[Met_help] [rt.rap.ucar.edu #68794] History for prepbufr/pbobs2nc question

Thu Aug 28 16:24:07 MDT 2014

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

Hello (John),

Recently I have been running validation of WRF v351 runs for the summer
of 2012 using METv4.1. I want to compare to stats generated earlier for
the same period during 2012 but using WRF v341 runs and obs processed
with METv4.0.

For the new validation, I allowed METv4.1 to download fresh prepbufr files
and
 convert to nc format. I have noticed that, not only are the ncdf files a
 different size in part due to the presence now of a quality flag, but
the count of matched pairs in the MPR files and the obs values are
actually different. This is a
big problem for an apples-to-apples stats comparison.

The differences could exist be because a) the input prepbufr files have
changed, or (more likely) b) something is different in METv4.1 vs
METv4.0. We corresponded a while ago about the problematic quality flag
in the GDAS prepbufr files (set to 9 by NCEP), plus there was a
stack-inversion problem in an earlier version of MET, however, the
details and relevancy escape me now.

I suspect that I need to regenerate stats for the WRF v341
runs using the just-created ncdf files and discard the earlier stats
generated using METv4.0.

I welcome your advice.

Thanks.

Regards,

John Henderson

AER

________________________________

This email is intended solely for the recipient. It may contain privileged, proprietary or confidential information or material. If you are not the intended recipient, please delete this email and any attachments and notify the sender of the error.

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

Subject: prepbufr/pbobs2nc question
From: John Halley Gotway
Time: Thu Aug 28 15:46:33 2014

John,

There is a write-up about a problem in extracting the virtual vs
sensible
temperature from the PREPBUFR files on the known issue pages for the
METv4.1.  Search for "pb2nc" here:
   http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php

And here's a direct link to the write-up:
   http://www.dtcenter.org/met/users/docs/write_ups/virtual_temp_bug.pdf

The differences you're seeing may be caused by this bug.  It would be
safest to rerun both analyses using the same version of the code.

Lastly, the next release of MET, version 5.0, is imminent.  It'll
likely be
next week after we finish testing using different platforms/compilers
and
combing through the documentation.  Here's a link to the release notes
describing the changes:

http://www.dtcenter.org/met/users/support/release_notes/METv5.0_release_notes.php

Applicable to you would be the addition of some new stats in the CNT
and
CTS line types.

Thanks,
John

On Thu, Aug 28, 2014 at 12:34 PM, jhenders at aer.com via RT
<met_help at ucar.edu
> wrote:

>
> Thu Aug 28 12:34:12 2014: Request 68794 was acted upon.
> Transaction: Ticket created by jhenders at aer.com
>        Queue: met_help
>      Subject: prepbufr/pbobs2nc question
>        Owner: Nobody
>   Requestors: jhenders at aer.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=68794 >
>
>
> Hello (John),
>
> Recently I have been running validation of WRF v351 runs for the
summer
> of 2012 using METv4.1. I want to compare to stats generated earlier
for
> the same period during 2012 but using WRF v341 runs and obs
processed
> with METv4.0.
>
> For the new validation, I allowed METv4.1 to download fresh prepbufr
files
> and
>  convert to nc format. I have noticed that, not only are the ncdf
files a
>  different size in part due to the presence now of a quality flag,
but
> the count of matched pairs in the MPR files and the obs values are
> actually different. This is a
> big problem for an apples-to-apples stats comparison.
>
> The differences could exist be because a) the input prepbufr files
have
> changed, or (more likely) b) something is different in METv4.1 vs
> METv4.0. We corresponded a while ago about the problematic quality
flag
> in the GDAS prepbufr files (set to 9 by NCEP), plus there was a
> stack-inversion problem in an earlier version of MET, however, the
> details and relevancy escape me now.
>
> I suspect that I need to regenerate stats for the WRF v341
> runs using the just-created ncdf files and discard the earlier stats
> generated using METv4.0.
>
> I welcome your advice.
>
> Thanks.
>
> Regards,
>
> John Henderson
>
> AER
>
>
> ________________________________
>
> This email is intended solely for the recipient. It may contain
> privileged, proprietary or confidential information or material. If
you are
> not the intended recipient, please delete this email and any
attachments
> and notify the sender of the error.
>
>
>
>

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #68794] prepbufr/pbobs2nc question
From: jhenders at aer.com
Time: Thu Aug 28 15:53:55 2014

John,

Thanks for your response. I believe that my METv4.1 is fully patched
and
it was compiled in November 2013, so I think I can successfully rerun
the
old stats (originally done with v4.0) using v4.1. The temp differences
were tenths of a degree to a degree based on a very, very quick check.

Thanks for pointing this out to me.

Regards,

John

On 8/28/14, 5:46 PM, "John Halley Gotway via RT" <met_help at ucar.edu>
wrote:

>John,
>
>There is a write-up about a problem in extracting the virtual vs
sensible
>temperature from the PREPBUFR files on the known issue pages for the
>METv4.1.  Search for "pb2nc" here:
>
>http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php
>
>And here's a direct link to the write-up:
>
http://www.dtcenter.org/met/users/docs/write_ups/virtual_temp_bug.pdf
>
>The differences you're seeing may be caused by this bug.  It would be
>safest to rerun both analyses using the same version of the code.
>
>Lastly, the next release of MET, version 5.0, is imminent.  It'll
likely
>be
>next week after we finish testing using different platforms/compilers
and
>combing through the documentation.  Here's a link to the release
notes
>describing the changes:
>
>http://www.dtcenter.org/met/users/support/release_notes/METv5.0_release_no
>tes.php
>
>Applicable to you would be the addition of some new stats in the CNT
and
>CTS line types.
>
>Thanks,
>John
>
>
>On Thu, Aug 28, 2014 at 12:34 PM, jhenders at aer.com via RT
><met_help at ucar.edu
>> wrote:
>
>>
>> Thu Aug 28 12:34:12 2014: Request 68794 was acted upon.
>> Transaction: Ticket created by jhenders at aer.com
>>        Queue: met_help
>>      Subject: prepbufr/pbobs2nc question
>>        Owner: Nobody
>>   Requestors: jhenders at aer.com
>>       Status: new
>>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=68794 >
>>
>>
>> Hello (John),
>>
>> Recently I have been running validation of WRF v351 runs for the
summer
>> of 2012 using METv4.1. I want to compare to stats generated earlier
for
>> the same period during 2012 but using WRF v341 runs and obs
processed
>> with METv4.0.
>>
>> For the new validation, I allowed METv4.1 to download fresh
prepbufr
>>files
>> and
>>  convert to nc format. I have noticed that, not only are the ncdf
files
>>a
>>  different size in part due to the presence now of a quality flag,
but
>> the count of matched pairs in the MPR files and the obs values are
>> actually different. This is a
>> big problem for an apples-to-apples stats comparison.
>>
>> The differences could exist be because a) the input prepbufr files
have
>> changed, or (more likely) b) something is different in METv4.1 vs
>> METv4.0. We corresponded a while ago about the problematic quality
flag
>> in the GDAS prepbufr files (set to 9 by NCEP), plus there was a
>> stack-inversion problem in an earlier version of MET, however, the
>> details and relevancy escape me now.
>>
>> I suspect that I need to regenerate stats for the WRF v341
>> runs using the just-created ncdf files and discard the earlier
stats
>> generated using METv4.0.
>>
>> I welcome your advice.
>>
>> Thanks.
>>
>> Regards,
>>
>> John Henderson
>>
>> AER
>>
>>
>> ________________________________
>>
>> This email is intended solely for the recipient. It may contain
>> privileged, proprietary or confidential information or material. If
you
>>are
>> not the intended recipient, please delete this email and any
attachments
>> and notify the sender of the error.
>>
>>
>>
>>
>

________________________________

This email is intended solely for the recipient. It may contain
privileged, proprietary or confidential information or material. If
you are not the intended recipient, please delete this email and any
attachments and notify the sender of the error.

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