20 20 30000 -110000 128 50000 -90000 1000 1000 64"
ST2ml2005080800.Grb_G212 ST2ml2005080800.Grb_latlon
../../../bin/pcp_combine -add ST2ml2005080800.Grb_latlon 01
ST2ml2005080800.Grb_latlon.nc
The resulting NetCDF file is attached. And here's the header for that
file:
netcdf ST2ml2005080800.Grb_latlon {
dimensions:
lat = 20 ;
lon = 20 ;
variables:
float lat(lat, lon) ;
lat:long_name = "latitude" ;
lat:units = "degrees_north" ;
lat:standard_name = "latitude" ;
float lon(lat, lon) ;
lon:long_name = "longitude" ;
lon:units = "degrees_east" ;
lon:standard_name = "longitude" ;
float APCP_01(lat, lon) ;
APCP_01:name = "APCP" ;
APCP_01:long_name = "Total precipitation" ;
APCP_01:level = "A1" ;
APCP_01:units = "kg/m^2" ;
APCP_01:grib_code = 61 ;
APCP_01:_FillValue = -9999.f ;
APCP_01:init_time = "20050807_230000" ;
APCP_01:init_time_ut = 1123455600 ;
APCP_01:valid_time = "20050808_000000" ;
APCP_01:valid_time_ut = 1123459200 ;
APCP_01:accum_time = "010000" ;
APCP_01:accum_time_sec = 3600 ;
// global attributes:
:FileOrigins = "File ST2ml2005080800.Grb_latlon.nc
generated 20101210_165425 UTC on host rambler by the MET pcp_combine
tool" ;
:MET_version = "V3.0" ;
:MET_tool = "pcp_combine" ;
:RunCommand = "Addition: 1 files." ;
:Projection = "LatLon" ;
:lat_ll = "30.000000 degrees_north" ;
:lon_ll = "-110.000000 degrees_east" ;
:delta_lat = "1.000000 degrees" ;
:delta_lon = "1.000000 degrees" ;
:Nlat = "20 grid_points" ;
:Nlon = "20 grid_points" ;
}
Specifically, you're NetCDF file should contain the same type of
global attributes this one does:
MET_version, Projection, lat_ll, lon_ll, delta_lat, delta_lon,
Nlat, and Nlon.
These define the grid on which your data resides.
Next, your NetCDF file should use dimensions named lat and lon. You
do NOT need to include the lat(lat,lon) and lon(lat,lon) variables.
But each you variables containing the reflectivity data should
be sized using the (lat, lon) dimensions. Also, they should include
the variable attributes listed above. The attributes ending in "_ut"
are for unix time - the number of seconds since Jan 1, 1970.
Hope this helps. Let me know if you get stuck.
Thanks,
John Halley Gotway
On 12/10/2010 04:42 AM, RAL HelpDesk {for soma senroy} wrote:
>
> Fri Dec 10 04:42:21 2010: Request 42668 was acted upon.
> Transaction: Ticket created by somasenroy at yahoo.com
> Queue: met_help
> Subject: queries regarding MODE component of METv3.0 software
> Owner: Nobody
> Requestors: somasenroy at yahoo.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=42668 >
>
>
> Hi,
> I am a Meteorologist at the India Meteorological Department, New
Delhi.
> I wish to use the MET software (more specifically the MODE
component) to validate the output of the Nowcasting softwares that
were operationally run during the Commonwealth Games, 2010 at Delhi. I
have the following queries, and will be grateful if you can help in
this regard.
>
> 1. My data is reflectivity values in a 1 km x 1 km domain over a
part of the Indian region. Projection is lat-lon. i checked the link
http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html
> But am unable to decide which of the grids to use. can you please
help.
>
> 2. I need an example netcdf file of a similar quantity in lat-lon
projection that i can input into the MODE software. the examples files
provided with the installation package are all in lambert conformal
projection. I need a file in whose model i will convert my files, to
input into the MODE package.
> 3. Any other corrections i need to make?
>
> thank you and regards
>
> Dr. Soma Sen Roy
> Meteorologist
> India Meteorological Department
> Delhi 110034
> India
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #42668] queries regarding MODE component of METv3.0 software
From: soma senroy
Time: Mon Dec 13 01:43:23 2010
Dear Dr. Gotway,
Thank you for the prompt reply. I will get back to you I have any
other queries.
regards
soma sen roy
--- On Fri, 10/12/10, RAL HelpDesk {for John Halley Gotway}
<met_help at ucar.edu> wrote:
> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
> Subject: Re: [rt.rap.ucar.edu #42668] queries regarding MODE
component of METv3.0 software
> To: somasenroy at yahoo.com
> Date: Friday, 10 December, 2010, 11:59 AM
> Dr. Soma Sen Roy,
>
> Based on your first question, it sounds like there's some
> misunderstanding about what type of gridded data the MET
> tools can support. You are not limited to just the
> grids defined in that GRIB
> documentation
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html).
> Instead,
MET parses the grid definition information from the
> input GRIB or NetCDF files and instantiates the grid those
> parameters define. So really, MET can handle any
> regular grid defined on a polar stereographic, lat/lon,
> lambert conformal, or mercator projection.
>
> The copygb tool is useful for regridding data in GRIB
> format. However, it sounds like you're dealing with
> NetCDF data and would like to set it up for the MET tools to
> read. In order to do that, I'd
> suggest making it look like the output of the
> MET/PCP-Combine tool. That way, MODE should be able to
> read it.
>
> Using the commands listed below, I ran some sample data
> through copygb to regrid it to a lat/lon grid, and then I
> ran it through pcp_combine to get it into NetCDF format:
>
> cd METv3.0/data/sample_obs/ST2ml
> copygb -xg"255 0 20 20 30000 -110000 128 50000 -90000 1000
> 1000 64" ST2ml2005080800.Grb_G212
> ST2ml2005080800.Grb_latlon
> ../../../bin/pcp_combine -add ST2ml2005080800.Grb_latlon 01
> ST2ml2005080800.Grb_latlon.nc
>
> The resulting NetCDF file is attached. And here's the
> header for that file:
>
> netcdf ST2ml2005080800.Grb_latlon {
> dimensions:
> lat = 20 ;
> lon = 20 ;
> variables:
> float lat(lat, lon) ;
>
> lat:long_name = "latitude" ;
>
> lat:units = "degrees_north" ;
>
> lat:standard_name = "latitude" ;
> float lon(lat, lon) ;
>
> lon:long_name = "longitude" ;
>
> lon:units = "degrees_east" ;
>
> lon:standard_name = "longitude" ;
> float APCP_01(lat, lon) ;
>
> APCP_01:name = "APCP" ;
>
> APCP_01:long_name = "Total precipitation" ;
>
> APCP_01:level = "A1" ;
>
> APCP_01:units = "kg/m^2" ;
>
> APCP_01:grib_code = 61 ;
>
> APCP_01:_FillValue = -9999.f ;
>
> APCP_01:init_time = "20050807_230000" ;
>
> APCP_01:init_time_ut = 1123455600 ;
>
> APCP_01:valid_time = "20050808_000000" ;
>
> APCP_01:valid_time_ut = 1123459200 ;
>
> APCP_01:accum_time = "010000" ;
>
> APCP_01:accum_time_sec = 3600 ;
>
> // global attributes:
>
> :FileOrigins = "File ST2ml2005080800.Grb_latlon.nc
generated
> 20101210_165425 UTC on host rambler by the MET pcp_combine
> tool" ;
>
> :MET_version = "V3.0" ;
>
> :MET_tool = "pcp_combine" ;
>
> :RunCommand = "Addition: 1 files." ;
>
> :Projection = "LatLon" ;
>
> :lat_ll = "30.000000 degrees_north" ;
>
> :lon_ll = "-110.000000 degrees_east" ;
>
> :delta_lat = "1.000000 degrees" ;
>
> :delta_lon = "1.000000 degrees" ;
>
> :Nlat = "20 grid_points" ;
>
> :Nlon = "20 grid_points" ;
> }
>
>
> Specifically, you're NetCDF file should contain the same
> type of global attributes this one does:
> MET_version, Projection, lat_ll, lon_ll,
> delta_lat, delta_lon, Nlat, and Nlon.
> These define the grid on which your data resides.
>
> Next, your NetCDF file should use dimensions named lat and
> lon. You do NOT need to include the lat(lat,lon) and
> lon(lat,lon) variables. But each you variables
> containing the reflectivity data should
> be sized using the (lat, lon) dimensions. Also, they
> should include the variable attributes listed above.
> The
attributes ending in "_ut" are for unix time - the
> number of seconds since Jan 1, 1970.
>
> Hope this helps. Let me know if you get stuck.
>
> Thanks,
> John Halley Gotway
>
>
>
>
> On 12/10/2010 04:42 AM, RAL HelpDesk {for soma senroy}
> wrote:
> >
> > Fri Dec 10 04:42:21 2010: Request 42668 was acted
> upon.
> > Transaction: Ticket created by somasenroy at yahoo.com
> > Queue: met_help
> > Subject: queries regarding MODE
> component of METv3.0 software
> > Owner: Nobody
> > Requestors: somasenroy at yahoo.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=42668
> >
> >
> >
> > Hi,
> > I am a Meteorologist at the India Meteorological
> Department, New Delhi.
> > I wish to use the MET software (more specifically the
> MODE component) to validate the output of the Nowcasting
> softwares that were operationally run during the
> Commonwealth Games, 2010 at Delhi. I have the following
> queries, and will be grateful if you can help in this
> regard.
> >
> > 1. My data is reflectivity values in a 1 km x 1 km
> domain over a part of the Indian region. Projection is
> lat-lon. i checked the link
http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html
> > But am unable to decide which of the grids to use. can
> you please help.
> >
> > 2. I need an example netcdf file of a similar quantity
> in lat-lon projection that i can input into the MODE
> software. the examples files provided with the installation
> package are all in lambert conformal projection. I need a
> file in whose model i will convert my files, to input into
> the MODE package.
> > 3. Any other corrections i need to make?
> >
> > thank you and regards
> >
> > Dr. Soma Sen Roy
> > Meteorologist
> > India Meteorological Department
> > Delhi 110034
> > India
> >
> >
>
>
------------------------------------------------
Subject: queries regarding MODE component of METv3.0 software
From: soma senroy
Time: Fri Dec 17 04:26:07 2010
Dear Dr. Gotway,
i have one additional question with regards the data. Please see the
postscript file that i have generated and the gif file generated for
the same time in NCL graphics (0928 UTC or 1458 IST). The picture in
the ps file appears to be rotated. i would like to know:
Is the whole map as plotted in the .ps file rotated i.e. plotted for a
landscape view? or is it my mistake in plotting the latitudes as
longitudes and viceversa.
I ask this because there is no lat-long marking on the pictures of the
.ps files.
Thank you and regards
soma sen roy
--- On Mon, 13/12/10, RAL HelpDesk {for John Halley Gotway}
<met_help at ucar.edu> wrote:
> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
> Subject: [rt.rap.ucar.edu #42668] Resolved: queries regarding MODE
component of METv3.0 software
> To: somasenroy at yahoo.com
> Date: Monday, 13 December, 2010, 11:04 AM
> According to our records, your
> request has been resolved. If you have any
> further questions or concerns, please respond to this
> message.
>
Dear Dr. Gotway,
Thank you for the prompt reply. I will get back to you I have any
other queries.
regards
soma sen roy
--- On Fri, 10/12/10, RAL HelpDesk {for John Halley Gotway}
<met_help at ucar.edu> wrote:
> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
> Subject: Re: [rt.rap.ucar.edu #42668] queries regarding MODE
component of METv3.0 software
> To: somasenroy at yahoo.com
> Date: Friday, 10 December, 2010, 11:59 AM
> Dr. Soma Sen Roy,
>
> Based on your first question, it sounds like there's some
> misunderstanding about what type of gridded data the MET
> tools can support. You are not limited to just the
> grids defined in that GRIB
> documentation
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html).
> Instead, MET parses the grid definition information from the
> input GRIB or NetCDF files and instantiates the grid those
> parameters define. So really, MET can handle any
> regular grid defined on a polar stereographic, lat/lon,
> lambert conformal, or mercator projection.
>
> The copygb tool is useful for regridding data in GRIB
> format. However, it sounds like you're dealing with
> NetCDF data and would like to set it up for the MET tools to
> read. In order to do that, I'd
> suggest making it look like the output of the
> MET/PCP-Combine tool. That way, MODE should be able to
> read it.
>
> Using the commands listed below, I ran some sample data
> through copygb to regrid it to a lat/lon grid, and then I
> ran it through pcp_combine to get it into NetCDF format:
>
> cd METv3.0/data/sample_obs/ST2ml
> copygb -xg"255 0 20 20 30000 -110000 128 50000 -90000 1000
> 1000 64" ST2ml2005080800.Grb_G212
> ST2ml2005080800.Grb_latlon
> ../../../bin/pcp_combine -add ST2ml2005080800.Grb_latlon 01
> ST2ml2005080800.Grb_latlon.nc
>
> The resulting NetCDF file is attached. And here's the
> header for that file:
>
> netcdf ST2ml2005080800.Grb_latlon {
> dimensions:
> lat = 20 ;
> lon = 20 ;
> variables:
> float lat(lat, lon) ;
>
> lat:long_name = "latitude" ;
>
> lat:units = "degrees_north" ;
>
> lat:standard_name = "latitude" ;
> float lon(lat, lon) ;
>
> lon:long_name = "longitude" ;
>
> lon:units = "degrees_east" ;
>
> lon:standard_name = "longitude" ;
> float APCP_01(lat, lon) ;
>
> APCP_01:name = "APCP" ;
>
> APCP_01:long_name = "Total precipitation" ;
>
> APCP_01:level = "A1" ;
>
> APCP_01:units = "kg/m^2" ;
>
> APCP_01:grib_code = 61 ;
>
> APCP_01:_FillValue = -9999.f ;
>
> APCP_01:init_time = "20050807_230000" ;
>
> APCP_01:init_time_ut = 1123455600 ;
>
> APCP_01:valid_time = "20050808_000000" ;
>
> APCP_01:valid_time_ut = 1123459200 ;
>
> APCP_01:accum_time = "010000" ;
>
> APCP_01:accum_time_sec = 3600 ;
>
> // global attributes:
>
> :FileOrigins = "File ST2ml2005080800.Grb_latlon.nc generated
> 20101210_165425 UTC on host rambler by the MET pcp_combine
> tool" ;
>
> :MET_version = "V3.0" ;
>
> :MET_tool = "pcp_combine" ;
>
> :RunCommand = "Addition: 1 files." ;
>
> :Projection = "LatLon" ;
>
> :lat_ll = "30.000000 degrees_north" ;
>
> :lon_ll = "-110.000000 degrees_east" ;
>
> :delta_lat = "1.000000 degrees" ;
>
> :delta_lon = "1.000000 degrees" ;
>
> :Nlat = "20 grid_points" ;
>
> :Nlon = "20 grid_points" ;
> }
>
>
> Specifically, you're NetCDF file should contain the same
> type of global attributes this one does:
> MET_version, Projection, lat_ll, lon_ll,
> delta_lat, delta_lon, Nlat, and Nlon.
> These define the grid on which your data resides.
>
> Next, your NetCDF file should use dimensions named lat and
> lon. You do NOT need to include the lat(lat,lon) and
> lon(lat,lon) variables. But each you variables
> containing the reflectivity data should
> be sized using the (lat, lon) dimensions. Also, they
> should include the variable attributes listed above.
> The attributes ending in "_ut" are for unix time - the
> number of seconds since Jan 1, 1970.
>
> Hope this helps. Let me know if you get stuck.
>
> Thanks,
> John Halley Gotway
>
>
>
>
> On 12/10/2010 04:42 AM, RAL HelpDesk {for soma senroy}
> wrote:
> >
> > Fri Dec 10 04:42:21 2010: Request 42668 was acted
> upon.
> > Transaction: Ticket created by somasenroy at yahoo.com
> > Queue: met_help
> > Subject: queries regarding MODE
> component of METv3.0 software
> > Owner: Nobody
> > Requestors: somasenroy at yahoo.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=42668
> >
> >
> >
> > Hi,
> > I am a Meteorologist at the India Meteorological
> Department, New Delhi.
> > I wish to use the MET software (more specifically the
> MODE component) to validate the output of the Nowcasting
> softwares that were operationally run during the
> Commonwealth Games, 2010 at Delhi. I have the following
> queries, and will be grateful if you can help in this
> regard.
> >
> > 1. My data is reflectivity values in a 1 km x 1 km
> domain over a part of the Indian region. Projection is
> lat-lon. i checked the link
http://www.nco.ncep.noaa.gov/pmb/docs/on388/tableb.html
> > But am unable to decide which of the grids to use. can
> you please help.
> >
> > 2. I need an example netcdf file of a similar quantity
> in lat-lon projection that i can input into the MODE
> software. the examples files provided with the installation
> package are all in lambert conformal projection. I need a
> file in whose model i will convert my files, to input into
> the MODE package.
> > 3. Any other corrections i need to make?
> >
> > thank you and regards
> >
> > Dr. Soma Sen Roy
> > Meteorologist
> > India Meteorological Department
> > Delhi 110034
> > India
> >
> >
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding MODE component of METv3.0 software
From: John Halley Gotway
Time: Fri Dec 17 16:00:29 2010
soma sen roy,
I do not believe there is any sort of rotation occurring in the
PostScript plot. You claim that these images are for the same time
but the timestamps on the images don't seem to line up.
In your NCL plot, I see a timestamp of 20101022 at 1458 IST. However,
in your PostScript plot from MODE, I see a timestamp of 20101022 at
09:28:44. So there appears to be at least a 30 minute offset
- 58 versus 28 minutes. Please try to take a look at 2 images with
*exactly* the same timestamp to determine whether or not MET is
reading the data correctly. If you still see a problem in how the
data lines up, please send us some sample data to work with here.
We'd need:
(1) Your forecast file for MODE.
(2) Your observation file for MODE.
(3) Your MODE configuration file.
(4) Your NCL output image and PostScript output file showing the
difference.
You can post that data to our anonymous ftp site following the
directions listed here:
http://www.dtcenter.org/met/users/support/met_help.php#ftp
If you do send us some data, please write me back to let me know when
it's there.
John
On 12/17/2010 04:26 AM, RAL HelpDesk {for soma senroy} wrote:
> i have one add
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding MODE component of METv3.0 software
From: soma senroy
Time: Tue Dec 21 05:48:05 2010
Dear Dr. Gotway,
i have created a folder senroy_data in the incoming/irap/met_help
location of your ftp site ftp.rap.ucar.edu.
This contains a tar-gzipped folder ssr.tgz containing the following:
1. An NCL plot for the data of 0928 UTC. The timing may have confused
you earlier since (IST= UTC+ 0530 hrs). So this is the picture of 1458
hrs IST (0928 hrs UTC) for the observation data file.
2. A 60 minute forecast file for input to mode : S20101022-
082756_60.nc
3. An observation data file for input to mode : S20101022-092844_00.nc
4. Mode config file: WrfModeConfig_modif
5. An output postscript file:
mode_000000L_20101022_082756V_010000A_new.ps
I had discovered a mistake in my writing of the data in the netcdf
file, but that has been corrected. However, i am still getting a
latitudinal flip in the image in the postscript file. I just want to
confirm, if there is still a bug in my program, or is it just that the
plotting is wrong in the ps file. It will be helpful if there are some
lat-lon markings in the postscript file.
I also find that this command takes a long time to execute (on an IBM-
AIX platform) - about an hour or so, when i change the grid_res to 1
instead of the default value in the WrfModeConfig_modif file. Is it
normal? I need to compare maybe 100 file sets at a go (for a typical
rainfall case). That will take ages then...
Also Any other parameter i need to change in this file (e.g.
fcst_conv_radius) since the reflectivity field is more irregular than
an average precipitation field?
Hope this helps. thank you and seasons greeting to you.
regards
soma
--- On Fri, 17/12/10, RAL HelpDesk {for John Halley Gotway}
<met_help at ucar.edu> wrote:
> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
> Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding
MODE component of METv3.0 software
> To: somasenroy at yahoo.com
> Date: Friday, 17 December, 2010, 6:00 PM
> soma sen roy,
>
> I do not believe there is any sort of rotation occurring in
> the PostScript plot. You claim that these images are
> for the same time but the timestamps on the images don't
> seem to line up.
>
> In your NCL plot, I see a timestamp of 20101022 at 1458
> IST. However, in your PostScript plot from MODE, I see
> a timestamp of 20101022 at 09:28:44. So there appears
> to be at least a 30 minute offset
> - 58 versus 28 minutes. Please try to take a look at
> 2 images with *exactly* the same timestamp to determine
> whether or not MET is reading the data correctly. If
> you still see a problem in how the
> data lines up, please send us some sample data to work with
> here. We'd need:
> (1) Your forecast file for MODE.
> (2) Your observation file for MODE.
> (3) Your MODE configuration file.
> (4) Your NCL output image and PostScript output file
> showing the difference.
>
> You can post that data to our anonymous ftp site following
> the directions listed here:
> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>
> If you do send us some data, please write me back to let me
> know when it's there.
>
> John
>
> On 12/17/2010 04:26 AM, RAL HelpDesk {for soma senroy}
> wrote:
> > i have one add
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding MODE component of METv3.0 software
From: Paul Oldenburg
Time: Wed Dec 22 14:03:28 2010
Soma,
Regarding the longitudinal flip that you see, MET uses the typical
convention of negative west longitude and positive
east longitude. Please make sure that you adhere to those conventions
in your own software.
Regarding MODE performance, the engineer responsible for MODE made
these suggestions:
(1) "fcst_area_thresh" and "obs_area_thresh" could be set
to some nonzero value. This will tell MODE to toss
small objects from the respective fields, so it won't
waste time applying the matching and merging algorithms
to them. Often the user is only interested in the
medium-to-large objects, so tossing the small ones can
really help.
(2) "match_flag" could be set to 3, rather than 1. This will
prevent any further merging during the fuzzy-logic matching
process. Double-threshold merging is (usually) fastest.
(3) The user could mask the verification region to some small
area that they're interested in, rather than doing the
whole grid. I see from the config file that no masking
was done.
Generally, parameters like total interest threshold, convolution
radius and threshold, etc., will affect the algorithm run time, but
the user should be setting these so as to get objects and matches
that look reasonable to him/her, and not tuning them for run-time
speed.
Please let us know if you have any questions.
Thanks,
Paul
On 12/21/2010 05:48 AM, RAL HelpDesk {for soma senroy} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=42668 >
>
> Dear Dr. Gotway,
> i have created a folder senroy_data in the incoming/irap/met_help
location of your ftp site ftp.rap.ucar.edu.
> This contains a tar-gzipped folder ssr.tgz containing the following:
> 1. An NCL plot for the data of 0928 UTC. The timing may have
confused you earlier since (IST= UTC+ 0530 hrs). So this is the
picture of 1458 hrs IST (0928 hrs UTC) for the observation data file.
> 2. A 60 minute forecast file for input to mode : S20101022-
082756_60.nc
> 3. An observation data file for input to mode : S20101022-
092844_00.nc
> 4. Mode config file: WrfModeConfig_modif
> 5. An output postscript file:
mode_000000L_20101022_082756V_010000A_new.ps
>
> I had discovered a mistake in my writing of the data in the netcdf
file, but that has been corrected. However, i am still getting a
latitudinal flip in the image in the postscript file. I just want to
confirm, if there is still a bug in my program, or is it just that the
plotting is wrong in the ps file. It will be helpful if there are some
lat-lon markings in the postscript file.
>
> I also find that this command takes a long time to execute (on an
IBM-AIX platform) - about an hour or so, when i change the grid_res to
1 instead of the default value in the WrfModeConfig_modif file. Is it
normal? I need to compare maybe 100 file sets at a go (for a typical
rainfall case). That will take ages then...
> Also Any other parameter i need to change in this file (e.g.
fcst_conv_radius) since the reflectivity field is more irregular than
an average precipitation field?
>
> Hope this helps. thank you and seasons greeting to you.
> regards
> soma
>
>
> --- On Fri, 17/12/10, RAL HelpDesk {for John Halley Gotway}
<met_help at ucar.edu> wrote:
>
>> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
>> Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding
MODE component of METv3.0 software
>> To: somasenroy at yahoo.com
>> Date: Friday, 17 December, 2010, 6:00 PM
>> soma sen roy,
>>
>> I do not believe there is any sort of rotation occurring in
>> the PostScript plot. You claim that these images are
>> for the same time but the timestamps on the images don't
>> seem to line up.
>>
>> In your NCL plot, I see a timestamp of 20101022 at 1458
>> IST. However, in your PostScript plot from MODE, I see
>> a timestamp of 20101022 at 09:28:44. So there appears
>> to be at least a 30 minute offset
>> - 58 versus 28 minutes. Please try to take a look at
>> 2 images with *exactly* the same timestamp to determine
>> whether or not MET is reading the data correctly. If
>> you still see a problem in how the
>> data lines up, please send us some sample data to work with
>> here. We'd need:
>> (1) Your forecast file for MODE.
>> (2) Your observation file for MODE.
>> (3) Your MODE configuration file.
>> (4) Your NCL output image and PostScript output file
>> showing the difference.
>>
>> You can post that data to our anonymous ftp site following
>> the directions listed here:
>> http://www.dtcenter.org/met/users/support/met_help.php#ftp
>>
>> If you do send us some data, please write me back to let me
>> know when it's there.
>>
>> John
>>
>> On 12/17/2010 04:26 AM, RAL HelpDesk {for soma senroy}
>> wrote:
>>> i have one add
>>
>>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding MODE component of METv3.0 software
From: soma senroy
Time: Thu Dec 23 04:29:30 2010
Hi Paul,
I will look into my latitudinal flip.
and thank you for the additional information about mode.
regards
soma
--- On Wed, 22/12/10, RAL HelpDesk {for Paul Oldenburg}
<met_help at ucar.edu> wrote:
> From: RAL HelpDesk {for Paul Oldenburg} <met_help at ucar.edu>
> Subject: Re: [rt.rap.ucar.edu #42668] Resolved: queries regarding
MODE component of METv3.0 software
> To: somasenroy at yahoo.com
> Date: Wednesday, 22 December, 2010, 4:03 PM
> Soma,
>
> Regarding the longitudinal flip that you see, MET uses the
> typical convention of negative west longitude and positive
> east longitude. Please make sure that you adhere to
> those conventions in your own software.
>
> Regarding MODE performance, the engineer responsible for
> MODE made these suggestions:
>
> (1) "fcst_area_thresh" and "obs_area_thresh" could be
> set
> to some nonzero value. This
> will tell MODE to toss
> small objects from the respective
> fields, so it won't
> waste time applying the matching
> and merging algorithms
> to them. Often the user is
> only interested in the
> medium-to-large objects, so
> tossing the small ones can
> really help.
>
> (2) "match_flag" could be set to 3, rather than
> 1. This will
> prevent any further merging during
> the fuzzy-logic matching
> process. Double-threshold
> merging is (usually) fastest.
>
> (3) The user could mask the verification region to
> some small
> area that they're interested in,
> rather than doing the
> whole grid. I see from the
> config file that no masking
> was done.
>
> Generally, parameters like total interest threshold,
> convolution
> radius and threshold, etc., will affect the algorithm run
> time, but
> the user should be setting these so as to get objects and
> matches
> that look reasonable to him/her, and not tuning them for
> run-time
> speed.
>
> Please let us know if you have any questions.
>
> Thanks,
>
> Paul
>
>
>
> On 12/21/2010 05:48 AM, RAL HelpDesk {for soma senroy}
> wrote:
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=42668
> >
> >
> > Dear Dr. Gotway,
> > i have created a folder senroy_data in the
> incoming/irap/met_help location of your ftp site
> ftp.rap.ucar.edu.
> > This contains a tar-gzipped folder ssr.tgz containing
> the following:
> > 1. An NCL plot for the data of 0928 UTC. The timing
> may have confused you earlier since (IST= UTC+ 0530 hrs). So
> this is the picture of 1458 hrs IST (0928 hrs UTC) for the
> observation data file.
> > 2. A 60 minute forecast file for input to mode :
> S20101022-
082756_60.nc
> > 3. An observation data file for input to mode :
> S20101022-092844_00.nc
> > 4. Mode config file: WrfModeConfig_modif
> > 5. An output postscript file:
> mode_000000L_20101022_082756V_010000A_new.ps
> >
> > I had discovered a mistake in my writing of the data
> in the netcdf file, but that has been corrected. However, i
> am still getting a latitudinal flip in the image in the
> postscript file. I just want to confirm, if there is still a
> bug in my program, or is it just that the plotting is wrong
> in the ps file. It will be helpful if there are some lat-lon
> markings in the postscript file.
> >
> > I also find that this command takes a long time to
> execute (on an IBM-AIX platform) - about an hour or so, when
> i change the grid_res to 1 instead of the default value in
> the WrfModeConfig_modif file. Is it normal? I need to
> compare maybe 100 file sets at a go (for a typical rainfall
> case). That will take ages then...
> > Also Any other parameter i need to change in this file
> (e.g. fcst_conv_radius) since the reflectivity field
> is more irregular than an average precipitation field?
> >
> > Hope this helps. thank you and seasons greeting to
> you.
> > regards
> > soma
> >
> >
> > --- On Fri, 17/12/10, RAL HelpDesk {for John Halley
> Gotway} <met_help at ucar.edu>
> wrote:
> >
> >> From: RAL HelpDesk {for John Halley Gotway} <met_help at ucar.edu>
> >> Subject: Re: [rt.rap.ucar.edu #42668] Resolved:
> queries regarding MODE component of METv3.0 software
> >> To: somasenroy at yahoo.com
> >> Date: Friday, 17 December, 2010, 6:00 PM
> >> soma sen roy,
> >>
> >> I do not believe there is any sort of rotation
> occurring in
> >> the PostScript plot. You claim that these
> images are
> >> for the same time but the timestamps on the images
> don't
> >> seem to line up.
> >>
> >> In your NCL plot, I see a timestamp of 20101022 at
> 1458
> >> IST. However, in your PostScript plot from
> MODE, I see
> >> a timestamp of 20101022 at 09:28:44. So
> there appears
> >> to be at least a 30 minute offset
> >> - 58 versus 28 minutes. Please try to take a
> look at
> >> 2 images with *exactly* the same timestamp to
> determine
> >> whether or not MET is reading the data
> correctly. If
> >> you still see a problem in how the
> >> data lines up, please send us some sample data to
> work with
> >> here. We'd need:
> >> (1) Your forecast file for MODE.
> >> (2) Your observation file for MODE.
> >> (3) Your MODE configuration file.
> >> (4) Your NCL output image and PostScript output
> file
> >> showing the difference.
> >>
> >> You can post that data to our anonymous ftp site
> following
> >> the directions listed here:
> >> http://www.dtcenter.org/met/users/support/met_help.php#ftp
> >>
> >> If you do send us some data, please write me back
> to let me
> >> know when it's there.
> >>
> >> John
> >>
> >> On 12/17/2010 04:26 AM, RAL HelpDesk {for soma
> senroy}
> >> wrote:
> >>> i have one add
> >>
> >>
> >
> >
>
>
>
------------------------------------------------
More information about the Met_help
mailing list