[Met_help] [rt.rap.ucar.edu #45249] History for Point-Stat questions
RAL HelpDesk {for John Halley Gotway}
met_help at ucar.edu
Fri Mar 11 12:17:40 MST 2011
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
met_help,
I made some progress but got stuck in next step..
read_levels_pinterp() -> no range specified for parsing TT(0,5,*,*) from the p_interp NetCD
I found you earlier response on this which is
In reading the data for the "TT" variable, we found a bug that
Point-Stat was expecting to find a range of pressure levels to be
verified rather than just a single one - like you were requesting
in your config file. We've modified a single source file to enable
the verification of a single level. Please replace the file
"METv3.0/lib/vx_met_util/read_gridded_data.cc" with the attached
version of that file.
How to get read_gridded_data.cc to fix this.
Thanks,
Anil
--------------------------------------------------------
met_help,
Can you help me with to get initial success in MET..I am trying lot and but get stuck..
1. First I am dealing with ARW, so first converted output to pressure level output (p_interp)
2. I have prebuffr data (prepbufr.gdas.2008062312.wo40) download from NCAR MSS and I used pb2nc tool and created 2008062312.nc file. Initially I am just putting Temperature field so in pb2nc config file
I set obs_grib_code[] = [ "TMP" ] and message_type[] = ["ADPUPA"] and successfully ran pb2nc tool.
3. Pointsat config file, I set fcst_field[] = [ "TT(0,5,*,*)" ]--(at 950 mb level) and obs_field[] = [ "11" ]; as my output is in netcdf format and I can't left blank obs_field so I put 11(code for temperature), when I am running I get this message
----------------------
../bin/point_stat wrfout_d01_2008-06-23_12:00:00_PLEV 2008062312.nc PointStatConfig
ERROR: GCInfo::set_gcinfo_grib() -> each GRIB code specified must be followed by an accumulation, level, or presssure level indicator "11".
-----------------------------
Can you please help me out if I am missing anything..
Thanks,
Anil
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
From: John Halley Gotway
Time: Thu Mar 10 14:43:56 2011
Anil,
I've created a met_help ticket for your requests. Please send future
questions about MET directly to met_help at ucar.edu. That way we're
able to track the community support we provide.
Regarding your second question, please ensure that you've compiled MET
with the latest set of bugfixes which can be found here:
http://www.dtcenter.org/met/users/support/known_issues/METv3.0/index.php
Just follow the instructions listed at the top of the page to apply
the patches and recompile MET. These patches include the bugfix to
which you're referring.
Regarding your first question, I believe your config file settings
should look like this:
fcst_field[] = [ "TT(0,5,*,*)" ];
obs_field[] = [ "11/P950" ];
That tells MET to compare the TT forecast data that you specified to
the observations that occur at the 950mb level.
Hope that helps,
John Halley Gotway
met_help at ucar.edu
On 03/10/2011 02:40 PM, RAL HelpDesk {for John Halley Gotway} wrote:
>
> Thu Mar 10 14:40:32 2011: Request 45249 was acted upon.
> Transaction: Ticket created by johnhg
> Queue: met_help
> Subject: Point-Stat questions
> Owner: johnhg
> Requestors: anil.kumar at nasa.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
>
> met_help,
>
> I made some progress but got stuck in next step..
> read_levels_pinterp() -> no range specified for parsing TT(0,5,*,*)
from the p_interp NetCD
>
> I found you earlier response on this which is
> In reading the data for the "TT" variable, we found a bug that
> Point-Stat was expecting to find a range of pressure levels to be
> verified rather than just a single one - like you were requesting
> in your config file. We've modified a single source file to enable
> the verification of a single level. Please replace the file
> "METv3.0/lib/vx_met_util/read_gridded_data.cc" with the attached
> version of that file.
>
>
> How to get read_gridded_data.cc to fix this.
>
> Thanks,
> Anil
>
> --------------------------------------------------------
>
> met_help,
>
> Can you help me with to get initial success in MET..I am trying lot
and but get stuck..
>
> 1. First I am dealing with ARW, so first converted output to
pressure level output (p_interp)
>
> 2. I have prebuffr data (prepbufr.gdas.2008062312.wo40) download
from NCAR MSS and I used pb2nc tool and created 2008062312.nc file.
Initially I am just putting Temperature field so in pb2nc config file
> I set obs_grib_code[] = [ "TMP" ] and message_type[] = ["ADPUPA"]
and successfully ran pb2nc tool.
>
> 3. Pointsat config file, I set fcst_field[] = [ "TT(0,5,*,*)" ]--(at
950 mb level) and obs_field[] = [ "11" ]; as my output is in netcdf
format and I can't left blank obs_field so I put 11(code for
temperature), when I am running I get this message
> ----------------------
> ../bin/point_stat wrfout_d01_2008-06-23_12:00:00_PLEV 2008062312.nc
PointStatConfig
> ERROR: GCInfo::set_gcinfo_grib() -> each GRIB code specified must be
followed by an accumulation, level, or presssure level indicator "11".
> -----------------------------
>
>
> Can you please help me out if I am missing anything..
>
> Thanks,
> Anil
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #45249] Point-Stat questions
From: Kumar, Anil[UNIV MARYLAND COLLEGE PARK]
Time: Thu Mar 10 14:54:02 2011
Thanks John!!
Just one basic question: How to evaluate T2 and Q2(bias
and rmse), for this I have to use only ADPSFC correct? And what I
should put for T2/Q2 ( fcst_field[] = [ "T2", "Q2" ] ) in obs_field[]
= [ "??" ]-corresponding to 2-m field
Thanks,
Anil
-----Original Message-----
From: RAL HelpDesk {for John Halley
Gotway} [mailto:met_help at ucar.edu]
Sent: Thursday, March 10, 2011
4:44 PM
To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
Anil,
I've created a met_help ticket for your requests. Please send future
questions about MET directly to met_help at ucar.edu. That way we're
able to track the community support we provide.
Regarding your
second question, please ensure that you've compiled MET with the
latest set of bugfixes which can be found here:
http://www.dtcenter.org/met/users/support/known_issues/METv3.0/index.php
Just follow the instructions listed at the top of the page to apply
the patches and recompile MET. These patches include the bugfix to
which you're referring.
Regarding your first question, I believe
your config file settings should look like this:
fcst_field[] = [
"TT(0,5,*,*)" ];
obs_field[] = [ "11/P950" ];
That tells MET
to compare the TT forecast data that you specified to the observations
that occur at the 950mb level.
Hope that helps,
John Halley Gotway
met_help at ucar.edu
On 03/10/2011 02:40 PM, RAL HelpDesk {for John
Halley Gotway} wrote:
>
> Thu Mar 10 14:40:32 2011: Request 45249
was acted upon.
> Transaction: Ticket created by johnhg
>
Queue: met_help
> Subject: Point-Stat questions
> Owner:
johnhg
> Requestors: anil.kumar at nasa.gov
> Status: new
>
Ticket <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
>
> met_help,
>
> I made some progress but got stuck in next
step..
> read_levels_pinterp() -> no range specified for parsing
TT(0,5,*,*) from the p_interp NetCD
>
> I found you earlier
response on this which is
> In reading the data for the "TT"
variable, we found a bug that
> Point-Stat was expecting to find a
range of pressure levels to be
> verified rather than just a single
one - like you were requesting
> in your config file. We've modified
a single source file to enable
> the verification of a single level.
Please replace the file
>
"METv3.0/lib/vx_met_util/read_gridded_data.cc" with the attached
>
version of that file.
>
>
> How to get read_gridded_data.cc to
fix this.
>
> Thanks,
> Anil
>
>
--------------------------------------------------------
>
>
met_help,
>
> Can you help me with to get initial success in MET..I
am trying lot and but get stuck..
>
> 1. First I am dealing with
ARW, so first converted output to pressure level output (p_interp)
>
> 2. I have prebuffr data (prepbufr.gdas.2008062312.wo40) download
from NCAR MSS and I used pb2nc tool and created 2008062312.nc file.
Initially I am just putting Temperature field so in pb2nc config file
> I set obs_grib_code[] = [ "TMP" ] and message_type[] = ["ADPUPA"]
and successfully ran pb2nc tool.
>
> 3. Pointsat config file, I set
fcst_field[] = [ "TT(0,5,*,*)" ]--(at 950 mb level) and obs_field[] =
[ "11" ]; as my output is in netcdf format and I can't left blank
obs_field so I put 11(code for temperature), when I am running I get
this message
> ----------------------
> ../bin/point_stat
wrfout_d01_2008-06-23_12:00:00_PLEV 2008062312.nc PointStatConfig
>
ERROR: GCInfo::set_gcinfo_grib() -> each GRIB code specified must be
followed by an accumulation, level, or presssure level indicator "11".
> -----------------------------
>
>
> Can you please help me out
if I am missing anything..
>
> Thanks,
> Anil
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
From: John Halley Gotway
Time: Fri Mar 11 09:28:15 2011
Anil,
Sorry for the delay in getting back with you. I saw your message
yesterday but was headed out the door.
To verify 2-meter temperature and 2-meter specific humidity, I think
the following would do the trick:
fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
obs_field[] = [ "TMP/Z2", "SPFH/Z2" ];
obtype[] = [ "ADPUPA" ];
This tells Point-Stat to compare T2 temp and Q2 specific humidity in
the pinterp output to observations of temperature (TMP/Z2) and
specific humidity (SPFH/Z2), respectively. Note that the
observations are still specified using the the GRIB conventions.
Also, I want to warn you that it's up to you to interpret these
results. For each field you verify, you should ensure that the units
used in the WRF output match the units prescribed by the GRIB
conventions (http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While
METv3.0 can read the NetCDF output of pinterp, we don't run it
regularly in our work. Instead, we use the WRF-PostProcessor which
outputs GRIB directly. So while METv3.0 reads the pinterp output,
please be sure to analyze the output closely.
Hope that helps.
Thanks,
John
On 03/10/2011 02:54 PM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
> Just one basic question: How to evaluate T2 and Q2(bias and rmse),
for this I have to use only ADPSFC correct? And what I should put for
T2/Q2 ( fcst_field[] = [ "T2", "Q2" ] ) in obs_field[] = [ "??" ]-
corresponding to 2-m field
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #45249] Point-Stat questions
From: Kumar, Anil[UNIV MARYLAND COLLEGE PARK]
Time: Fri Mar 11 10:03:57 2011
Okay, I did these setting and it seems this work fine and created some
files
But when I am checking in files there is no numbers and just a
first row.
Point-state runs fine with this statement (Is I am still
missing something)
discover08 153% ../../../nu-wrf_v2dev3-
3.2.1/MET/bin/point_stat wrfout_d01_2008-06-22_00:00:00_PLEV
2008062312.nc PointStatConfig -outdir point_stat -v 2
GSL_RNG_TYPE=mt19937
GSL_RNG_SEED=18446744072020523597
Forecast
File: wrfout_d01_2008-06-22_00:00:00_PLEV
Climatology File: none
Configuration File: PointStatConfig
Observation File: 2008062312.nc
--------------------------------------------------------------------------------
Reading records for T2(0,*,*).
For T2(0,*,*) found 1 forecast levels
and 0 climatology levels.
--------------------------------------------------------------------------------
Reading records for Q2(0,*,*).
For Q2(0,*,*) found 1 forecast levels
and 0 climatology levels.
--------------------------------------------------------------------------------
Searching 577 observations from 7 PrepBufr messages.
--------------------------------------------------------------------------------
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA, over
region FULL, for interpolation method MEDIAN(4), using 0 pairs.
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA, over
region SPL, for interpolation method MEDIAN(4), using 0 pairs.
--------------------------------------------------------------------------------
Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA, over
region FULL, for interpolation method MEDIAN(4), using 0 pairs.
Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA, over
region SPL, for interpolation method MEDIAN(4), using 0 pairs.
--------------------------------------------------------------------------------
Output file: point_stat/point_stat_14640000L_20080622_000000V.stat
Output file: point_stat/point_stat_14640000L_20080622_000000V_fho.txt
Output file: point_stat/point_stat_14640000L_20080622_000000V_ctc.txt
Output file: point_stat/point_stat_14640000L_20080622_000000V_cts.txt
Output file: point_stat/point_stat_14640000L_20080622_000000V_mctc.txt
-----Original Message-----
From: RAL HelpDesk {for John Halley
Gotway} [mailto:met_help at ucar.edu]
Sent: Friday, March 11, 2011
11:28 AM
To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
Anil,
Sorry for the delay in getting back with you. I saw your
message yesterday but was headed out the door.
To verify 2-meter
temperature and 2-meter specific humidity, I think the following would
do the trick:
fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
obs_field[] = [ "TMP/Z2", "SPFH/Z2" ];
obtype[] = [ "ADPUPA" ];
This tells Point-Stat to compare T2 temp and Q2 specific humidity in
the pinterp output to observations of temperature (TMP/Z2) and
specific humidity (SPFH/Z2), respectively. Note that the
observations are still specified using the the GRIB conventions.
Also, I want to warn you that it's up to you to interpret these
results. For each field you verify, you should ensure that the units
used in the WRF output match the units prescribed by the GRIB
conventions (http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While
METv3.0 can read the NetCDF output of pinterp, we don't run it
regularly in our work. Instead, we use the WRF-PostProcessor which
outputs GRIB directly. So while METv3.0 reads the pinterp output,
please be sure to analyze the output closely.
Hope that helps.
Thanks,
John
On 03/10/2011 02:54 PM, RAL HelpDesk {for Kumar,
Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
> Just one basic question:
How to evaluate T2 and Q2(bias and rmse), for this I have to use only
ADPSFC correct? And what I should put for T2/Q2 ( fcst_field[] = [
"T2", "Q2" ] ) in obs_field[] = [ "??" ]-corresponding to 2-m field
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
From: John Halley Gotway
Time: Fri Mar 11 11:23:51 2011
Anil,
If you look at what Point-Stat printed to the screen, you'll see that
it's using 0 matched pairs:
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
That's why you're getting no output statistics - Point-Stat isn't
finding any matched pairs. And I believe the reason for this is the
"obtype" you're using. Since you're trying to verify 2-meter
fields, you should be verifying against observations at the surface.
For MET, the surface observations are of type ADPSFC and/or SFCSHP for
land surface and water surface observations, respectively.
In your Point-Stat config file, please try changing ADPUPA to ADPSFC
and rerun.
Thanks,
John
On 03/11/2011 10:03 AM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
> Okay, I did these setting and it seems this work fine and created
some files
> But when I am checking in files there is no numbers and just a first
row.
>
> Point-state runs fine with this statement (Is I am still missing
something)
>
> discover08 153% ../../../nu-wrf_v2dev3-3.2.1/MET/bin/point_stat
wrfout_d01_2008-06-22_00:00:00_PLEV 2008062312.nc PointStatConfig
-outdir point_stat -v 2
> GSL_RNG_TYPE=mt19937
> GSL_RNG_SEED=18446744072020523597
> Forecast File: wrfout_d01_2008-06-22_00:00:00_PLEV
> Climatology File: none
> Configuration File: PointStatConfig
> Observation File: 2008062312.nc
>
>
--------------------------------------------------------------------------------
>
> Reading records for T2(0,*,*).
> For T2(0,*,*) found 1 forecast levels and 0 climatology levels.
>
>
--------------------------------------------------------------------------------
>
> Reading records for Q2(0,*,*).
> For Q2(0,*,*) found 1 forecast levels and 0 climatology levels.
>
>
--------------------------------------------------------------------------------
>
> Searching 577 observations from 7 PrepBufr messages.
>
>
--------------------------------------------------------------------------------
>
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>
>
--------------------------------------------------------------------------------
>
> Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
> Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>
>
--------------------------------------------------------------------------------
>
> Output file: point_stat/point_stat_14640000L_20080622_000000V.stat
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_fho.txt
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_ctc.txt
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_cts.txt
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_mctc.txt
>
>
>
>
>
>
> -----Original Message-----
> From: RAL HelpDesk {for John Halley Gotway}
[mailto:met_help at ucar.edu]
> Sent: Friday, March 11, 2011 11:28 AM
> To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
> Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
>
>
>
> Anil,
>
> Sorry for the delay in getting back with you. I saw your message
yesterday but was headed out the door.
>
> To verify 2-meter temperature and 2-meter specific humidity, I think
the following would do the trick:
>
> fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
> obs_field[] = [ "TMP/Z2", "SPFH/Z2" ];
> obtype[] = [ "ADPUPA" ];
>
> This tells Point-Stat to compare T2 temp and Q2 specific humidity in
the pinterp output to observations of temperature (TMP/Z2) and
specific humidity (SPFH/Z2), respectively. Note that the
> observations are still specified using the the GRIB conventions.
>
> Also, I want to warn you that it's up to you to interpret these
results. For each field you verify, you should ensure that the units
used in the WRF output match the units prescribed by the GRIB
> conventions (http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While
METv3.0 can read the NetCDF output of pinterp, we don't run it
regularly in our work. Instead, we use the WRF-PostProcessor which
> outputs GRIB directly. So while METv3.0 reads the pinterp output,
please be sure to analyze the output closely.
>
> Hope that helps.
>
> Thanks,
> John
>
> On 03/10/2011 02:54 PM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
>> Just one basic question: How to evaluate T2 and Q2(bias and rmse),
for this I have to use only ADPSFC correct? And what I should put for
T2/Q2 ( fcst_field[] = [ "T2", "Q2" ] ) in obs_field[] = [ "??" ]-
corresponding to 2-m field
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #45249] Point-Stat questions
From: Kumar, Anil[UNIV MARYLAND COLLEGE PARK]
Time: Fri Mar 11 11:43:30 2011
Finally it works!!!!! Thanks a lot!!!!
Just one more favor, to get
bias and rmse which flag should be made output (Is it 3?)
Thanks
again!
Anil
-----Original Message-----
From: RAL HelpDesk {for
John Halley Gotway} [mailto:met_help at ucar.edu]
Sent: Friday, March
11, 2011 1:24 PM
To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE
PARK]
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
Anil,
If you look at what Point-Stat printed to the screen, you'll
see that it's using 0 matched pairs:
Processing T2(0,*,*) versus
TMP/Z2, for observation type ADPUPA, over region FULL, for
interpolation method MEDIAN(4), using 0 pairs.
Processing
T2(0,*,*) versus TMP/Z2, for observation type ADPUPA, over region SPL,
for interpolation method MEDIAN(4), using 0 pairs.
That's why
you're getting no output statistics - Point-Stat isn't finding any
matched pairs. And I believe the reason for this is the "obtype"
you're using. Since you're trying to verify 2-meter
fields, you
should be verifying against observations at the surface. For MET, the
surface observations are of type ADPSFC and/or SFCSHP for land surface
and water surface observations, respectively.
In your Point-Stat
config file, please try changing ADPUPA to ADPSFC and rerun.
Thanks,
John
On 03/11/2011 10:03 AM, RAL HelpDesk {for Kumar,
Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
>
> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
> Okay,
I did these setting and it seems this work fine and created some files
> But when I am checking in files there is no numbers and just a first
row.
>
> Point-state runs fine with this statement (Is I am still
missing something)
>
> discover08 153% ../../../nu-wrf_v2dev3-
3.2.1/MET/bin/point_stat wrfout_d01_2008-06-22_00:00:00_PLEV
2008062312.nc PointStatConfig -outdir point_stat -v 2
>
GSL_RNG_TYPE=mt19937
> GSL_RNG_SEED=18446744072020523597
> Forecast
File: wrfout_d01_2008-06-22_00:00:00_PLEV
> Climatology File: none
>
Configuration File: PointStatConfig
> Observation File: 2008062312.nc
>
>
--------------------------------------------------------------------------------
>
> Reading records for T2(0,*,*).
> For T2(0,*,*) found 1 forecast
levels and 0 climatology levels.
>
>
--------------------------------------------------------------------------------
>
> Reading records for Q2(0,*,*).
> For Q2(0,*,*) found 1 forecast
levels and 0 climatology levels.
>
>
--------------------------------------------------------------------------------
>
> Searching 577 observations from 7 PrepBufr messages.
>
>
--------------------------------------------------------------------------------
>
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>
>
--------------------------------------------------------------------------------
>
> Processing Q2(0,*,*) versus SPFH/Z2, for observation type
ADPUPA, over region FULL, for interpolation method MEDIAN(4), using 0
pairs.
> Processing Q2(0,*,*) versus SPFH/Z2, for observation type
ADPUPA, over region SPL, for interpolation method MEDIAN(4), using 0
pairs.
>
>
--------------------------------------------------------------------------------
>
> Output file:
point_stat/point_stat_14640000L_20080622_000000V.stat
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_fho.txt
> Output
file: point_stat/point_stat_14640000L_20080622_000000V_ctc.txt
>
Output file: point_stat/point_stat_14640000L_20080622_000000V_cts.txt
> Output file:
point_stat/point_stat_14640000L_20080622_000000V_mctc.txt
>
>
>
>
>
>
> -----Original Message-----
> From: RAL HelpDesk {for
John Halley Gotway} [mailto:met_help at ucar.edu]
> Sent: Friday, March
11, 2011 11:28 AM
> To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND
COLLEGE PARK]
> Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat
questions
>
>
>
> Anil,
>
> Sorry for the delay in getting
back with you. I saw your message yesterday but was headed out the
door.
>
> To verify 2-meter temperature and 2-meter specific
humidity, I think the following would do the trick:
>
>
fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
> obs_field[] = [
"TMP/Z2", "SPFH/Z2" ];
> obtype[] = [ "ADPUPA" ];
>
> This
tells Point-Stat to compare T2 temp and Q2 specific humidity in the
pinterp output to observations of temperature (TMP/Z2) and specific
humidity (SPFH/Z2), respectively. Note that the
> observations are
still specified using the the GRIB conventions.
>
> Also, I want to
warn you that it's up to you to interpret these results. For each
field you verify, you should ensure that the units used in the WRF
output match the units prescribed by the GRIB
> conventions
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While METv3.0 can
read the NetCDF output of pinterp, we don't run it regularly in our
work. Instead, we use the WRF-PostProcessor which
> outputs GRIB
directly. So while METv3.0 reads the pinterp output, please be sure
to analyze the output closely.
>
> Hope that helps.
>
> Thanks,
> John
>
> On 03/10/2011 02:54 PM, RAL HelpDesk {for Kumar,
Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
>> Just one basic question:
How to evaluate T2 and Q2(bias and rmse), for this I have to use only
ADPSFC correct? And what I should put for T2/Q2 ( fcst_field[] = [
"T2", "Q2" ] ) in obs_field[] = [ "??" ]-corresponding to 2-m field
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
From: John Halley Gotway
Time: Fri Mar 11 12:01:48 2011
Anil,
The "output_flag" in the Point-Stat config file controls what line
types are output. And the comments in that config file list out the
statistics that are included in each line type. RMSE is
included in the continuous statistics line type (CNT) - which is
output flag value number 6 . As for bias, there are a couple types of
statistical bias. The multiplicative bias (mean forecast - mean
obs) is also included in the CNT line type, and I assume that's the
one you're after. But there also a "frequency bias" (FBIAS) defined
in the contingency table statistics line type (CTS).
Please refer to the comments of the config file:
METv3.0/data/config/PointStatConfig_default
And to Appendix C of the MET User's Guide for definitions for each of
the statistics MET produces:
METv3.0/doc/MET_Users_Guide.pdf
Thanks,
John
On 03/11/2011 11:43 AM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
> Finally it works!!!!! Thanks a lot!!!!
> Just one more favor, to get bias and rmse which flag should be made
output (Is it 3?)
> Thanks again!
> Anil
>
>
> -----Original Message-----
> From: RAL HelpDesk {for John Halley Gotway}
[mailto:met_help at ucar.edu]
> Sent: Friday, March 11, 2011 1:24 PM
> To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
> Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
>
> Anil,
>
> If you look at what Point-Stat printed to the screen, you'll see
that it's using 0 matched pairs:
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>
> That's why you're getting no output statistics - Point-Stat isn't
finding any matched pairs. And I believe the reason for this is the
"obtype" you're using. Since you're trying to verify 2-meter
> fields, you should be verifying against observations at the surface.
For MET, the surface observations are of type ADPSFC and/or SFCSHP for
land surface and water surface observations, respectively.
>
> In your Point-Stat config file, please try changing ADPUPA to ADPSFC
and rerun.
>
> Thanks,
> John
>
> On 03/11/2011 10:03 AM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>>
>> Okay, I did these setting and it seems this work fine and created
some files
>> But when I am checking in files there is no numbers and just a
first row.
>>
>> Point-state runs fine with this statement (Is I am still missing
something)
>>
>> discover08 153% ../../../nu-wrf_v2dev3-3.2.1/MET/bin/point_stat
wrfout_d01_2008-06-22_00:00:00_PLEV 2008062312.nc PointStatConfig
-outdir point_stat -v 2
>> GSL_RNG_TYPE=mt19937
>> GSL_RNG_SEED=18446744072020523597
>> Forecast File: wrfout_d01_2008-06-22_00:00:00_PLEV
>> Climatology File: none
>> Configuration File: PointStatConfig
>> Observation File: 2008062312.nc
>>
>>
--------------------------------------------------------------------------------
>>
>> Reading records for T2(0,*,*).
>> For T2(0,*,*) found 1 forecast levels and 0 climatology levels.
>>
>>
--------------------------------------------------------------------------------
>>
>> Reading records for Q2(0,*,*).
>> For Q2(0,*,*) found 1 forecast levels and 0 climatology levels.
>>
>>
--------------------------------------------------------------------------------
>>
>> Searching 577 observations from 7 PrepBufr messages.
>>
>>
--------------------------------------------------------------------------------
>>
>> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
>> Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>>
>>
--------------------------------------------------------------------------------
>>
>> Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA,
over region FULL, for interpolation method MEDIAN(4), using 0 pairs.
>> Processing Q2(0,*,*) versus SPFH/Z2, for observation type ADPUPA,
over region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>>
>>
--------------------------------------------------------------------------------
>>
>> Output file: point_stat/point_stat_14640000L_20080622_000000V.stat
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_fho.txt
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_ctc.txt
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_cts.txt
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_mctc.txt
>>
>>
>>
>>
>>
>>
>> -----Original Message-----
>> From: RAL HelpDesk {for John Halley Gotway}
[mailto:met_help at ucar.edu]
>> Sent: Friday, March 11, 2011 11:28 AM
>> To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
>> Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat questions
>>
>>
>>
>> Anil,
>>
>> Sorry for the delay in getting back with you. I saw your message
yesterday but was headed out the door.
>>
>> To verify 2-meter temperature and 2-meter specific humidity, I
think the following would do the trick:
>>
>> fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
>> obs_field[] = [ "TMP/Z2", "SPFH/Z2" ];
>> obtype[] = [ "ADPUPA" ];
>>
>> This tells Point-Stat to compare T2 temp and Q2 specific humidity
in the pinterp output to observations of temperature (TMP/Z2) and
specific humidity (SPFH/Z2), respectively. Note that the
>> observations are still specified using the the GRIB conventions.
>>
>> Also, I want to warn you that it's up to you to interpret these
results. For each field you verify, you should ensure that the units
used in the WRF output match the units prescribed by the GRIB
>> conventions (http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While
METv3.0 can read the NetCDF output of pinterp, we don't run it
regularly in our work. Instead, we use the WRF-PostProcessor which
>> outputs GRIB directly. So while METv3.0 reads the pinterp output,
please be sure to analyze the output closely.
>>
>> Hope that helps.
>>
>> Thanks,
>> John
>>
>> On 03/10/2011 02:54 PM, RAL HelpDesk {for Kumar, Anil[UNIV MARYLAND
COLLEGE PARK]} wrote:
>>> Just one basic question: How to evaluate T2 and Q2(bias and rmse),
for this I have to use only ADPSFC correct? And what I should put for
T2/Q2 ( fcst_field[] = [ "T2", "Q2" ] ) in obs_field[] = [ "??" ]-
corresponding to 2-m field
>>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #45249] Point-Stat questions
From: Kumar, Anil[UNIV MARYLAND COLLEGE PARK]
Time: Fri Mar 11 12:12:55 2011
Thanks John!!!
Have a great Weekend!!
Anil
-----Original
Message-----
From: RAL HelpDesk {for John Halley Gotway}
[mailto:met_help at ucar.edu]
Sent: Friday, March 11, 2011 2:02 PM
To:
Kumar, Anil (GSFC-614.3)[UNIV MARYLAND COLLEGE PARK]
Subject: Re:
[rt.rap.ucar.edu #45249] Point-Stat questions
Anil,
The
"output_flag" in the Point-Stat config file controls what line types
are output. And the comments in that config file list out the
statistics that are included in each line type. RMSE is
included in
the continuous statistics line type (CNT) - which is output flag value
number 6 . As for bias, there are a couple types of statistical bias.
The multiplicative bias (mean forecast - mean
obs) is also included
in the CNT line type, and I assume that's the one you're after. But
there also a "frequency bias" (FBIAS) defined in the contingency table
statistics line type (CTS).
Please refer to the comments of the
config file:
METv3.0/data/config/PointStatConfig_default
And to
Appendix C of the MET User's Guide for definitions for each of the
statistics MET produces:
METv3.0/doc/MET_Users_Guide.pdf
Thanks,
John
On 03/11/2011 11:43 AM, RAL HelpDesk {for Kumar,
Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
>
> <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>
>
Finally it works!!!!! Thanks a lot!!!!
> Just one more favor, to get
bias and rmse which flag should be made output (Is it 3?)
> Thanks
again!
> Anil
>
>
> -----Original Message-----
> From: RAL
HelpDesk {for John Halley Gotway} [mailto:met_help at ucar.edu]
> Sent:
Friday, March 11, 2011 1:24 PM
> To: Kumar, Anil (GSFC-614.3)[UNIV
MARYLAND COLLEGE PARK]
> Subject: Re: [rt.rap.ucar.edu #45249] Point-
Stat questions
>
> Anil,
>
> If you look at what Point-Stat
printed to the screen, you'll see that it's using 0 matched pairs:
>
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA, over
region FULL, for interpolation method MEDIAN(4), using 0 pairs.
>
Processing T2(0,*,*) versus TMP/Z2, for observation type ADPUPA, over
region SPL, for interpolation method MEDIAN(4), using 0 pairs.
>
>
That's why you're getting no output statistics - Point-Stat isn't
finding any matched pairs. And I believe the reason for this is the
"obtype" you're using. Since you're trying to verify 2-meter
>
fields, you should be verifying against observations at the surface.
For MET, the surface observations are of type ADPSFC and/or SFCSHP for
land surface and water surface observations, respectively.
>
> In
your Point-Stat config file, please try changing ADPUPA to ADPSFC and
rerun.
>
> Thanks,
> John
>
> On 03/11/2011 10:03 AM, RAL
HelpDesk {for Kumar, Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
>>
>>
<URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=45249 >
>>
>> Okay, I did these setting and it seems this work fine and created
some files
>> But when I am checking in files there is no numbers and
just a first row.
>>
>> Point-state runs fine with this statement
(Is I am still missing something)
>>
>> discover08 153% ../../../nu-
wrf_v2dev3-3.2.1/MET/bin/point_stat wrfout_d01_2008-06-
22_00:00:00_PLEV 2008062312.nc PointStatConfig -outdir point_stat -v 2
>> GSL_RNG_TYPE=mt19937
>> GSL_RNG_SEED=18446744072020523597
>>
Forecast File: wrfout_d01_2008-06-22_00:00:00_PLEV
>> Climatology
File: none
>> Configuration File: PointStatConfig
>> Observation
File: 2008062312.nc
>>
>>
--------------------------------------------------------------------------------
>>
>> Reading records for T2(0,*,*).
>> For T2(0,*,*) found 1
forecast levels and 0 climatology levels.
>>
>>
--------------------------------------------------------------------------------
>>
>> Reading records for Q2(0,*,*).
>> For Q2(0,*,*) found 1
forecast levels and 0 climatology levels.
>>
>>
--------------------------------------------------------------------------------
>>
>> Searching 577 observations from 7 PrepBufr messages.
>>
>>
--------------------------------------------------------------------------------
>>
>> Processing T2(0,*,*) versus TMP/Z2, for observation type
ADPUPA, over region FULL, for interpolation method MEDIAN(4), using 0
pairs.
>> Processing T2(0,*,*) versus TMP/Z2, for observation type
ADPUPA, over region SPL, for interpolation method MEDIAN(4), using 0
pairs.
>>
>>
--------------------------------------------------------------------------------
>>
>> Processing Q2(0,*,*) versus SPFH/Z2, for observation type
ADPUPA, over region FULL, for interpolation method MEDIAN(4), using 0
pairs.
>> Processing Q2(0,*,*) versus SPFH/Z2, for observation type
ADPUPA, over region SPL, for interpolation method MEDIAN(4), using 0
pairs.
>>
>>
--------------------------------------------------------------------------------
>>
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V.stat
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_fho.txt
>> Output
file: point_stat/point_stat_14640000L_20080622_000000V_ctc.txt
>>
Output file: point_stat/point_stat_14640000L_20080622_000000V_cts.txt
>> Output file:
point_stat/point_stat_14640000L_20080622_000000V_mctc.txt
>>
>>
>>
>>
>>
>>
>> -----Original Message-----
>> From: RAL HelpDesk {for
John Halley Gotway} [mailto:met_help at ucar.edu]
>> Sent: Friday,
March 11, 2011 11:28 AM
>> To: Kumar, Anil (GSFC-614.3)[UNIV MARYLAND
COLLEGE PARK]
>> Subject: Re: [rt.rap.ucar.edu #45249] Point-Stat
questions
>>
>>
>>
>> Anil,
>>
>> Sorry for the delay in getting
back with you. I saw your message yesterday but was headed out the
door.
>>
>> To verify 2-meter temperature and 2-meter specific
humidity, I think the following would do the trick:
>>
>>
fcst_field[] = [ "T2(0,*,*)", "Q2(0,*,*)" ];
>> obs_field[] = [
"TMP/Z2", "SPFH/Z2" ];
>> obtype[] = [ "ADPUPA" ];
>>
>> This
tells Point-Stat to compare T2 temp and Q2 specific humidity in the
pinterp output to observations of temperature (TMP/Z2) and specific
humidity (SPFH/Z2), respectively. Note that the
>> observations are
still specified using the the GRIB conventions.
>>
>> Also, I want
to warn you that it's up to you to interpret these results. For each
field you verify, you should ensure that the units used in the WRF
output match the units prescribed by the GRIB
>> conventions
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/). While METv3.0 can
read the NetCDF output of pinterp, we don't run it regularly in our
work. Instead, we use the WRF-PostProcessor which
>> outputs GRIB
directly. So while METv3.0 reads the pinterp output, please be sure
to analyze the output closely.
>>
>> Hope that helps.
>>
>>
Thanks,
>> John
>>
>> On 03/10/2011 02:54 PM, RAL HelpDesk {for
Kumar, Anil[UNIV MARYLAND COLLEGE PARK]} wrote:
>>> Just one basic
question: How to evaluate T2 and Q2(bias and rmse), for this I have to
use only ADPSFC correct? And what I should put for T2/Q2 (
fcst_field[] = [ "T2", "Q2" ] ) in obs_field[] = [ "??" ]-
corresponding to 2-m field
>>
>
------------------------------------------------
More information about the Met_help
mailing list