[Met_help] [rt.rap.ucar.edu #46901] History for Question about map projection in stat_point
RAL HelpDesk {for John Halley Gotway}
met_help at ucar.edu
Tue May 17 15:52:40 MDT 2011
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
To whom it may concern,
My simulation results is from wrf, and observation is in prepbufr
format. Wind observation in prepbufr is in earth coordinate, while wind
results from wrf output always in projection map. Observation and
simulation results are not in the same coordination. Is there any wind
rotation included in met? or should I rotate the observation wind before
pb2nc? Thank you very much.
Yujie
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46901] Question about map projection in stat_point
From: John Halley Gotway
Time: Tue May 17 14:53:24 2011
Yujie,
You are correct. The U and V wind observations in PREPBUFR files are
earth-relative while the output of WRF is grid-relative. Can you tell
me, are you running your WRF output through the WRF
PostProcessor and generating GRIB output to verify?
If so, then everything is fine. When verifying U and V, MET checks a
flag in the GRIB file to see if the winds are grid-relative. If so,
it rotates them to be earth relative prior to verifying them.
If you are using the pinterp tool instead of WPP, I do not believe
that rotation from grid to earth relative is occurring. I'll make a
note of that to see how we can address it.
Thanks,
John Halley Gotway
met_help at ucar.edu
On 05/17/2011 02:12 PM, RAL HelpDesk {for Yujie Pan} wrote:
>
> Tue May 17 14:12:39 2011: Request 46901 was acted upon.
> Transaction: Ticket created by panyujie at gmail.com
> Queue: met_help
> Subject: Question about map projection in stat_point
> Owner: Nobody
> Requestors: panyujie at gmail.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901 >
>
>
> To whom it may concern,
>
> My simulation results is from wrf, and observation is in prepbufr
> format. Wind observation in prepbufr is in earth coordinate, while
wind
> results from wrf output always in projection map. Observation and
> simulation results are not in the same coordination. Is there any
wind
> rotation included in met? or should I rotate the observation wind
before
> pb2nc? Thank you very much.
>
>
> Yujie
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46901] Question about map projection in stat_point
From: Yujie Pan
Time: Tue May 17 15:07:14 2011
Thank you very much, John. You help me a lot.
I am running WRF output through the WRF PostProcessor and generating
GRIB output to verify.
And one more question, is there any tools to find out whether the flag
is correctly set in GRIB file? Or is there any tips in log file to
make
sure that the winds already have been rotated?
I need to make sure it has done in my experiment.
Yujie
On 5/17/2011 3:53 PM, RAL HelpDesk {for John Halley Gotway} wrote:
> Yujie,
>
> You are correct. The U and V wind observations in PREPBUFR files
are earth-relative while the output of WRF is grid-relative. Can you
tell me, are you running your WRF output through the WRF
> PostProcessor and generating GRIB output to verify?
>
> If so, then everything is fine. When verifying U and V, MET checks
a flag in the GRIB file to see if the winds are grid-relative. If so,
it rotates them to be earth relative prior to verifying them.
>
> If you are using the pinterp tool instead of WPP, I do not believe
that rotation from grid to earth relative is occurring. I'll make a
note of that to see how we can address it.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
> On 05/17/2011 02:12 PM, RAL HelpDesk {for Yujie Pan} wrote:
>> Tue May 17 14:12:39 2011: Request 46901 was acted upon.
>> Transaction: Ticket created by panyujie at gmail.com
>> Queue: met_help
>> Subject: Question about map projection in stat_point
>> Owner: Nobody
>> Requestors: panyujie at gmail.com
>> Status: new
>> Ticket<URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901>
>>
>>
>> To whom it may concern,
>>
>> My simulation results is from wrf, and observation is in prepbufr
>> format. Wind observation in prepbufr is in earth coordinate, while
wind
>> results from wrf output always in projection map. Observation and
>> simulation results are not in the same coordination. Is there any
wind
>> rotation included in met? or should I rotate the observation wind
before
>> pb2nc? Thank you very much.
>>
>>
>> Yujie
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46901] Question about map projection in stat_point
From: John Halley Gotway
Time: Tue May 17 15:24:58 2011
Here's a couple of tools you may find helpful.
First, I'd suggest using "wgrib" to look at header information for
each GRIB record. Try running the following command on some sample
GRIB data that's distributed with MET:
wgrib -GDS10
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212
Listed below is the output for the last two records, which are U and V
at 10 meters:
162:2148250:d=05080700:UGRD:kpds5=33:kpds6=105:kpds7=10:TR=0:P1=0:P2=0:TimeU=1:10
m above gnd:anl:NAve=0:GDS10= 0 0 42 0 255 3 0 185 0 129 0 47 158 130
9 83 8 129 115 24 0 158 187 0 158 187 0 64 0
97 168 0 97 168 0 0 0 0 0 0 0 0
163:2160186:d=05080700:VGRD:kpds5=34:kpds6=105:kpds7=10:TR=0:P1=0:P2=0:TimeU=1:10
m above gnd:anl:NAve=0:GDS10= 0 0 42 0 255 3 0 185 0 129 0 47 158 130
9 83 8 129 115 24 0 158 187 0 158 187 0 64 0
97 168 0 97 168 0 0 0 0 0 0 0 0
The -GDS10 tells wgrib to dump out the "Grid Description Section".
And here's some documentation about how to interpret the GRIB header
information:
http://www.nco.ncep.noaa.gov/pmb/docs/on388
If you look in table 7
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/table7.html), you'll see
that the flag for determining if the winds are grid or earth relative
is set by bit 5 of octet 17. That's
where MET looks for the flag. In the wgrib output above, you'll see
that the 17th entry in the GDS is 8. Now this is where it gets
confusing. Since it's set to a value of 8, that means that the bit
for 2^3 is set to a value of 1. And you count backwards here, so bit
number 8 - 3 = 5 is turned on. That means that the 5th bit is set to
one. And looking at Table 7, that means that U and V are
defined as grid-relative.
Second, you may find the tool "ncl_convert2nc" to be of use. It will
read in a GRIB file and dump it out into a NetCDF format. Try running
the following commands:
cp METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212.grb
ncl_convert2nc
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212.grb
ncdump -h wrfprs_ruc13_00.tm00_G212.nc | grep gridrot_212
That field named "gridrot_212" gives the angle of rotation necessary
for rotating the grid from grid-relative to earth-relative.
Hope that helps.
John
On 05/17/2011 03:07 PM, RAL HelpDesk {for Yujie Pan} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901 >
>
> Thank you very much, John. You help me a lot.
> I am running WRF output through the WRF PostProcessor and generating
> GRIB output to verify.
> And one more question, is there any tools to find out whether the
flag
> is correctly set in GRIB file? Or is there any tips in log file to
make
> sure that the winds already have been rotated?
> I need to make sure it has done in my experiment.
>
> Yujie
>
> On 5/17/2011 3:53 PM, RAL HelpDesk {for John Halley Gotway} wrote:
>> Yujie,
>>
>> You are correct. The U and V wind observations in PREPBUFR files
are earth-relative while the output of WRF is grid-relative. Can you
tell me, are you running your WRF output through the WRF
>> PostProcessor and generating GRIB output to verify?
>>
>> If so, then everything is fine. When verifying U and V, MET checks
a flag in the GRIB file to see if the winds are grid-relative. If so,
it rotates them to be earth relative prior to verifying them.
>>
>> If you are using the pinterp tool instead of WPP, I do not believe
that rotation from grid to earth relative is occurring. I'll make a
note of that to see how we can address it.
>>
>> Thanks,
>> John Halley Gotway
>> met_help at ucar.edu
>>
>> On 05/17/2011 02:12 PM, RAL HelpDesk {for Yujie Pan} wrote:
>>> Tue May 17 14:12:39 2011: Request 46901 was acted upon.
>>> Transaction: Ticket created by panyujie at gmail.com
>>> Queue: met_help
>>> Subject: Question about map projection in stat_point
>>> Owner: Nobody
>>> Requestors: panyujie at gmail.com
>>> Status: new
>>> Ticket<URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901>
>>>
>>>
>>> To whom it may concern,
>>>
>>> My simulation results is from wrf, and observation is in prepbufr
>>> format. Wind observation in prepbufr is in earth coordinate, while
wind
>>> results from wrf output always in projection map. Observation and
>>> simulation results are not in the same coordination. Is there any
wind
>>> rotation included in met? or should I rotate the observation wind
before
>>> pb2nc? Thank you very much.
>>>
>>>
>>> Yujie
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46901] Question about map projection in stat_point
From: Yujie Pan
Time: Tue May 17 15:50:01 2011
John, Thank you very much! I check it, and now I am sure the winds
have
rotated!
Thank you again!
Have a great day!
Yujie
On 5/17/2011 4:24 PM, RAL HelpDesk {for John Halley Gotway} wrote:
> Here's a couple of tools you may find helpful.
>
> First, I'd suggest using "wgrib" to look at header information for
each GRIB record. Try running the following command on some sample
GRIB data that's distributed with MET:
> wgrib -GDS10
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212
>
> Listed below is the output for the last two records, which are U and
V at 10 meters:
>
162:2148250:d=05080700:UGRD:kpds5=33:kpds6=105:kpds7=10:TR=0:P1=0:P2=0:TimeU=1:10
m above gnd:anl:NAve=0:GDS10= 0 0 42 0 255 3 0 185 0 129 0 47 158 130
9 83 8 129 115 24 0 158 187 0 158 187 0 64 0
> 97 168 0 97 168 0 0 0 0 0 0 0 0
>
163:2160186:d=05080700:VGRD:kpds5=34:kpds6=105:kpds7=10:TR=0:P1=0:P2=0:TimeU=1:10
m above gnd:anl:NAve=0:GDS10= 0 0 42 0 255 3 0 185 0 129 0 47 158 130
9 83 8 129 115 24 0 158 187 0 158 187 0 64 0
> 97 168 0 97 168 0 0 0 0 0 0 0 0
>
> The -GDS10 tells wgrib to dump out the "Grid Description Section".
And here's some documentation about how to interpret the GRIB header
information:
> http://www.nco.ncep.noaa.gov/pmb/docs/on388
>
> If you look in table 7
(http://www.nco.ncep.noaa.gov/pmb/docs/on388/table7.html), you'll see
that the flag for determining if the winds are grid or earth relative
is set by bit 5 of octet 17. That's
> where MET looks for the flag. In the wgrib output above, you'll see
that the 17th entry in the GDS is 8. Now this is where it gets
confusing. Since it's set to a value of 8, that means that the bit
> for 2^3 is set to a value of 1. And you count backwards here, so
bit number 8 - 3 = 5 is turned on. That means that the 5th bit is set
to one. And looking at Table 7, that means that U and V are
> defined as grid-relative.
>
> Second, you may find the tool "ncl_convert2nc" to be of use. It
will read in a GRIB file and dump it out into a NetCDF format. Try
running the following commands:
> cp
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212.grb
> ncl_convert2nc
METv3.0.1/data/sample_fcst/2005080700/wrfprs_ruc13_00.tm00_G212.grb
> ncdump -h wrfprs_ruc13_00.tm00_G212.nc | grep gridrot_212
>
> That field named "gridrot_212" gives the angle of rotation necessary
for rotating the grid from grid-relative to earth-relative.
>
> Hope that helps.
>
> John
>
>
> On 05/17/2011 03:07 PM, RAL HelpDesk {for Yujie Pan} wrote:
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901>
>>
>> Thank you very much, John. You help me a lot.
>> I am running WRF output through the WRF PostProcessor and
generating
>> GRIB output to verify.
>> And one more question, is there any tools to find out whether the
flag
>> is correctly set in GRIB file? Or is there any tips in log file to
make
>> sure that the winds already have been rotated?
>> I need to make sure it has done in my experiment.
>>
>> Yujie
>>
>> On 5/17/2011 3:53 PM, RAL HelpDesk {for John Halley Gotway} wrote:
>>> Yujie,
>>>
>>> You are correct. The U and V wind observations in PREPBUFR files
are earth-relative while the output of WRF is grid-relative. Can you
tell me, are you running your WRF output through the WRF
>>> PostProcessor and generating GRIB output to verify?
>>>
>>> If so, then everything is fine. When verifying U and V, MET
checks a flag in the GRIB file to see if the winds are grid-relative.
If so, it rotates them to be earth relative prior to verifying them.
>>>
>>> If you are using the pinterp tool instead of WPP, I do not believe
that rotation from grid to earth relative is occurring. I'll make a
note of that to see how we can address it.
>>>
>>> Thanks,
>>> John Halley Gotway
>>> met_help at ucar.edu
>>>
>>> On 05/17/2011 02:12 PM, RAL HelpDesk {for Yujie Pan} wrote:
>>>> Tue May 17 14:12:39 2011: Request 46901 was acted upon.
>>>> Transaction: Ticket created by panyujie at gmail.com
>>>> Queue: met_help
>>>> Subject: Question about map projection in stat_point
>>>> Owner: Nobody
>>>> Requestors: panyujie at gmail.com
>>>> Status: new
>>>> Ticket<URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46901>
>>>>
>>>>
>>>> To whom it may concern,
>>>>
>>>> My simulation results is from wrf, and observation is in prepbufr
>>>> format. Wind observation in prepbufr is in earth coordinate,
while wind
>>>> results from wrf output always in projection map. Observation and
>>>> simulation results are not in the same coordination. Is there any
wind
>>>> rotation included in met? or should I rotate the observation wind
before
>>>> pb2nc? Thank you very much.
>>>>
>>>>
>>>> Yujie
------------------------------------------------
More information about the Met_help
mailing list