[Met_help] [rt.rap.ucar.edu #87597] History for Mismatch of fcst and obs images
John Halley Gotway via RT
met_help at ucar.edu
Tue Jul 9 12:07:19 MDT 2019
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Sir,
I am able to run the mode of met tools. I am in a dilemma while analysing
the fcst and obs files by running mode.
The confusion is in the matching and merging of fcst and obs of
precipitation images.
I could see the observation and fcst of precipitation separately in the
region of my study from the 3B42RT.2018082003.7.nc4 file as well as
APCP_fcst file but the regions are not matching while running the mode.
I am sending herewith the following files.
1. ncview output (in .ps format)of observed netcdf file for precipitation
[ncview 3B42RT.2018082003.7.nc4].
2. .ps file created after running mode [mode APCP_3h_03utc20Aug.nc
PYTHON_NUMPY MODEConfig_3B42RT_py2018 -outdir ./output/mode -v 2].
3. MODE_Config_file [MODEConfig_3B42RT_py2018]
4. TRMM observation file [3B42RT.2018082003.7.nc4]
5. ncview output (in .ps format)of fcst netcdf file for precipitation
[ncview APCP_3h_03utc20Aug.nc].
Kindly look into these files.
Thanking You
With Best Regards
________________________
Dr. BIRANCHI KUMAR MAHALA
Asst. Prof., Mathematics
*Kalinga Institute of Industrial Technology,*
*Deemed to be University, Bhubaneswar*
*Mobile- 09861182688*
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Mismatch of fcst and obs images
From: John Halley Gotway
Time: Wed Oct 31 15:47:22 2018
Dr Biranchi,
Judging from the PostScript files you sent, I see that the forecast
objects
look reasonable. However, the raw observation field does not. It's
just a
constant field containing all 0's.
However, I'm not able to replicate the behavior you describe. I reran
MODE
using the data you sent:
/usr/local/met-8.0/bin/mode \
APCP_3h_03utc20Aug.nc PYTHON_NUMPY \
MODEConfig_3B42RT_py2018_JHG -outdir out
My config file is the same as yours, except that I updated the paths
to the
python script to run.
That produced the attached PostScript image, which looks good.
So the question is this, why is it working well for me and not you?
Another thing to try out would be:
- In the "regrid" section, try setting "to_grid = OBS".
- In the "regrid" section, try using "method = NEAREST; width = 1;" or
"method = BILIN; width = 2;"
Also, prior to running MODE, make sure that the python plotting is
still
working well with MET by running:
/usr/local/met-8.0/bin/plot_data_plane PYTHON_NUMPY
3B42RT.2018082003.7.ps
'name="./read_3B42RT.py ./3B42RT.2018082003.7.nc4 precipitation";'
Do those produce any more reasonable results?
Thanks,
John
On Wed, Oct 31, 2018 at 10:39 AM Biranchi Mahala via RT
<met_help at ucar.edu>
wrote:
>
> Wed Oct 31 10:39:13 2018: Request 87597 was acted upon.
> Transaction: Ticket created by biranchimahala at gmail.com
> Queue: met_help
> Subject: Mismatch of fcst and obs images
> Owner: Nobody
> Requestors: biranchimahala at gmail.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=87597 >
>
>
> Sir,
> I am able to run the mode of met tools. I am in a dilemma while
analysing
> the fcst and obs files by running mode.
> The confusion is in the matching and merging of fcst and obs of
> precipitation images.
> I could see the observation and fcst of precipitation separately in
the
> region of my study from the 3B42RT.2018082003.7.nc4 file as well as
> APCP_fcst file but the regions are not matching while running the
mode.
> I am sending herewith the following files.
> 1. ncview output (in .ps format)of observed netcdf file for
precipitation
> [ncview 3B42RT.2018082003.7.nc4].
> 2. .ps file created after running mode [mode APCP_3h_03utc20Aug.nc
> PYTHON_NUMPY MODEConfig_3B42RT_py2018 -outdir ./output/mode -v 2].
> 3. MODE_Config_file [MODEConfig_3B42RT_py2018]
> 4. TRMM observation file [3B42RT.2018082003.7.nc4]
> 5. ncview output (in .ps format)of fcst netcdf file for
precipitation
> [ncview APCP_3h_03utc20Aug.nc].
> Kindly look into these files.
>
> Thanking You
> With Best Regards
> ________________________
> Dr. BIRANCHI KUMAR MAHALA
> Asst. Prof., Mathematics
> *Kalinga Institute of Industrial Technology,*
> *Deemed to be University, Bhubaneswar*
> *Mobile- 09861182688*
>
>
------------------------------------------------
Subject: Mismatch of fcst and obs images
From: Biranchi Mahala
Time: Thu Nov 01 11:29:54 2018
Sir,
I am not getting any improved result by changing;
In the "regrid" section, "to_grid = OBS".
In the "regrid" section, "method = NEAREST; width = 1;" or
"method = BILIN; width = 2;" separately.
Also python plotting is working well as;
python read_3B42RT.py 3B42RT.2018082003.7.nc4 precipitation
Input File: '3B42RT.2018082003.7.nc4'
Variable Name : 'precipitation'
Data Shape: (480, 1440)
Data Type: dtype('float64')
Begin Time = 2018-08-20 01:30:00
End Time = 2018-08-20 04:30:00
Accum Time = 030000
Attributes: {'long_name': 'precipitation', 'init': '20180820_043000',
'valid': '20180820_043000', 'grid': {'delta_lat': 0.25, 'lon_ll':
-179.875,
'name': '3B42-Grid', 'Nlat': 480, 'lat_ll': -59.875, 'delta_lon':
0.25,
'type': 'LatLon', 'Nlon': 1440}, 'name': 'precipitation', 'lead':
'00',
'level': 'Surface', 'units': 'mm', 'accum': '030000'}
/MET_V8/bin/mode APCP_3h_03utc20Aug.nc PYTHON_NUMPY
MODEConfig_3B42RT_py2018 -outdir output/mode
produce the following;
DEBUG 1: Default Config File:
/MET_V8/share/met//config/MODEConfig_default
DEBUG 1: Match Config File: MODEConfig_3B42RT_py2018
DEBUG 1: Merge Config File: MODEConfig_3B42RT_py2018
DEBUG 1: Forecast File: APCP_3h_03utc20Aug.nc
DEBUG 1: Observation File: PYTHON_NUMPY
Input File: '/home/MET_working_dir/3B42RT.2018082003.7.nc4'
Variable Name : 'precipitation'
Data Shape: (480, 1440)
Data Type: dtype('float64')
Begin Time = 2018-08-20 01:30:00
End Time = 2018-08-20 04:30:00
Accum Time = 030000
Attributes: {'long_name': 'precipitation', 'init': '20180820_043000',
'valid': '20180820_043000', 'grid': {'delta_lat': 0.25, 'lon_ll':
-179.875,
'name': '3B42-Grid', 'Nlat': 480, 'lat_ll': -59.875, 'delta_lon':
0.25,
'type': 'LatLon', 'Nlon': 1440}, 'name': 'precipitation', 'lead':
'00',
'level': 'Surface', 'units': 'mm', 'accum': '030000'}
DEBUG 1: Regridding observation precipitation/Surface to the
verification
grid.
WARNING:
WARNING: setup_fcst_obs_data() -> Forecast and observation valid times
do
not match 20180820_030000 != 20180820_043000 for APCP_03(*,*) versus
precipitation/Surface.
WARNING:
DEBUG 1: Forecast Field: APCP_03 at A3
DEBUG 1: Observation Field: precipitation at Surface
DEBUG 2: Processing masking regions.
DEBUG 2: Identifying objects in the forecast and observation fields...
DEBUG 2: Computing contingency table statistics...
DEBUG 2: Identified: 8 forecast objects and 0 observation objects.
DEBUG 2: Performing merging (THRESH) in the forecast field.
DEBUG 2: Performing merging (THRESH) in the observation field.
DEBUG 2: Remaining: 8 forecast objects and 0 observation objects.
DEBUG 2: Performing matching (MERGE_BOTH) between the forecast and
observation fields.
DEBUG 1: Creating Fcst-Obs Object Statistics file:
output/mode/mode_270000L_20180820_030000V_030000A_obj.txt
DEBUG 1: Creating Contingency Table Statistics file:
output/mode/mode_270000L_20180820_030000V_030000A_cts.txt
DEBUG 1: Creating Object NetCDF file:
output/mode/mode_270000L_20180820_030000V_030000A_obj.nc
DEBUG 1: Loading forecast raw color table:
/MET_V8/share/met//colortables/met_default.ctable
DEBUG 1: Loading observation raw color table:
/MET_V8/share/met//colortables/met_default.ctable
DEBUG 1: Creating postscript file:
output/mode/mode_270000L_20180820_030000V_030000A.ps
For information.
Thanking You
With Best Regards
________________________
Dr. BIRANCHI KUMAR MAHALA
Asst. Prof., Mathematics
*Kalinga Institute of Industrial Technology,*
*Deemed to be University, Bhubaneswar*
*Mobile- 09861182688*
On Thu, Nov 1, 2018 at 3:17 AM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:
> Dr Biranchi,
>
> Judging from the PostScript files you sent, I see that the forecast
objects
> look reasonable. However, the raw observation field does not. It's
just a
> constant field containing all 0's.
>
> However, I'm not able to replicate the behavior you describe. I
reran MODE
> using the data you sent:
> /usr/local/met-8.0/bin/mode \
> APCP_3h_03utc20Aug.nc PYTHON_NUMPY \
> MODEConfig_3B42RT_py2018_JHG -outdir out
>
> My config file is the same as yours, except that I updated the paths
to the
> python script to run.
>
> That produced the attached PostScript image, which looks good.
>
> So the question is this, why is it working well for me and not you?
>
> Another thing to try out would be:
> - In the "regrid" section, try setting "to_grid = OBS".
> - In the "regrid" section, try using "method = NEAREST; width = 1;"
or
> "method = BILIN; width = 2;"
>
> Also, prior to running MODE, make sure that the python plotting is
still
> working well with MET by running:
>
> /usr/local/met-8.0/bin/plot_data_plane PYTHON_NUMPY
3B42RT.2018082003.7.ps
> 'name="./read_3B42RT.py ./3B42RT.2018082003.7.nc4 precipitation";'
>
> Do those produce any more reasonable results?
>
> Thanks,
> John
>
> On Wed, Oct 31, 2018 at 10:39 AM Biranchi Mahala via RT
<met_help at ucar.edu
> >
> wrote:
>
> >
> > Wed Oct 31 10:39:13 2018: Request 87597 was acted upon.
> > Transaction: Ticket created by biranchimahala at gmail.com
> > Queue: met_help
> > Subject: Mismatch of fcst and obs images
> > Owner: Nobody
> > Requestors: biranchimahala at gmail.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=87597 >
> >
> >
> > Sir,
> > I am able to run the mode of met tools. I am in a dilemma while
analysing
> > the fcst and obs files by running mode.
> > The confusion is in the matching and merging of fcst and obs of
> > precipitation images.
> > I could see the observation and fcst of precipitation separately
in the
> > region of my study from the 3B42RT.2018082003.7.nc4 file as well
as
> > APCP_fcst file but the regions are not matching while running the
mode.
> > I am sending herewith the following files.
> > 1. ncview output (in .ps format)of observed netcdf file for
precipitation
> > [ncview 3B42RT.2018082003.7.nc4].
> > 2. .ps file created after running mode [mode APCP_3h_03utc20Aug.nc
> > PYTHON_NUMPY MODEConfig_3B42RT_py2018 -outdir ./output/mode -v 2].
> > 3. MODE_Config_file [MODEConfig_3B42RT_py2018]
> > 4. TRMM observation file [3B42RT.2018082003.7.nc4]
> > 5. ncview output (in .ps format)of fcst netcdf file for
precipitation
> > [ncview APCP_3h_03utc20Aug.nc].
> > Kindly look into these files.
> >
> > Thanking You
> > With Best Regards
> > ________________________
> > Dr. BIRANCHI KUMAR MAHALA
> > Asst. Prof., Mathematics
> > *Kalinga Institute of Industrial Technology,*
> > *Deemed to be University, Bhubaneswar*
> > *Mobile- 09861182688*
> >
> >
>
>
------------------------------------------------
Subject: Mismatch of fcst and obs images
From: John Halley Gotway
Time: Fri Nov 02 15:56:49 2018
Dr. Biranchi,
I recall us already doing this, but let's make sure that they Python
script
can still be run through the plot_data_plane tool to make a reasonable
plot
of the data:
*/MET_V8/bin/plot_data_plane PYTHON_NUMPY 3B42RT.2018082003.7.ps
<http://3B42RT.2018082003.7.ps> 'name="read_3B42RT.py
3B42RT.2018082003.7.nc4 precipitation";'*
Does the resulting image "*3B42RT.2018082003.7.ps
<http://3B42RT.2018082003.7.ps>*" look good? I would expect that if
plot_data_plane reads the data well, then so would the MODE tool.
Thanks,
John
On Thu, Nov 1, 2018 at 11:30 AM Biranchi Mahala via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=87597 >
>
> Sir,
> I am not getting any improved result by changing;
> In the "regrid" section, "to_grid = OBS".
> In the "regrid" section, "method = NEAREST; width = 1;" or
> "method = BILIN; width = 2;" separately.
> Also python plotting is working well as;
> python read_3B42RT.py 3B42RT.2018082003.7.nc4 precipitation
> Input File: '3B42RT.2018082003.7.nc4'
> Variable Name : 'precipitation'
> Data Shape: (480, 1440)
> Data Type: dtype('float64')
> Begin Time = 2018-08-20 01:30:00
> End Time = 2018-08-20 04:30:00
> Accum Time = 030000
> Attributes: {'long_name': 'precipitation', 'init':
'20180820_043000',
> 'valid': '20180820_043000', 'grid': {'delta_lat': 0.25, 'lon_ll':
-179.875,
> 'name': '3B42-Grid', 'Nlat': 480, 'lat_ll': -59.875, 'delta_lon':
0.25,
> 'type': 'LatLon', 'Nlon': 1440}, 'name': 'precipitation', 'lead':
'00',
> 'level': 'Surface', 'units': 'mm', 'accum': '030000'}
>
> /MET_V8/bin/mode APCP_3h_03utc20Aug.nc PYTHON_NUMPY
> MODEConfig_3B42RT_py2018 -outdir output/mode
> produce the following;
>
> DEBUG 1: Default Config File:
/MET_V8/share/met//config/MODEConfig_default
> DEBUG 1: Match Config File: MODEConfig_3B42RT_py2018
> DEBUG 1: Merge Config File: MODEConfig_3B42RT_py2018
> DEBUG 1: Forecast File: APCP_3h_03utc20Aug.nc
> DEBUG 1: Observation File: PYTHON_NUMPY
> Input File: '/home/MET_working_dir/3B42RT.2018082003.7.nc4'
> Variable Name : 'precipitation'
> Data Shape: (480, 1440)
> Data Type: dtype('float64')
> Begin Time = 2018-08-20 01:30:00
> End Time = 2018-08-20 04:30:00
> Accum Time = 030000
> Attributes: {'long_name': 'precipitation', 'init':
'20180820_043000',
> 'valid': '20180820_043000', 'grid': {'delta_lat': 0.25, 'lon_ll':
-179.875,
> 'name': '3B42-Grid', 'Nlat': 480, 'lat_ll': -59.875, 'delta_lon':
0.25,
> 'type': 'LatLon', 'Nlon': 1440}, 'name': 'precipitation', 'lead':
'00',
> 'level': 'Surface', 'units': 'mm', 'accum': '030000'}
> DEBUG 1: Regridding observation precipitation/Surface to the
verification
> grid.
> WARNING:
> WARNING: setup_fcst_obs_data() -> Forecast and observation valid
times do
> not match 20180820_030000 != 20180820_043000 for APCP_03(*,*) versus
> precipitation/Surface.
> WARNING:
> DEBUG 1: Forecast Field: APCP_03 at A3
> DEBUG 1: Observation Field: precipitation at Surface
> DEBUG 2: Processing masking regions.
> DEBUG 2: Identifying objects in the forecast and observation
fields...
> DEBUG 2: Computing contingency table statistics...
> DEBUG 2: Identified: 8 forecast objects and 0 observation objects.
> DEBUG 2: Performing merging (THRESH) in the forecast field.
> DEBUG 2: Performing merging (THRESH) in the observation field.
> DEBUG 2: Remaining: 8 forecast objects and 0 observation objects.
> DEBUG 2: Performing matching (MERGE_BOTH) between the forecast and
> observation fields.
> DEBUG 1: Creating Fcst-Obs Object Statistics file:
> output/mode/mode_270000L_20180820_030000V_030000A_obj.txt
> DEBUG 1: Creating Contingency Table Statistics file:
> output/mode/mode_270000L_20180820_030000V_030000A_cts.txt
> DEBUG 1: Creating Object NetCDF file:
> output/mode/mode_270000L_20180820_030000V_030000A_obj.nc
> DEBUG 1: Loading forecast raw color table:
> /MET_V8/share/met//colortables/met_default.ctable
> DEBUG 1: Loading observation raw color table:
> /MET_V8/share/met//colortables/met_default.ctable
> DEBUG 1: Creating postscript file:
> output/mode/mode_270000L_20180820_030000V_030000A.ps
>
> For information.
> Thanking You
>
> With Best Regards
> ________________________
> Dr. BIRANCHI KUMAR MAHALA
> Asst. Prof., Mathematics
> *Kalinga Institute of Industrial Technology,*
> *Deemed to be University, Bhubaneswar*
> *Mobile- 09861182688*
>
>
>
> On Thu, Nov 1, 2018 at 3:17 AM John Halley Gotway via RT <
> met_help at ucar.edu>
> wrote:
>
> > Dr Biranchi,
> >
> > Judging from the PostScript files you sent, I see that the
forecast
> objects
> > look reasonable. However, the raw observation field does not.
It's
> just a
> > constant field containing all 0's.
> >
> > However, I'm not able to replicate the behavior you describe. I
reran
> MODE
> > using the data you sent:
> > /usr/local/met-8.0/bin/mode \
> > APCP_3h_03utc20Aug.nc PYTHON_NUMPY \
> > MODEConfig_3B42RT_py2018_JHG -outdir out
> >
> > My config file is the same as yours, except that I updated the
paths to
> the
> > python script to run.
> >
> > That produced the attached PostScript image, which looks good.
> >
> > So the question is this, why is it working well for me and not
you?
> >
> > Another thing to try out would be:
> > - In the "regrid" section, try setting "to_grid = OBS".
> > - In the "regrid" section, try using "method = NEAREST; width =
1;" or
> > "method = BILIN; width = 2;"
> >
> > Also, prior to running MODE, make sure that the python plotting is
still
> > working well with MET by running:
> >
> > /usr/local/met-8.0/bin/plot_data_plane PYTHON_NUMPY
> 3B42RT.2018082003.7.ps
> > 'name="./read_3B42RT.py ./3B42RT.2018082003.7.nc4 precipitation";'
> >
> > Do those produce any more reasonable results?
> >
> > Thanks,
> > John
> >
> > On Wed, Oct 31, 2018 at 10:39 AM Biranchi Mahala via RT <
> met_help at ucar.edu
> > >
> > wrote:
> >
> > >
> > > Wed Oct 31 10:39:13 2018: Request 87597 was acted upon.
> > > Transaction: Ticket created by biranchimahala at gmail.com
> > > Queue: met_help
> > > Subject: Mismatch of fcst and obs images
> > > Owner: Nobody
> > > Requestors: biranchimahala at gmail.com
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=87597
> >
> > >
> > >
> > > Sir,
> > > I am able to run the mode of met tools. I am in a dilemma while
> analysing
> > > the fcst and obs files by running mode.
> > > The confusion is in the matching and merging of fcst and obs of
> > > precipitation images.
> > > I could see the observation and fcst of precipitation separately
in the
> > > region of my study from the 3B42RT.2018082003.7.nc4 file as well
as
> > > APCP_fcst file but the regions are not matching while running
the mode.
> > > I am sending herewith the following files.
> > > 1. ncview output (in .ps format)of observed netcdf file for
> precipitation
> > > [ncview 3B42RT.2018082003.7.nc4].
> > > 2. .ps file created after running mode [mode
APCP_3h_03utc20Aug.nc
> > > PYTHON_NUMPY MODEConfig_3B42RT_py2018 -outdir ./output/mode -v
2].
> > > 3. MODE_Config_file [MODEConfig_3B42RT_py2018]
> > > 4. TRMM observation file [3B42RT.2018082003.7.nc4]
> > > 5. ncview output (in .ps format)of fcst netcdf file for
precipitation
> > > [ncview APCP_3h_03utc20Aug.nc].
> > > Kindly look into these files.
> > >
> > > Thanking You
> > > With Best Regards
> > > ________________________
> > > Dr. BIRANCHI KUMAR MAHALA
> > > Asst. Prof., Mathematics
> > > *Kalinga Institute of Industrial Technology,*
> > > *Deemed to be University, Bhubaneswar*
> > > *Mobile- 09861182688*
> > >
> > >
> >
> >
>
>
------------------------------------------------
More information about the Met_help
mailing list