[Met_help] [rt.rap.ucar.edu #69099] History for Comparing WRF-ARW outputs and CCSM4 outputs
John Halley Gotway via RT
met_help at ucar.edu
Fri Oct 24 10:33:31 MDT 2014
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Dear Met Help,
Hi.
I am looking for a way to verify WRF model performance as a climate model.
I'm completed 10-year tropical channel climate model simulations using the
CCSM4 data as input data (dynamical downscaling) for two different time
periods: one for the LGM and the other for the modern as a control run.
The WRF simulation produced 6 hourly output for ten years periods each.
Also the CCSM4 simulations come with 6 hourly outputs.
Basically, I'd like to compare the two simulation results, verifying the
WRF model simulations.
(WRF ARW was implemented for the experiments.)
Series-Analysis in the MET v5.0 seems to do the similar jobs but I need
much more help from you than the MET Online Tutorial to make a step forward.
Any help will be highly appreciated.
Thank you.
Regards,
Jinwoong Yoo
Postdoc
Department of Earth and Planetary Sciences
UNM
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Thu Sep 25 08:52:38 2014
Jinwoong,
I see that you'd like some help running the series-analysis tool on
your
wrf-arw climate runs. One thing to note is that MET can't read the
raw
wrfout files directly. Instead you first need to post process them
using
either UPP (produces GRIB output) or the p_interp utility (produced
NetCDF
output). After you've post processed some files, please send us some
sample data by posting it to our anonymous ftp site. Also send a
description of the comparisons you'd like to make between the 2
datasets.
Here's instructions for sending us data:
http://www.dtcenter.org/met/users/support/met_help.php
Thanks
John Halley Gotway
met_help at ucar.edu
On Wednesday, September 24, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> Transaction: Ticket created by jinwoong.yoo at gmail.com <javascript:;>
> Queue: met_help
> Subject: Comparing WRF-ARW outputs and CCSM4 outputs
> Owner: Nobody
> Requestors: jinwoong.yoo at gmail.com <javascript:;>
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
>
> Dear Met Help,
>
> Hi.
> I am looking for a way to verify WRF model performance as a climate
model.
> I'm completed 10-year tropical channel climate model simulations
using the
> CCSM4 data as input data (dynamical downscaling) for two different
time
> periods: one for the LGM and the other for the modern as a control
run.
> The WRF simulation produced 6 hourly output for ten years periods
each.
> Also the CCSM4 simulations come with 6 hourly outputs.
>
> Basically, I'd like to compare the two simulation results, verifying
the
> WRF model simulations.
> (WRF ARW was implemented for the experiments.)
> Series-Analysis in the MET v5.0 seems to do the similar jobs but I
need
> much more help from you than the MET Online Tutorial to make a step
> forward.
> Any help will be highly appreciated.
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> Postdoc
> Department of Earth and Planetary Sciences
> UNM
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Sep 25 09:44:29 2014
Dear John,
Thank you for your kind email with the direction.
Let me first post process some of my WRF-ARW outputs using UPP and let
you
know.
Thank you very much.
Regards,
Jinwoong Yoo
UNM
On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Jinwoong,
>
> I see that you'd like some help running the series-analysis tool on
your
> wrf-arw climate runs. One thing to note is that MET can't read the
raw
> wrfout files directly. Instead you first need to post process them
using
> either UPP (produces GRIB output) or the p_interp utility (produced
NetCDF
> output). After you've post processed some files, please send us
some
> sample data by posting it to our anonymous ftp site. Also send a
> description of the comparisons you'd like to make between the 2
datasets.
>
> Here's instructions for sending us data:
> http://www.dtcenter.org/met/users/support/met_help.php
>
> Thanks
> John Halley Gotway
> met_help at ucar.edu
>
> On Wednesday, September 24, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> > Transaction: Ticket created by jinwoong.yoo at gmail.com
<javascript:;>
> > Queue: met_help
> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
> > Owner: Nobody
> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> >
> > Dear Met Help,
> >
> > Hi.
> > I am looking for a way to verify WRF model performance as a
climate
> model.
> > I'm completed 10-year tropical channel climate model simulations
using
> the
> > CCSM4 data as input data (dynamical downscaling) for two different
time
> > periods: one for the LGM and the other for the modern as a control
run.
> > The WRF simulation produced 6 hourly output for ten years periods
each.
> > Also the CCSM4 simulations come with 6 hourly outputs.
> >
> > Basically, I'd like to compare the two simulation results,
verifying the
> > WRF model simulations.
> > (WRF ARW was implemented for the experiments.)
> > Series-Analysis in the MET v5.0 seems to do the similar jobs but I
need
> > much more help from you than the MET Online Tutorial to make a
step
> > forward.
> > Any help will be highly appreciated.
> > Thank you.
> >
> > Regards,
> >
> > Jinwoong Yoo
> > Postdoc
> > Department of Earth and Planetary Sciences
> > UNM
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Sep 25 10:34:16 2014
Dear John,
By the way, what about the CCSM4 file in NetCDF format?
Do I need to post process of the CCSM4 files also to run MET along
with the
WRF-ARW files?
Thank you.
Regards,
Jinwoong Yoo
UNM
On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> Dear John,
>
> Thank you for your kind email with the direction.
> Let me first post process some of my WRF-ARW outputs using UPP and
let you
> know.
> Thank you very much.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
>> Jinwoong,
>>
>> I see that you'd like some help running the series-analysis tool on
your
>> wrf-arw climate runs. One thing to note is that MET can't read the
raw
>> wrfout files directly. Instead you first need to post process them
using
>> either UPP (produces GRIB output) or the p_interp utility (produced
NetCDF
>> output). After you've post processed some files, please send us
some
>> sample data by posting it to our anonymous ftp site. Also send a
>> description of the comparisons you'd like to make between the 2
datasets.
>>
>> Here's instructions for sending us data:
>> http://www.dtcenter.org/met/users/support/met_help.php
>>
>> Thanks
>> John Halley Gotway
>> met_help at ucar.edu
>>
>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
>> wrote:
>>
>> >
>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
<javascript:;>
>> > Queue: met_help
>> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
>> > Owner: Nobody
>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
>> > Status: new
>> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>> >
>> >
>> > Dear Met Help,
>> >
>> > Hi.
>> > I am looking for a way to verify WRF model performance as a
climate
>> model.
>> > I'm completed 10-year tropical channel climate model simulations
using
>> the
>> > CCSM4 data as input data (dynamical downscaling) for two
different time
>> > periods: one for the LGM and the other for the modern as a
control run.
>> > The WRF simulation produced 6 hourly output for ten years periods
each.
>> > Also the CCSM4 simulations come with 6 hourly outputs.
>> >
>> > Basically, I'd like to compare the two simulation results,
verifying the
>> > WRF model simulations.
>> > (WRF ARW was implemented for the experiments.)
>> > Series-Analysis in the MET v5.0 seems to do the similar jobs but
I need
>> > much more help from you than the MET Online Tutorial to make a
step
>> > forward.
>> > Any help will be highly appreciated.
>> > Thank you.
>> >
>> > Regards,
>> >
>> > Jinwoong Yoo
>> > Postdoc
>> > Department of Earth and Planetary Sciences
>> > UNM
>> >
>> >
>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Sep 25 15:13:39 2014
Dear John,
I ran UPP with my multiple wrfout files, first with the first month of
the
10 year simulation. However, after executing the run_unipost script,
UPP
produced only four files: unipost_d01.00.out, unipost_d01.06.out,
unipost_d01.12.out, and unipost_d01.18.out. It seems that UPP
processed
only the first day's output among the one month data.
If you are familiar with UPP, how can I make sure for UPP to process
multiple wrfout files more than one day?
Thank you.
Regards,
Jinwoong Yoo
UNM
On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
wrote:
> Dear John,
>
> By the way, what about the CCSM4 file in NetCDF format?
> Do I need to post process of the CCSM4 files also to run MET along
with
> the WRF-ARW files?
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> Dear John,
>>
>> Thank you for your kind email with the direction.
>> Let me first post process some of my WRF-ARW outputs using UPP and
let
>> you know.
>> Thank you very much.
>>
>> Regards,
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
>> met_help at ucar.edu> wrote:
>>
>>> Jinwoong,
>>>
>>> I see that you'd like some help running the series-analysis tool
on your
>>> wrf-arw climate runs. One thing to note is that MET can't read
the raw
>>> wrfout files directly. Instead you first need to post process
them using
>>> either UPP (produces GRIB output) or the p_interp utility
(produced
>>> NetCDF
>>> output). After you've post processed some files, please send us
some
>>> sample data by posting it to our anonymous ftp site. Also send a
>>> description of the comparisons you'd like to make between the 2
datasets.
>>>
>>> Here's instructions for sending us data:
>>> http://www.dtcenter.org/met/users/support/met_help.php
>>>
>>> Thanks
>>> John Halley Gotway
>>> met_help at ucar.edu
>>>
>>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu
>>> >
>>> wrote:
>>>
>>> >
>>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
>>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
<javascript:;>
>>> > Queue: met_help
>>> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
>>> > Owner: Nobody
>>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
>>> > Status: new
>>> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>> >
>>> >
>>> >
>>> > Dear Met Help,
>>> >
>>> > Hi.
>>> > I am looking for a way to verify WRF model performance as a
climate
>>> model.
>>> > I'm completed 10-year tropical channel climate model simulations
using
>>> the
>>> > CCSM4 data as input data (dynamical downscaling) for two
different time
>>> > periods: one for the LGM and the other for the modern as a
control run.
>>> > The WRF simulation produced 6 hourly output for ten years
periods each.
>>> > Also the CCSM4 simulations come with 6 hourly outputs.
>>> >
>>> > Basically, I'd like to compare the two simulation results,
verifying
>>> the
>>> > WRF model simulations.
>>> > (WRF ARW was implemented for the experiments.)
>>> > Series-Analysis in the MET v5.0 seems to do the similar jobs but
I need
>>> > much more help from you than the MET Online Tutorial to make a
step
>>> > forward.
>>> > Any help will be highly appreciated.
>>> > Thank you.
>>> >
>>> > Regards,
>>> >
>>> > Jinwoong Yoo
>>> > Postdoc
>>> > Department of Earth and Planetary Sciences
>>> > UNM
>>> >
>>> >
>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Fri Sep 26 20:21:31 2014
Jinwoong,
UPP support is typically provided through wrf_help at ucar.edu. I'll
give it
a shot, but if this doesn't solve it, I'll refer you to wrf_help.
I assume you're running the script named "run_unipost". Looking in
there,
you'll see:
export startdate=2005012300
export fhr=00
export lastfhr=18
export incrementhr=03
I'd suggest modifying the...
- "startdate" to be the model initialization time to be processed.
- Leave "fhr" set to 00 to start at the analysis time.
- Set "lastfhr" to be maybe 744 (which is 31 days * 24 hours/day).
- Set "incrementhr" to 06 to process output every 6 hours.
Give that a shot and let me know if that gets you output past the
first day.
Are you going to be evaluating precipitation? I ask that because
precip is
an accumulated interval, meaning that it has a beginning accumulation
time
and ending accumulation time. GRIB1 is limited in the amount of space
available for storing times. When you get past 255 forecast hours,
you can
no longer store beginning/ending times for precip because it runs out
of
room in the header. Do you have a 10 year simulation computed from a
single initialization? Meaning, do you have forecast lead times out
to
87,600 hours (= 10 years * 365 days * 24 hours)? I'm wondering if
we'll
run into problems formatting those time strings.
Thanks,
John
On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Dear John,
>
> I ran UPP with my multiple wrfout files, first with the first month
of the
> 10 year simulation. However, after executing the run_unipost script,
UPP
> produced only four files: unipost_d01.00.out, unipost_d01.06.out,
> unipost_d01.12.out, and unipost_d01.18.out. It seems that UPP
processed
> only the first day's output among the one month data.
> If you are familiar with UPP, how can I make sure for UPP to process
> multiple wrfout files more than one day?
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
> > Dear John,
> >
> > By the way, what about the CCSM4 file in NetCDF format?
> > Do I need to post process of the CCSM4 files also to run MET along
with
> > the WRF-ARW files?
> > Thank you.
> >
> > Regards,
> >
> > Jinwoong Yoo
> > UNM
> >
> > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > wrote:
> >
> >> Dear John,
> >>
> >> Thank you for your kind email with the direction.
> >> Let me first post process some of my WRF-ARW outputs using UPP
and let
> >> you know.
> >> Thank you very much.
> >>
> >> Regards,
> >>
> >> Jinwoong Yoo
> >> UNM
> >>
> >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
> >> met_help at ucar.edu> wrote:
> >>
> >>> Jinwoong,
> >>>
> >>> I see that you'd like some help running the series-analysis tool
on
> your
> >>> wrf-arw climate runs. One thing to note is that MET can't read
the raw
> >>> wrfout files directly. Instead you first need to post process
them
> using
> >>> either UPP (produces GRIB output) or the p_interp utility
(produced
> >>> NetCDF
> >>> output). After you've post processed some files, please send us
some
> >>> sample data by posting it to our anonymous ftp site. Also send
a
> >>> description of the comparisons you'd like to make between the 2
> datasets.
> >>>
> >>> Here's instructions for sending us data:
> >>> http://www.dtcenter.org/met/users/support/met_help.php
> >>>
> >>> Thanks
> >>> John Halley Gotway
> >>> met_help at ucar.edu
> >>>
> >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT <
> met_help at ucar.edu
> >>> >
> >>> wrote:
> >>>
> >>> >
> >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> >>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
<javascript:;>
> >>> > Queue: met_help
> >>> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
> >>> > Owner: Nobody
> >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> >>> > Status: new
> >>> > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>> >
> >>> >
> >>> >
> >>> > Dear Met Help,
> >>> >
> >>> > Hi.
> >>> > I am looking for a way to verify WRF model performance as a
climate
> >>> model.
> >>> > I'm completed 10-year tropical channel climate model
simulations
> using
> >>> the
> >>> > CCSM4 data as input data (dynamical downscaling) for two
different
> time
> >>> > periods: one for the LGM and the other for the modern as a
control
> run.
> >>> > The WRF simulation produced 6 hourly output for ten years
periods
> each.
> >>> > Also the CCSM4 simulations come with 6 hourly outputs.
> >>> >
> >>> > Basically, I'd like to compare the two simulation results,
verifying
> >>> the
> >>> > WRF model simulations.
> >>> > (WRF ARW was implemented for the experiments.)
> >>> > Series-Analysis in the MET v5.0 seems to do the similar jobs
but I
> need
> >>> > much more help from you than the MET Online Tutorial to make a
step
> >>> > forward.
> >>> > Any help will be highly appreciated.
> >>> > Thank you.
> >>> >
> >>> > Regards,
> >>> >
> >>> > Jinwoong Yoo
> >>> > Postdoc
> >>> > Department of Earth and Planetary Sciences
> >>> > UNM
> >>> >
> >>> >
> >>>
> >>>
> >>
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Fri Sep 26 21:46:25 2014
Dear John,
I see. Let me change "export lastfhr=18" to "export lastfhr=744".
Thank you for your comments.
Yes, I'd like to evaluate precipitation, too, and my 10 year
simulation was
integrated from a single initialization. Then, will the UPP
run_unipost
crash after the 255 forecast hours or just will precipitation not be
stored
any more aster the 255 forecast hours?
Thank you.
Regards,
Jinwoong Yoo
UNM
On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Jinwoong,
>
> UPP support is typically provided through wrf_help at ucar.edu. I'll
give it
> a shot, but if this doesn't solve it, I'll refer you to wrf_help.
>
> I assume you're running the script named "run_unipost". Looking in
there,
> you'll see:
>
> export startdate=2005012300
>
> export fhr=00
> export lastfhr=18
> export incrementhr=03
>
> I'd suggest modifying the...
> - "startdate" to be the model initialization time to be processed.
> - Leave "fhr" set to 00 to start at the analysis time.
> - Set "lastfhr" to be maybe 744 (which is 31 days * 24 hours/day).
> - Set "incrementhr" to 06 to process output every 6 hours.
>
> Give that a shot and let me know if that gets you output past the
first
> day.
>
> Are you going to be evaluating precipitation? I ask that because
precip is
> an accumulated interval, meaning that it has a beginning
accumulation time
> and ending accumulation time. GRIB1 is limited in the amount of
space
> available for storing times. When you get past 255 forecast hours,
you can
> no longer store beginning/ending times for precip because it runs
out of
> room in the header. Do you have a 10 year simulation computed from
a
> single initialization? Meaning, do you have forecast lead times out
to
> 87,600 hours (= 10 years * 365 days * 24 hours)? I'm wondering if
we'll
> run into problems formatting those time strings.
>
> Thanks,
> John
>
>
> On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> > Dear John,
> >
> > I ran UPP with my multiple wrfout files, first with the first
month of
> the
> > 10 year simulation. However, after executing the run_unipost
script, UPP
> > produced only four files: unipost_d01.00.out, unipost_d01.06.out,
> > unipost_d01.12.out, and unipost_d01.18.out. It seems that UPP
processed
> > only the first day's output among the one month data.
> > If you are familiar with UPP, how can I make sure for UPP to
process
> > multiple wrfout files more than one day?
> > Thank you.
> >
> > Regards,
> >
> > Jinwoong Yoo
> > UNM
> >
> > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > wrote:
> >
> > > Dear John,
> > >
> > > By the way, what about the CCSM4 file in NetCDF format?
> > > Do I need to post process of the CCSM4 files also to run MET
along with
> > > the WRF-ARW files?
> > > Thank you.
> > >
> > > Regards,
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > > wrote:
> > >
> > >> Dear John,
> > >>
> > >> Thank you for your kind email with the direction.
> > >> Let me first post process some of my WRF-ARW outputs using UPP
and let
> > >> you know.
> > >> Thank you very much.
> > >>
> > >> Regards,
> > >>
> > >> Jinwoong Yoo
> > >> UNM
> > >>
> > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
> > >> met_help at ucar.edu> wrote:
> > >>
> > >>> Jinwoong,
> > >>>
> > >>> I see that you'd like some help running the series-analysis
tool on
> > your
> > >>> wrf-arw climate runs. One thing to note is that MET can't
read the
> raw
> > >>> wrfout files directly. Instead you first need to post process
them
> > using
> > >>> either UPP (produces GRIB output) or the p_interp utility
(produced
> > >>> NetCDF
> > >>> output). After you've post processed some files, please send
us some
> > >>> sample data by posting it to our anonymous ftp site. Also
send a
> > >>> description of the comparisons you'd like to make between the
2
> > datasets.
> > >>>
> > >>> Here's instructions for sending us data:
> > >>> http://www.dtcenter.org/met/users/support/met_help.php
> > >>>
> > >>> Thanks
> > >>> John Halley Gotway
> > >>> met_help at ucar.edu
> > >>>
> > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT <
> > met_help at ucar.edu
> > >>> >
> > >>> wrote:
> > >>>
> > >>> >
> > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> > >>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
> <javascript:;>
> > >>> > Queue: met_help
> > >>> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
> > >>> > Owner: Nobody
> > >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> > >>> > Status: new
> > >>> > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >>> >
> > >>> >
> > >>> >
> > >>> > Dear Met Help,
> > >>> >
> > >>> > Hi.
> > >>> > I am looking for a way to verify WRF model performance as a
climate
> > >>> model.
> > >>> > I'm completed 10-year tropical channel climate model
simulations
> > using
> > >>> the
> > >>> > CCSM4 data as input data (dynamical downscaling) for two
different
> > time
> > >>> > periods: one for the LGM and the other for the modern as a
control
> > run.
> > >>> > The WRF simulation produced 6 hourly output for ten years
periods
> > each.
> > >>> > Also the CCSM4 simulations come with 6 hourly outputs.
> > >>> >
> > >>> > Basically, I'd like to compare the two simulation results,
> verifying
> > >>> the
> > >>> > WRF model simulations.
> > >>> > (WRF ARW was implemented for the experiments.)
> > >>> > Series-Analysis in the MET v5.0 seems to do the similar jobs
but I
> > need
> > >>> > much more help from you than the MET Online Tutorial to make
a step
> > >>> > forward.
> > >>> > Any help will be highly appreciated.
> > >>> > Thank you.
> > >>> >
> > >>> > Regards,
> > >>> >
> > >>> > Jinwoong Yoo
> > >>> > Postdoc
> > >>> > Department of Earth and Planetary Sciences
> > >>> > UNM
> > >>> >
> > >>> >
> > >>>
> > >>>
> > >>
> > >
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Fri Sep 26 23:36:31 2014
Jinwoo,
I believe UPP will still run, just using a single time value as if it
were
an instantaneous field rather than an accumulation interval. The
default
for ARW is to store precip as a runtime accumulation. For forecasts
hours
< 255, like 12, I suspect it'll be stored in GRIB as an accumulation
from 0
to 12. For hours > 255, like 300, I suspect it'll be stored as an
instantaneous 300 hour forecast. It's a subtle difference, but
something
to be aware of when you set up MET config files to verify the output.
Give UPP a try and let me know how it goes.
John
On Friday, September 26, 2014, Jinwoong Yoo via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Dear John,
>
> I see. Let me change "export lastfhr=18" to "export lastfhr=744".
> Thank you for your comments.
> Yes, I'd like to evaluate precipitation, too, and my 10 year
simulation was
> integrated from a single initialization. Then, will the UPP
run_unipost
> crash after the 255 forecast hours or just will precipitation not be
stored
> any more aster the 255 forecast hours?
>
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
> met_help at ucar.edu <javascript:;>> wrote:
>
> > Jinwoong,
> >
> > UPP support is typically provided through wrf_help at ucar.edu
> <javascript:;>. I'll give it
> > a shot, but if this doesn't solve it, I'll refer you to wrf_help.
> >
> > I assume you're running the script named "run_unipost". Looking
in
> there,
> > you'll see:
> >
> > export startdate=2005012300
> >
> > export fhr=00
> > export lastfhr=18
> > export incrementhr=03
> >
> > I'd suggest modifying the...
> > - "startdate" to be the model initialization time to be
processed.
> > - Leave "fhr" set to 00 to start at the analysis time.
> > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
hours/day).
> > - Set "incrementhr" to 06 to process output every 6 hours.
> >
> > Give that a shot and let me know if that gets you output past the
first
> > day.
> >
> > Are you going to be evaluating precipitation? I ask that because
precip
> is
> > an accumulated interval, meaning that it has a beginning
accumulation
> time
> > and ending accumulation time. GRIB1 is limited in the amount of
space
> > available for storing times. When you get past 255 forecast
hours, you
> can
> > no longer store beginning/ending times for precip because it runs
out of
> > room in the header. Do you have a 10 year simulation computed
from a
> > single initialization? Meaning, do you have forecast lead times
out to
> > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm wondering if
we'll
> > run into problems formatting those time strings.
> >
> > Thanks,
> > John
> >
> >
> > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT
<met_help at ucar.edu
> <javascript:;>>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > >
> > > Dear John,
> > >
> > > I ran UPP with my multiple wrfout files, first with the first
month of
> > the
> > > 10 year simulation. However, after executing the run_unipost
script,
> UPP
> > > produced only four files: unipost_d01.00.out,
unipost_d01.06.out,
> > > unipost_d01.12.out, and unipost_d01.18.out. It seems that UPP
processed
> > > only the first day's output among the one month data.
> > > If you are familiar with UPP, how can I make sure for UPP to
process
> > > multiple wrfout files more than one day?
> > > Thank you.
> > >
> > > Regards,
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com
> <javascript:;>>
> > > wrote:
> > >
> > > > Dear John,
> > > >
> > > > By the way, what about the CCSM4 file in NetCDF format?
> > > > Do I need to post process of the CCSM4 files also to run MET
along
> with
> > > > the WRF-ARW files?
> > > > Thank you.
> > > >
> > > > Regards,
> > > >
> > > > Jinwoong Yoo
> > > > UNM
> > > >
> > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com <javascript:;>>
> > > > wrote:
> > > >
> > > >> Dear John,
> > > >>
> > > >> Thank you for your kind email with the direction.
> > > >> Let me first post process some of my WRF-ARW outputs using
UPP and
> let
> > > >> you know.
> > > >> Thank you very much.
> > > >>
> > > >> Regards,
> > > >>
> > > >> Jinwoong Yoo
> > > >> UNM
> > > >>
> > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT <
> > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > >>
> > > >>> Jinwoong,
> > > >>>
> > > >>> I see that you'd like some help running the series-analysis
tool on
> > > your
> > > >>> wrf-arw climate runs. One thing to note is that MET can't
read the
> > raw
> > > >>> wrfout files directly. Instead you first need to post
process them
> > > using
> > > >>> either UPP (produces GRIB output) or the p_interp utility
(produced
> > > >>> NetCDF
> > > >>> output). After you've post processed some files, please
send us
> some
> > > >>> sample data by posting it to our anonymous ftp site. Also
send a
> > > >>> description of the comparisons you'd like to make between
the 2
> > > datasets.
> > > >>>
> > > >>> Here's instructions for sending us data:
> > > >>> http://www.dtcenter.org/met/users/support/met_help.php
> > > >>>
> > > >>> Thanks
> > > >>> John Halley Gotway
> > > >>> met_help at ucar.edu <javascript:;>
> > > >>>
> > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT <
> > > met_help at ucar.edu <javascript:;>
> > > >>> >
> > > >>> wrote:
> > > >>>
> > > >>> >
> > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> > > >>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
> <javascript:;>
> > <javascript:;>
> > > >>> > Queue: met_help
> > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4 outputs
> > > >>> > Owner: Nobody
> > > >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> <javascript:;>
> > > >>> > Status: new
> > > >>> > Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > >>> >
> > > >>> >
> > > >>> >
> > > >>> > Dear Met Help,
> > > >>> >
> > > >>> > Hi.
> > > >>> > I am looking for a way to verify WRF model performance as
a
> climate
> > > >>> model.
> > > >>> > I'm completed 10-year tropical channel climate model
simulations
> > > using
> > > >>> the
> > > >>> > CCSM4 data as input data (dynamical downscaling) for two
> different
> > > time
> > > >>> > periods: one for the LGM and the other for the modern as a
> control
> > > run.
> > > >>> > The WRF simulation produced 6 hourly output for ten years
periods
> > > each.
> > > >>> > Also the CCSM4 simulations come with 6 hourly outputs.
> > > >>> >
> > > >>> > Basically, I'd like to compare the two simulation results,
> > verifying
> > > >>> the
> > > >>> > WRF model simulations.
> > > >>> > (WRF ARW was implemented for the experiments.)
> > > >>> > Series-Analysis in the MET v5.0 seems to do the similar
jobs but
> I
> > > need
> > > >>> > much more help from you than the MET Online Tutorial to
make a
> step
> > > >>> > forward.
> > > >>> > Any help will be highly appreciated.
> > > >>> > Thank you.
> > > >>> >
> > > >>> > Regards,
> > > >>> >
> > > >>> > Jinwoong Yoo
> > > >>> > Postdoc
> > > >>> > Department of Earth and Planetary Sciences
> > > >>> > UNM
> > > >>> >
> > > >>> >
> > > >>>
> > > >>>
> > > >>
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Tue Sep 30 14:26:32 2014
Dear John,
I ran UPP successfully on the UCAR Yellowstone.
I assume you can access to my glade scratch directory.
My UPP ouputs are located
under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
Against those files, I'd like to compare CCSM4 simulation outputs some
of
which are located also under
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
Those are six hourly files all together. If you cannot access to those
files, please let me know.
However, there are some variables among the CCSM4 outputs that are
available only monthly or annually. In this case, how can I post-
process my
WRF outputs for the usage of MET through UPP?
Thank you.
Regards,
Jinwoong Yoo
UNM
On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Jinwoo,
>
> I believe UPP will still run, just using a single time value as if
it were
> an instantaneous field rather than an accumulation interval. The
default
> for ARW is to store precip as a runtime accumulation. For forecasts
hours
> < 255, like 12, I suspect it'll be stored in GRIB as an accumulation
from 0
> to 12. For hours > 255, like 300, I suspect it'll be stored as an
> instantaneous 300 hour forecast. It's a subtle difference, but
something
> to be aware of when you set up MET config files to verify the
output.
>
> Give UPP a try and let me know how it goes.
>
> John
>
> On Friday, September 26, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> > Dear John,
> >
> > I see. Let me change "export lastfhr=18" to "export lastfhr=744".
> > Thank you for your comments.
> > Yes, I'd like to evaluate precipitation, too, and my 10 year
simulation
> was
> > integrated from a single initialization. Then, will the UPP
run_unipost
> > crash after the 255 forecast hours or just will precipitation not
be
> stored
> > any more aster the 255 forecast hours?
> >
> > Thank you.
> >
> > Regards,
> >
> > Jinwoong Yoo
> > UNM
> >
> > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
> > met_help at ucar.edu <javascript:;>> wrote:
> >
> > > Jinwoong,
> > >
> > > UPP support is typically provided through wrf_help at ucar.edu
> > <javascript:;>. I'll give it
> > > a shot, but if this doesn't solve it, I'll refer you to
wrf_help.
> > >
> > > I assume you're running the script named "run_unipost". Looking
in
> > there,
> > > you'll see:
> > >
> > > export startdate=2005012300
> > >
> > > export fhr=00
> > > export lastfhr=18
> > > export incrementhr=03
> > >
> > > I'd suggest modifying the...
> > > - "startdate" to be the model initialization time to be
processed.
> > > - Leave "fhr" set to 00 to start at the analysis time.
> > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
hours/day).
> > > - Set "incrementhr" to 06 to process output every 6 hours.
> > >
> > > Give that a shot and let me know if that gets you output past
the first
> > > day.
> > >
> > > Are you going to be evaluating precipitation? I ask that
because
> precip
> > is
> > > an accumulated interval, meaning that it has a beginning
accumulation
> > time
> > > and ending accumulation time. GRIB1 is limited in the amount of
space
> > > available for storing times. When you get past 255 forecast
hours, you
> > can
> > > no longer store beginning/ending times for precip because it
runs out
> of
> > > room in the header. Do you have a 10 year simulation computed
from a
> > > single initialization? Meaning, do you have forecast lead times
out to
> > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm wondering
if
> we'll
> > > run into problems formatting those time strings.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
> met_help at ucar.edu
> > <javascript:;>>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
> > > >
> > > > Dear John,
> > > >
> > > > I ran UPP with my multiple wrfout files, first with the first
month
> of
> > > the
> > > > 10 year simulation. However, after executing the run_unipost
script,
> > UPP
> > > > produced only four files: unipost_d01.00.out,
unipost_d01.06.out,
> > > > unipost_d01.12.out, and unipost_d01.18.out. It seems that UPP
> processed
> > > > only the first day's output among the one month data.
> > > > If you are familiar with UPP, how can I make sure for UPP to
process
> > > > multiple wrfout files more than one day?
> > > > Thank you.
> > > >
> > > > Regards,
> > > >
> > > > Jinwoong Yoo
> > > > UNM
> > > >
> > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com
> > <javascript:;>>
> > > > wrote:
> > > >
> > > > > Dear John,
> > > > >
> > > > > By the way, what about the CCSM4 file in NetCDF format?
> > > > > Do I need to post process of the CCSM4 files also to run MET
along
> > with
> > > > > the WRF-ARW files?
> > > > > Thank you.
> > > > >
> > > > > Regards,
> > > > >
> > > > > Jinwoong Yoo
> > > > > UNM
> > > > >
> > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> > jinwoong.yoo at gmail.com <javascript:;>>
> > > > > wrote:
> > > > >
> > > > >> Dear John,
> > > > >>
> > > > >> Thank you for your kind email with the direction.
> > > > >> Let me first post process some of my WRF-ARW outputs using
UPP and
> > let
> > > > >> you know.
> > > > >> Thank you very much.
> > > > >>
> > > > >> Regards,
> > > > >>
> > > > >> Jinwoong Yoo
> > > > >> UNM
> > > > >>
> > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via RT
<
> > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > > >>
> > > > >>> Jinwoong,
> > > > >>>
> > > > >>> I see that you'd like some help running the series-
analysis tool
> on
> > > > your
> > > > >>> wrf-arw climate runs. One thing to note is that MET can't
read
> the
> > > raw
> > > > >>> wrfout files directly. Instead you first need to post
process
> them
> > > > using
> > > > >>> either UPP (produces GRIB output) or the p_interp utility
> (produced
> > > > >>> NetCDF
> > > > >>> output). After you've post processed some files, please
send us
> > some
> > > > >>> sample data by posting it to our anonymous ftp site. Also
send a
> > > > >>> description of the comparisons you'd like to make between
the 2
> > > > datasets.
> > > > >>>
> > > > >>> Here's instructions for sending us data:
> > > > >>> http://www.dtcenter.org/met/users/support/met_help.php
> > > > >>>
> > > > >>> Thanks
> > > > >>> John Halley Gotway
> > > > >>> met_help at ucar.edu <javascript:;>
> > > > >>>
> > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT <
> > > > met_help at ucar.edu <javascript:;>
> > > > >>> >
> > > > >>> wrote:
> > > > >>>
> > > > >>> >
> > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted upon.
> > > > >>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
> > <javascript:;>
> > > <javascript:;>
> > > > >>> > Queue: met_help
> > > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4
outputs
> > > > >>> > Owner: Nobody
> > > > >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> > <javascript:;>
> > > > >>> > Status: new
> > > > >>> > Ticket <URL:
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > > >>> >
> > > > >>> >
> > > > >>> >
> > > > >>> > Dear Met Help,
> > > > >>> >
> > > > >>> > Hi.
> > > > >>> > I am looking for a way to verify WRF model performance
as a
> > climate
> > > > >>> model.
> > > > >>> > I'm completed 10-year tropical channel climate model
> simulations
> > > > using
> > > > >>> the
> > > > >>> > CCSM4 data as input data (dynamical downscaling) for two
> > different
> > > > time
> > > > >>> > periods: one for the LGM and the other for the modern as
a
> > control
> > > > run.
> > > > >>> > The WRF simulation produced 6 hourly output for ten
years
> periods
> > > > each.
> > > > >>> > Also the CCSM4 simulations come with 6 hourly outputs.
> > > > >>> >
> > > > >>> > Basically, I'd like to compare the two simulation
results,
> > > verifying
> > > > >>> the
> > > > >>> > WRF model simulations.
> > > > >>> > (WRF ARW was implemented for the experiments.)
> > > > >>> > Series-Analysis in the MET v5.0 seems to do the similar
jobs
> but
> > I
> > > > need
> > > > >>> > much more help from you than the MET Online Tutorial to
make a
> > step
> > > > >>> > forward.
> > > > >>> > Any help will be highly appreciated.
> > > > >>> > Thank you.
> > > > >>> >
> > > > >>> > Regards,
> > > > >>> >
> > > > >>> > Jinwoong Yoo
> > > > >>> > Postdoc
> > > > >>> > Department of Earth and Planetary Sciences
> > > > >>> > UNM
> > > > >>> >
> > > > >>> >
> > > > >>>
> > > > >>>
> > > > >>
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Julie Prestopnik
Time: Tue Sep 30 15:35:15 2014
Hello. John is out of the office, but will be back later this week. I
am
unable to help you with this task. Hopefully, John will be able to
respond
to this ticket later this week.
Julie
On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Dear John,
>
> I ran UPP successfully on the UCAR Yellowstone.
> I assume you can access to my glade scratch directory.
> My UPP ouputs are located
> under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>
> Against those files, I'd like to compare CCSM4 simulation outputs
some of
> which are located also under
> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>
> Those are six hourly files all together. If you cannot access to
those
> files, please let me know.
>
> However, there are some variables among the CCSM4 outputs that are
> available only monthly or annually. In this case, how can I post-
process my
> WRF outputs for the usage of MET through UPP?
>
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
>
>
>
> On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Jinwoo,
> >
> > I believe UPP will still run, just using a single time value as if
it
> were
> > an instantaneous field rather than an accumulation interval. The
default
> > for ARW is to store precip as a runtime accumulation. For
forecasts
> hours
> > < 255, like 12, I suspect it'll be stored in GRIB as an
accumulation
> from 0
> > to 12. For hours > 255, like 300, I suspect it'll be stored as an
> > instantaneous 300 hour forecast. It's a subtle difference, but
something
> > to be aware of when you set up MET config files to verify the
output.
> >
> > Give UPP a try and let me know how it goes.
> >
> > John
> >
> > On Friday, September 26, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > >
> > > Dear John,
> > >
> > > I see. Let me change "export lastfhr=18" to "export
lastfhr=744".
> > > Thank you for your comments.
> > > Yes, I'd like to evaluate precipitation, too, and my 10 year
simulation
> > was
> > > integrated from a single initialization. Then, will the UPP
run_unipost
> > > crash after the 255 forecast hours or just will precipitation
not be
> > stored
> > > any more aster the 255 forecast hours?
> > >
> > > Thank you.
> > >
> > > Regards,
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu <javascript:;>> wrote:
> > >
> > > > Jinwoong,
> > > >
> > > > UPP support is typically provided through wrf_help at ucar.edu
> > > <javascript:;>. I'll give it
> > > > a shot, but if this doesn't solve it, I'll refer you to
wrf_help.
> > > >
> > > > I assume you're running the script named "run_unipost".
Looking in
> > > there,
> > > > you'll see:
> > > >
> > > > export startdate=2005012300
> > > >
> > > > export fhr=00
> > > > export lastfhr=18
> > > > export incrementhr=03
> > > >
> > > > I'd suggest modifying the...
> > > > - "startdate" to be the model initialization time to be
processed.
> > > > - Leave "fhr" set to 00 to start at the analysis time.
> > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
hours/day).
> > > > - Set "incrementhr" to 06 to process output every 6 hours.
> > > >
> > > > Give that a shot and let me know if that gets you output past
the
> first
> > > > day.
> > > >
> > > > Are you going to be evaluating precipitation? I ask that
because
> > precip
> > > is
> > > > an accumulated interval, meaning that it has a beginning
accumulation
> > > time
> > > > and ending accumulation time. GRIB1 is limited in the amount
of
> space
> > > > available for storing times. When you get past 255 forecast
hours,
> you
> > > can
> > > > no longer store beginning/ending times for precip because it
runs out
> > of
> > > > room in the header. Do you have a 10 year simulation computed
from a
> > > > single initialization? Meaning, do you have forecast lead
times out
> to
> > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
wondering if
> > we'll
> > > > run into problems formatting those time strings.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > >
> > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
> > met_help at ucar.edu
> > > <javascript:;>>
> > > > wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > > > >
> > > > > Dear John,
> > > > >
> > > > > I ran UPP with my multiple wrfout files, first with the
first month
> > of
> > > > the
> > > > > 10 year simulation. However, after executing the run_unipost
> script,
> > > UPP
> > > > > produced only four files: unipost_d01.00.out,
unipost_d01.06.out,
> > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems that
UPP
> > processed
> > > > > only the first day's output among the one month data.
> > > > > If you are familiar with UPP, how can I make sure for UPP to
> process
> > > > > multiple wrfout files more than one day?
> > > > > Thank you.
> > > > >
> > > > > Regards,
> > > > >
> > > > > Jinwoong Yoo
> > > > > UNM
> > > > >
> > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> > jinwoong.yoo at gmail.com
> > > <javascript:;>>
> > > > > wrote:
> > > > >
> > > > > > Dear John,
> > > > > >
> > > > > > By the way, what about the CCSM4 file in NetCDF format?
> > > > > > Do I need to post process of the CCSM4 files also to run
MET
> along
> > > with
> > > > > > the WRF-ARW files?
> > > > > > Thank you.
> > > > > >
> > > > > > Regards,
> > > > > >
> > > > > > Jinwoong Yoo
> > > > > > UNM
> > > > > >
> > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> > > jinwoong.yoo at gmail.com <javascript:;>>
> > > > > > wrote:
> > > > > >
> > > > > >> Dear John,
> > > > > >>
> > > > > >> Thank you for your kind email with the direction.
> > > > > >> Let me first post process some of my WRF-ARW outputs
using UPP
> and
> > > let
> > > > > >> you know.
> > > > > >> Thank you very much.
> > > > > >>
> > > > > >> Regards,
> > > > > >>
> > > > > >> Jinwoong Yoo
> > > > > >> UNM
> > > > > >>
> > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via
RT <
> > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > > > >>
> > > > > >>> Jinwoong,
> > > > > >>>
> > > > > >>> I see that you'd like some help running the series-
analysis
> tool
> > on
> > > > > your
> > > > > >>> wrf-arw climate runs. One thing to note is that MET
can't read
> > the
> > > > raw
> > > > > >>> wrfout files directly. Instead you first need to post
process
> > them
> > > > > using
> > > > > >>> either UPP (produces GRIB output) or the p_interp
utility
> > (produced
> > > > > >>> NetCDF
> > > > > >>> output). After you've post processed some files, please
send
> us
> > > some
> > > > > >>> sample data by posting it to our anonymous ftp site.
Also
> send a
> > > > > >>> description of the comparisons you'd like to make
between the 2
> > > > > datasets.
> > > > > >>>
> > > > > >>> Here's instructions for sending us data:
> > > > > >>> http://www.dtcenter.org/met/users/support/met_help.php
> > > > > >>>
> > > > > >>> Thanks
> > > > > >>> John Halley Gotway
> > > > > >>> met_help at ucar.edu <javascript:;>
> > > > > >>>
> > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT <
> > > > > met_help at ucar.edu <javascript:;>
> > > > > >>> >
> > > > > >>> wrote:
> > > > > >>>
> > > > > >>> >
> > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted
upon.
> > > > > >>> > Transaction: Ticket created by jinwoong.yoo at gmail.com
> > > <javascript:;>
> > > > <javascript:;>
> > > > > >>> > Queue: met_help
> > > > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4
outputs
> > > > > >>> > Owner: Nobody
> > > > > >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> > > <javascript:;>
> > > > > >>> > Status: new
> > > > > >>> > Ticket <URL:
> > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > > > >>> >
> > > > > >>> >
> > > > > >>> >
> > > > > >>> > Dear Met Help,
> > > > > >>> >
> > > > > >>> > Hi.
> > > > > >>> > I am looking for a way to verify WRF model performance
as a
> > > climate
> > > > > >>> model.
> > > > > >>> > I'm completed 10-year tropical channel climate model
> > simulations
> > > > > using
> > > > > >>> the
> > > > > >>> > CCSM4 data as input data (dynamical downscaling) for
two
> > > different
> > > > > time
> > > > > >>> > periods: one for the LGM and the other for the modern
as a
> > > control
> > > > > run.
> > > > > >>> > The WRF simulation produced 6 hourly output for ten
years
> > periods
> > > > > each.
> > > > > >>> > Also the CCSM4 simulations come with 6 hourly outputs.
> > > > > >>> >
> > > > > >>> > Basically, I'd like to compare the two simulation
results,
> > > > verifying
> > > > > >>> the
> > > > > >>> > WRF model simulations.
> > > > > >>> > (WRF ARW was implemented for the experiments.)
> > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
similar jobs
> > but
> > > I
> > > > > need
> > > > > >>> > much more help from you than the MET Online Tutorial
to make
> a
> > > step
> > > > > >>> > forward.
> > > > > >>> > Any help will be highly appreciated.
> > > > > >>> > Thank you.
> > > > > >>> >
> > > > > >>> > Regards,
> > > > > >>> >
> > > > > >>> > Jinwoong Yoo
> > > > > >>> > Postdoc
> > > > > >>> > Department of Earth and Planetary Sciences
> > > > > >>> > UNM
> > > > > >>> >
> > > > > >>> >
> > > > > >>>
> > > > > >>>
> > > > > >>
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
--
Julie Prestopnik
National Center for Atmospheric Research
Research Applications Laboratory
Phone: 303.497.8399
Email: jpresto at ucar.edu
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Tue Sep 30 16:05:07 2014
Hi, Julie.
Thank you for heads-up.
I will wait for John to reply then.
Thank you.
Jinwoong Yoo
UNM
On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT
<met_help at ucar.edu>
wrote:
> Hello. John is out of the office, but will be back later this week.
I am
> unable to help you with this task. Hopefully, John will be able to
respond
> to this ticket later this week.
>
> Julie
>
> On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> > Dear John,
> >
> > I ran UPP successfully on the UCAR Yellowstone.
> > I assume you can access to my glade scratch directory.
> > My UPP ouputs are located
> > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> >
> > Against those files, I'd like to compare CCSM4 simulation outputs
some of
> > which are located also under
> > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> >
> > Those are six hourly files all together. If you cannot access to
those
> > files, please let me know.
> >
> > However, there are some variables among the CCSM4 outputs that are
> > available only monthly or annually. In this case, how can I post-
process
> my
> > WRF outputs for the usage of MET through UPP?
> >
> > Thank you.
> >
> > Regards,
> >
> > Jinwoong Yoo
> > UNM
> >
> >
> >
> >
> > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Jinwoo,
> > >
> > > I believe UPP will still run, just using a single time value as
if it
> > were
> > > an instantaneous field rather than an accumulation interval.
The
> default
> > > for ARW is to store precip as a runtime accumulation. For
forecasts
> > hours
> > > < 255, like 12, I suspect it'll be stored in GRIB as an
accumulation
> > from 0
> > > to 12. For hours > 255, like 300, I suspect it'll be stored as
an
> > > instantaneous 300 hour forecast. It's a subtle difference, but
> something
> > > to be aware of when you set up MET config files to verify the
output.
> > >
> > > Give UPP a try and let me know how it goes.
> > >
> > > John
> > >
> > > On Friday, September 26, 2014, Jinwoong Yoo via RT
<met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
> > > >
> > > > Dear John,
> > > >
> > > > I see. Let me change "export lastfhr=18" to "export
lastfhr=744".
> > > > Thank you for your comments.
> > > > Yes, I'd like to evaluate precipitation, too, and my 10 year
> simulation
> > > was
> > > > integrated from a single initialization. Then, will the UPP
> run_unipost
> > > > crash after the 255 forecast hours or just will precipitation
not be
> > > stored
> > > > any more aster the 255 forecast hours?
> > > >
> > > > Thank you.
> > > >
> > > > Regards,
> > > >
> > > > Jinwoong Yoo
> > > > UNM
> > > >
> > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu <javascript:;>> wrote:
> > > >
> > > > > Jinwoong,
> > > > >
> > > > > UPP support is typically provided through wrf_help at ucar.edu
> > > > <javascript:;>. I'll give it
> > > > > a shot, but if this doesn't solve it, I'll refer you to
wrf_help.
> > > > >
> > > > > I assume you're running the script named "run_unipost".
Looking in
> > > > there,
> > > > > you'll see:
> > > > >
> > > > > export startdate=2005012300
> > > > >
> > > > > export fhr=00
> > > > > export lastfhr=18
> > > > > export incrementhr=03
> > > > >
> > > > > I'd suggest modifying the...
> > > > > - "startdate" to be the model initialization time to be
processed.
> > > > > - Leave "fhr" set to 00 to start at the analysis time.
> > > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
hours/day).
> > > > > - Set "incrementhr" to 06 to process output every 6 hours.
> > > > >
> > > > > Give that a shot and let me know if that gets you output
past the
> > first
> > > > > day.
> > > > >
> > > > > Are you going to be evaluating precipitation? I ask that
because
> > > precip
> > > > is
> > > > > an accumulated interval, meaning that it has a beginning
> accumulation
> > > > time
> > > > > and ending accumulation time. GRIB1 is limited in the
amount of
> > space
> > > > > available for storing times. When you get past 255 forecast
hours,
> > you
> > > > can
> > > > > no longer store beginning/ending times for precip because it
runs
> out
> > > of
> > > > > room in the header. Do you have a 10 year simulation
computed
> from a
> > > > > single initialization? Meaning, do you have forecast lead
times
> out
> > to
> > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
wondering if
> > > we'll
> > > > > run into problems formatting those time strings.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > >
> > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
> > > met_help at ucar.edu
> > > > <javascript:;>>
> > > > > wrote:
> > > > >
> > > > > >
> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > > > > >
> > > > > > Dear John,
> > > > > >
> > > > > > I ran UPP with my multiple wrfout files, first with the
first
> month
> > > of
> > > > > the
> > > > > > 10 year simulation. However, after executing the
run_unipost
> > script,
> > > > UPP
> > > > > > produced only four files: unipost_d01.00.out,
unipost_d01.06.out,
> > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems that
UPP
> > > processed
> > > > > > only the first day's output among the one month data.
> > > > > > If you are familiar with UPP, how can I make sure for UPP
to
> > process
> > > > > > multiple wrfout files more than one day?
> > > > > > Thank you.
> > > > > >
> > > > > > Regards,
> > > > > >
> > > > > > Jinwoong Yoo
> > > > > > UNM
> > > > > >
> > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> > > jinwoong.yoo at gmail.com
> > > > <javascript:;>>
> > > > > > wrote:
> > > > > >
> > > > > > > Dear John,
> > > > > > >
> > > > > > > By the way, what about the CCSM4 file in NetCDF format?
> > > > > > > Do I need to post process of the CCSM4 files also to run
MET
> > along
> > > > with
> > > > > > > the WRF-ARW files?
> > > > > > > Thank you.
> > > > > > >
> > > > > > > Regards,
> > > > > > >
> > > > > > > Jinwoong Yoo
> > > > > > > UNM
> > > > > > >
> > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> > > > jinwoong.yoo at gmail.com <javascript:;>>
> > > > > > > wrote:
> > > > > > >
> > > > > > >> Dear John,
> > > > > > >>
> > > > > > >> Thank you for your kind email with the direction.
> > > > > > >> Let me first post process some of my WRF-ARW outputs
using UPP
> > and
> > > > let
> > > > > > >> you know.
> > > > > > >> Thank you very much.
> > > > > > >>
> > > > > > >> Regards,
> > > > > > >>
> > > > > > >> Jinwoong Yoo
> > > > > > >> UNM
> > > > > > >>
> > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway via
RT <
> > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > > > > >>
> > > > > > >>> Jinwoong,
> > > > > > >>>
> > > > > > >>> I see that you'd like some help running the series-
analysis
> > tool
> > > on
> > > > > > your
> > > > > > >>> wrf-arw climate runs. One thing to note is that MET
can't
> read
> > > the
> > > > > raw
> > > > > > >>> wrfout files directly. Instead you first need to post
> process
> > > them
> > > > > > using
> > > > > > >>> either UPP (produces GRIB output) or the p_interp
utility
> > > (produced
> > > > > > >>> NetCDF
> > > > > > >>> output). After you've post processed some files,
please send
> > us
> > > > some
> > > > > > >>> sample data by posting it to our anonymous ftp site.
Also
> > send a
> > > > > > >>> description of the comparisons you'd like to make
between
> the 2
> > > > > > datasets.
> > > > > > >>>
> > > > > > >>> Here's instructions for sending us data:
> > > > > > >>> http://www.dtcenter.org/met/users/support/met_help.php
> > > > > > >>>
> > > > > > >>> Thanks
> > > > > > >>> John Halley Gotway
> > > > > > >>> met_help at ucar.edu <javascript:;>
> > > > > > >>>
> > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via RT
<
> > > > > > met_help at ucar.edu <javascript:;>
> > > > > > >>> >
> > > > > > >>> wrote:
> > > > > > >>>
> > > > > > >>> >
> > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted
upon.
> > > > > > >>> > Transaction: Ticket created by
jinwoong.yoo at gmail.com
> > > > <javascript:;>
> > > > > <javascript:;>
> > > > > > >>> > Queue: met_help
> > > > > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4
outputs
> > > > > > >>> > Owner: Nobody
> > > > > > >>> > Requestors: jinwoong.yoo at gmail.com <javascript:;>
> > > > <javascript:;>
> > > > > > >>> > Status: new
> > > > > > >>> > Ticket <URL:
> > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > > > > >>> >
> > > > > > >>> >
> > > > > > >>> >
> > > > > > >>> > Dear Met Help,
> > > > > > >>> >
> > > > > > >>> > Hi.
> > > > > > >>> > I am looking for a way to verify WRF model
performance as a
> > > > climate
> > > > > > >>> model.
> > > > > > >>> > I'm completed 10-year tropical channel climate model
> > > simulations
> > > > > > using
> > > > > > >>> the
> > > > > > >>> > CCSM4 data as input data (dynamical downscaling) for
two
> > > > different
> > > > > > time
> > > > > > >>> > periods: one for the LGM and the other for the
modern as a
> > > > control
> > > > > > run.
> > > > > > >>> > The WRF simulation produced 6 hourly output for ten
years
> > > periods
> > > > > > each.
> > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
outputs.
> > > > > > >>> >
> > > > > > >>> > Basically, I'd like to compare the two simulation
results,
> > > > > verifying
> > > > > > >>> the
> > > > > > >>> > WRF model simulations.
> > > > > > >>> > (WRF ARW was implemented for the experiments.)
> > > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
similar
> jobs
> > > but
> > > > I
> > > > > > need
> > > > > > >>> > much more help from you than the MET Online Tutorial
to
> make
> > a
> > > > step
> > > > > > >>> > forward.
> > > > > > >>> > Any help will be highly appreciated.
> > > > > > >>> > Thank you.
> > > > > > >>> >
> > > > > > >>> > Regards,
> > > > > > >>> >
> > > > > > >>> > Jinwoong Yoo
> > > > > > >>> > Postdoc
> > > > > > >>> > Department of Earth and Planetary Sciences
> > > > > > >>> > UNM
> > > > > > >>> >
> > > > > > >>> >
> > > > > > >>>
> > > > > > >>>
> > > > > > >>
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
> --
> Julie Prestopnik
> National Center for Atmospheric Research
> Research Applications Laboratory
> Phone: 303.497.8399
> Email: jpresto at ucar.edu
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Thu Oct 02 12:02:47 2014
Jinwoong,
I'm able to access your data on Yellowstone. I see that you'd like to
compare your UPP output in GRIB format to some data that follows the
NetCDF
CF-1.0 convention.
UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
CCSM4 output: /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
Hopefully you're able to look at those files and identify which GRIB
records you'd like to compare to which NetCDF variables. When you run
the
MET tools, you'll define the forecast and observation fields to be
used.
Looking at your data, I see 2 immediate problems:
(1) Gridded forecast and observation data must be on the same grid
before
MET can compare them.
Here's a description from wgrib of the grid for your GRIB data:
Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
scan 64 mode 136 Latin 0.000000 bdsgrid 1
But it looks like the CCSM4 data is on a lat/lon grid.
I'd suggest using the copygb utility to regrid the GRIB1 output from
UPP to
the lat/lon grid of the CCSM4 data. Here's an example of doing that:
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
(2) There's an issue in the times in the accumulated precipitation.
Try
running the following 2 commands:
wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
grep APCP
wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
grep APCP
For the first, APCP is an accumulation of precip from hours 0 to 252.
The
second should be 0 to 258, but it's stored as 0 to 2! That's because
UPP
is failing to realize that the precipitation timestamps are
overflowing the
1 byte allocated for that second time unit. This problem was
addressed in
an earlier version of UPP but the problem is still showing up in the
version you're using. If you plan to verify precip, I'd suggest
writing
wrfhelp at ucar.edu, tell them the version of UPP you're using, and ask
them
if this should be fixed.
Hope that helps.
Thanks,
John
On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Hi, Julie.
> Thank you for heads-up.
> I will wait for John to reply then.
> Thank you.
>
> Jinwoong Yoo
> UNM
>
> On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
> met_help at ucar.edu>
> wrote:
>
> > Hello. John is out of the office, but will be back later this
week. I am
> > unable to help you with this task. Hopefully, John will be able
to
> respond
> > to this ticket later this week.
> >
> > Julie
> >
> > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > >
> > > Dear John,
> > >
> > > I ran UPP successfully on the UCAR Yellowstone.
> > > I assume you can access to my glade scratch directory.
> > > My UPP ouputs are located
> > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> > >
> > > Against those files, I'd like to compare CCSM4 simulation
outputs some
> of
> > > which are located also under
> > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> > >
> > > Those are six hourly files all together. If you cannot access to
those
> > > files, please let me know.
> > >
> > > However, there are some variables among the CCSM4 outputs that
are
> > > available only monthly or annually. In this case, how can I
> post-process
> > my
> > > WRF outputs for the usage of MET through UPP?
> > >
> > > Thank you.
> > >
> > > Regards,
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > >
> > >
> > >
> > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Jinwoo,
> > > >
> > > > I believe UPP will still run, just using a single time value
as if it
> > > were
> > > > an instantaneous field rather than an accumulation interval.
The
> > default
> > > > for ARW is to store precip as a runtime accumulation. For
forecasts
> > > hours
> > > > < 255, like 12, I suspect it'll be stored in GRIB as an
accumulation
> > > from 0
> > > > to 12. For hours > 255, like 300, I suspect it'll be stored
as an
> > > > instantaneous 300 hour forecast. It's a subtle difference,
but
> > something
> > > > to be aware of when you set up MET config files to verify the
output.
> > > >
> > > > Give UPP a try and let me know how it goes.
> > > >
> > > > John
> > > >
> > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
> met_help at ucar.edu>
> > > > wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > > > >
> > > > > Dear John,
> > > > >
> > > > > I see. Let me change "export lastfhr=18" to "export
lastfhr=744".
> > > > > Thank you for your comments.
> > > > > Yes, I'd like to evaluate precipitation, too, and my 10 year
> > simulation
> > > > was
> > > > > integrated from a single initialization. Then, will the UPP
> > run_unipost
> > > > > crash after the 255 forecast hours or just will
precipitation not
> be
> > > > stored
> > > > > any more aster the 255 forecast hours?
> > > > >
> > > > > Thank you.
> > > > >
> > > > > Regards,
> > > > >
> > > > > Jinwoong Yoo
> > > > > UNM
> > > > >
> > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT <
> > > > > met_help at ucar.edu <javascript:;>> wrote:
> > > > >
> > > > > > Jinwoong,
> > > > > >
> > > > > > UPP support is typically provided through
wrf_help at ucar.edu
> > > > > <javascript:;>. I'll give it
> > > > > > a shot, but if this doesn't solve it, I'll refer you to
wrf_help.
> > > > > >
> > > > > > I assume you're running the script named "run_unipost".
Looking
> in
> > > > > there,
> > > > > > you'll see:
> > > > > >
> > > > > > export startdate=2005012300
> > > > > >
> > > > > > export fhr=00
> > > > > > export lastfhr=18
> > > > > > export incrementhr=03
> > > > > >
> > > > > > I'd suggest modifying the...
> > > > > > - "startdate" to be the model initialization time to be
> processed.
> > > > > > - Leave "fhr" set to 00 to start at the analysis time.
> > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
> hours/day).
> > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
> > > > > >
> > > > > > Give that a shot and let me know if that gets you output
past the
> > > first
> > > > > > day.
> > > > > >
> > > > > > Are you going to be evaluating precipitation? I ask that
because
> > > > precip
> > > > > is
> > > > > > an accumulated interval, meaning that it has a beginning
> > accumulation
> > > > > time
> > > > > > and ending accumulation time. GRIB1 is limited in the
amount of
> > > space
> > > > > > available for storing times. When you get past 255
forecast
> hours,
> > > you
> > > > > can
> > > > > > no longer store beginning/ending times for precip because
it runs
> > out
> > > > of
> > > > > > room in the header. Do you have a 10 year simulation
computed
> > from a
> > > > > > single initialization? Meaning, do you have forecast lead
times
> > out
> > > to
> > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
wondering
> if
> > > > we'll
> > > > > > run into problems formatting those time strings.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > >
> > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
> > > > met_help at ucar.edu
> > > > > <javascript:;>>
> > > > > > wrote:
> > > > > >
> > > > > > >
> > > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >
> > > > > > >
> > > > > > > Dear John,
> > > > > > >
> > > > > > > I ran UPP with my multiple wrfout files, first with the
first
> > month
> > > > of
> > > > > > the
> > > > > > > 10 year simulation. However, after executing the
run_unipost
> > > script,
> > > > > UPP
> > > > > > > produced only four files: unipost_d01.00.out,
> unipost_d01.06.out,
> > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems
that UPP
> > > > processed
> > > > > > > only the first day's output among the one month data.
> > > > > > > If you are familiar with UPP, how can I make sure for
UPP to
> > > process
> > > > > > > multiple wrfout files more than one day?
> > > > > > > Thank you.
> > > > > > >
> > > > > > > Regards,
> > > > > > >
> > > > > > > Jinwoong Yoo
> > > > > > > UNM
> > > > > > >
> > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> > > > jinwoong.yoo at gmail.com
> > > > > <javascript:;>>
> > > > > > > wrote:
> > > > > > >
> > > > > > > > Dear John,
> > > > > > > >
> > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
> > > > > > > > Do I need to post process of the CCSM4 files also to
run MET
> > > along
> > > > > with
> > > > > > > > the WRF-ARW files?
> > > > > > > > Thank you.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > >
> > > > > > > > Jinwoong Yoo
> > > > > > > > UNM
> > > > > > > >
> > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> > > > > jinwoong.yoo at gmail.com <javascript:;>>
> > > > > > > > wrote:
> > > > > > > >
> > > > > > > >> Dear John,
> > > > > > > >>
> > > > > > > >> Thank you for your kind email with the direction.
> > > > > > > >> Let me first post process some of my WRF-ARW outputs
using
> UPP
> > > and
> > > > > let
> > > > > > > >> you know.
> > > > > > > >> Thank you very much.
> > > > > > > >>
> > > > > > > >> Regards,
> > > > > > > >>
> > > > > > > >> Jinwoong Yoo
> > > > > > > >> UNM
> > > > > > > >>
> > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway
via RT <
> > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > > > > > >>
> > > > > > > >>> Jinwoong,
> > > > > > > >>>
> > > > > > > >>> I see that you'd like some help running the series-
analysis
> > > tool
> > > > on
> > > > > > > your
> > > > > > > >>> wrf-arw climate runs. One thing to note is that MET
can't
> > read
> > > > the
> > > > > > raw
> > > > > > > >>> wrfout files directly. Instead you first need to
post
> > process
> > > > them
> > > > > > > using
> > > > > > > >>> either UPP (produces GRIB output) or the p_interp
utility
> > > > (produced
> > > > > > > >>> NetCDF
> > > > > > > >>> output). After you've post processed some files,
please
> send
> > > us
> > > > > some
> > > > > > > >>> sample data by posting it to our anonymous ftp site.
Also
> > > send a
> > > > > > > >>> description of the comparisons you'd like to make
between
> > the 2
> > > > > > > datasets.
> > > > > > > >>>
> > > > > > > >>> Here's instructions for sending us data:
> > > > > > > >>>
http://www.dtcenter.org/met/users/support/met_help.php
> > > > > > > >>>
> > > > > > > >>> Thanks
> > > > > > > >>> John Halley Gotway
> > > > > > > >>> met_help at ucar.edu <javascript:;>
> > > > > > > >>>
> > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via
RT <
> > > > > > > met_help at ucar.edu <javascript:;>
> > > > > > > >>> >
> > > > > > > >>> wrote:
> > > > > > > >>>
> > > > > > > >>> >
> > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted
upon.
> > > > > > > >>> > Transaction: Ticket created by
jinwoong.yoo at gmail.com
> > > > > <javascript:;>
> > > > > > <javascript:;>
> > > > > > > >>> > Queue: met_help
> > > > > > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4
outputs
> > > > > > > >>> > Owner: Nobody
> > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
> > > > > <javascript:;>
> > > > > > > >>> > Status: new
> > > > > > > >>> > Ticket <URL:
> > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > > > > > >>> >
> > > > > > > >>> >
> > > > > > > >>> >
> > > > > > > >>> > Dear Met Help,
> > > > > > > >>> >
> > > > > > > >>> > Hi.
> > > > > > > >>> > I am looking for a way to verify WRF model
performance
> as a
> > > > > climate
> > > > > > > >>> model.
> > > > > > > >>> > I'm completed 10-year tropical channel climate
model
> > > > simulations
> > > > > > > using
> > > > > > > >>> the
> > > > > > > >>> > CCSM4 data as input data (dynamical downscaling)
for two
> > > > > different
> > > > > > > time
> > > > > > > >>> > periods: one for the LGM and the other for the
modern as
> a
> > > > > control
> > > > > > > run.
> > > > > > > >>> > The WRF simulation produced 6 hourly output for
ten years
> > > > periods
> > > > > > > each.
> > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
outputs.
> > > > > > > >>> >
> > > > > > > >>> > Basically, I'd like to compare the two simulation
> results,
> > > > > > verifying
> > > > > > > >>> the
> > > > > > > >>> > WRF model simulations.
> > > > > > > >>> > (WRF ARW was implemented for the experiments.)
> > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
similar
> > jobs
> > > > but
> > > > > I
> > > > > > > need
> > > > > > > >>> > much more help from you than the MET Online
Tutorial to
> > make
> > > a
> > > > > step
> > > > > > > >>> > forward.
> > > > > > > >>> > Any help will be highly appreciated.
> > > > > > > >>> > Thank you.
> > > > > > > >>> >
> > > > > > > >>> > Regards,
> > > > > > > >>> >
> > > > > > > >>> > Jinwoong Yoo
> > > > > > > >>> > Postdoc
> > > > > > > >>> > Department of Earth and Planetary Sciences
> > > > > > > >>> > UNM
> > > > > > > >>> >
> > > > > > > >>> >
> > > > > > > >>>
> > > > > > > >>>
> > > > > > > >>
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Julie Prestopnik
> > National Center for Atmospheric Research
> > Research Applications Laboratory
> > Phone: 303.497.8399
> > Email: jpresto at ucar.edu
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Thu Oct 02 12:13:45 2014
Jinwoong,
FYI - I just talked to the person who provides support for UPP through
wrfhelp, and she's expecting an email from you.
Thanks,
John
On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway <johnhg at ucar.edu>
wrote:
> Jinwoong,
>
> I'm able to access your data on Yellowstone. I see that you'd like
to
> compare your UPP output in GRIB format to some data that follows the
NetCDF
> CF-1.0 convention.
>
> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> CCSM4 output: /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>
> Hopefully you're able to look at those files and identify which GRIB
> records you'd like to compare to which NetCDF variables. When you
run the
> MET tools, you'll define the forecast and observation fields to be
used.
>
> Looking at your data, I see 2 immediate problems:
>
> (1) Gridded forecast and observation data must be on the same grid
before
> MET can compare them.
> Here's a description from wgrib of the grid for your GRIB data:
> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> But it looks like the CCSM4 data is on a lat/lon grid.
>
> I'd suggest using the copygb utility to regrid the GRIB1 output from
UPP
> to the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>
> (2) There's an issue in the times in the accumulated precipitation.
Try
> running the following 2 commands:
> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
> grep APCP
> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
> grep APCP
>
> For the first, APCP is an accumulation of precip from hours 0 to
252. The
> second should be 0 to 258, but it's stored as 0 to 2! That's
because UPP
> is failing to realize that the precipitation timestamps are
overflowing the
> 1 byte allocated for that second time unit. This problem was
addressed in
> an earlier version of UPP but the problem is still showing up in the
> version you're using. If you plan to verify precip, I'd suggest
writing
> wrfhelp at ucar.edu, tell them the version of UPP you're using, and ask
them
> if this should be fixed.
>
> Hope that helps.
>
> Thanks,
> John
>
>
>
>
>
>
> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>
>> Hi, Julie.
>> Thank you for heads-up.
>> I will wait for John to reply then.
>> Thank you.
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>> met_help at ucar.edu>
>> wrote:
>>
>> > Hello. John is out of the office, but will be back later this
week. I
>> am
>> > unable to help you with this task. Hopefully, John will be able
to
>> respond
>> > to this ticket later this week.
>> >
>> > Julie
>> >
>> > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT
<met_help at ucar.edu
>> >
>> > wrote:
>> >
>> > >
>> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>> > >
>> > > Dear John,
>> > >
>> > > I ran UPP successfully on the UCAR Yellowstone.
>> > > I assume you can access to my glade scratch directory.
>> > > My UPP ouputs are located
>> > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>> > >
>> > > Against those files, I'd like to compare CCSM4 simulation
outputs
>> some of
>> > > which are located also under
>> > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>> > >
>> > > Those are six hourly files all together. If you cannot access
to those
>> > > files, please let me know.
>> > >
>> > > However, there are some variables among the CCSM4 outputs that
are
>> > > available only monthly or annually. In this case, how can I
>> post-process
>> > my
>> > > WRF outputs for the usage of MET through UPP?
>> > >
>> > > Thank you.
>> > >
>> > > Regards,
>> > >
>> > > Jinwoong Yoo
>> > > UNM
>> > >
>> > >
>> > >
>> > >
>> > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
>> > > met_help at ucar.edu> wrote:
>> > >
>> > > > Jinwoo,
>> > > >
>> > > > I believe UPP will still run, just using a single time value
as if
>> it
>> > > were
>> > > > an instantaneous field rather than an accumulation interval.
The
>> > default
>> > > > for ARW is to store precip as a runtime accumulation. For
forecasts
>> > > hours
>> > > > < 255, like 12, I suspect it'll be stored in GRIB as an
accumulation
>> > > from 0
>> > > > to 12. For hours > 255, like 300, I suspect it'll be stored
as an
>> > > > instantaneous 300 hour forecast. It's a subtle difference,
but
>> > something
>> > > > to be aware of when you set up MET config files to verify the
>> output.
>> > > >
>> > > > Give UPP a try and let me know how it goes.
>> > > >
>> > > > John
>> > > >
>> > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>> met_help at ucar.edu>
>> > > > wrote:
>> > > >
>> > > > >
>> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>> > > > >
>> > > > > Dear John,
>> > > > >
>> > > > > I see. Let me change "export lastfhr=18" to "export
lastfhr=744".
>> > > > > Thank you for your comments.
>> > > > > Yes, I'd like to evaluate precipitation, too, and my 10
year
>> > simulation
>> > > > was
>> > > > > integrated from a single initialization. Then, will the UPP
>> > run_unipost
>> > > > > crash after the 255 forecast hours or just will
precipitation not
>> be
>> > > > stored
>> > > > > any more aster the 255 forecast hours?
>> > > > >
>> > > > > Thank you.
>> > > > >
>> > > > > Regards,
>> > > > >
>> > > > > Jinwoong Yoo
>> > > > > UNM
>> > > > >
>> > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT
<
>> > > > > met_help at ucar.edu <javascript:;>> wrote:
>> > > > >
>> > > > > > Jinwoong,
>> > > > > >
>> > > > > > UPP support is typically provided through
wrf_help at ucar.edu
>> > > > > <javascript:;>. I'll give it
>> > > > > > a shot, but if this doesn't solve it, I'll refer you to
>> wrf_help.
>> > > > > >
>> > > > > > I assume you're running the script named "run_unipost".
>> Looking in
>> > > > > there,
>> > > > > > you'll see:
>> > > > > >
>> > > > > > export startdate=2005012300
>> > > > > >
>> > > > > > export fhr=00
>> > > > > > export lastfhr=18
>> > > > > > export incrementhr=03
>> > > > > >
>> > > > > > I'd suggest modifying the...
>> > > > > > - "startdate" to be the model initialization time to be
>> processed.
>> > > > > > - Leave "fhr" set to 00 to start at the analysis time.
>> > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
>> hours/day).
>> > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
>> > > > > >
>> > > > > > Give that a shot and let me know if that gets you output
past
>> the
>> > > first
>> > > > > > day.
>> > > > > >
>> > > > > > Are you going to be evaluating precipitation? I ask that
>> because
>> > > > precip
>> > > > > is
>> > > > > > an accumulated interval, meaning that it has a beginning
>> > accumulation
>> > > > > time
>> > > > > > and ending accumulation time. GRIB1 is limited in the
amount of
>> > > space
>> > > > > > available for storing times. When you get past 255
forecast
>> hours,
>> > > you
>> > > > > can
>> > > > > > no longer store beginning/ending times for precip because
it
>> runs
>> > out
>> > > > of
>> > > > > > room in the header. Do you have a 10 year simulation
computed
>> > from a
>> > > > > > single initialization? Meaning, do you have forecast
lead times
>> > out
>> > > to
>> > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
wondering
>> if
>> > > > we'll
>> > > > > > run into problems formatting those time strings.
>> > > > > >
>> > > > > > Thanks,
>> > > > > > John
>> > > > > >
>> > > > > >
>> > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
>> > > > met_help at ucar.edu
>> > > > > <javascript:;>>
>> > > > > > wrote:
>> > > > > >
>> > > > > > >
>> > > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>> >
>> > > > > > >
>> > > > > > > Dear John,
>> > > > > > >
>> > > > > > > I ran UPP with my multiple wrfout files, first with the
first
>> > month
>> > > > of
>> > > > > > the
>> > > > > > > 10 year simulation. However, after executing the
run_unipost
>> > > script,
>> > > > > UPP
>> > > > > > > produced only four files: unipost_d01.00.out,
>> unipost_d01.06.out,
>> > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems
that UPP
>> > > > processed
>> > > > > > > only the first day's output among the one month data.
>> > > > > > > If you are familiar with UPP, how can I make sure for
UPP to
>> > > process
>> > > > > > > multiple wrfout files more than one day?
>> > > > > > > Thank you.
>> > > > > > >
>> > > > > > > Regards,
>> > > > > > >
>> > > > > > > Jinwoong Yoo
>> > > > > > > UNM
>> > > > > > >
>> > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>> > > > jinwoong.yoo at gmail.com
>> > > > > <javascript:;>>
>> > > > > > > wrote:
>> > > > > > >
>> > > > > > > > Dear John,
>> > > > > > > >
>> > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
>> > > > > > > > Do I need to post process of the CCSM4 files also to
run MET
>> > > along
>> > > > > with
>> > > > > > > > the WRF-ARW files?
>> > > > > > > > Thank you.
>> > > > > > > >
>> > > > > > > > Regards,
>> > > > > > > >
>> > > > > > > > Jinwoong Yoo
>> > > > > > > > UNM
>> > > > > > > >
>> > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>> > > > > jinwoong.yoo at gmail.com <javascript:;>>
>> > > > > > > > wrote:
>> > > > > > > >
>> > > > > > > >> Dear John,
>> > > > > > > >>
>> > > > > > > >> Thank you for your kind email with the direction.
>> > > > > > > >> Let me first post process some of my WRF-ARW outputs
using
>> UPP
>> > > and
>> > > > > let
>> > > > > > > >> you know.
>> > > > > > > >> Thank you very much.
>> > > > > > > >>
>> > > > > > > >> Regards,
>> > > > > > > >>
>> > > > > > > >> Jinwoong Yoo
>> > > > > > > >> UNM
>> > > > > > > >>
>> > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway
via RT
>> <
>> > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>> > > > > > > >>
>> > > > > > > >>> Jinwoong,
>> > > > > > > >>>
>> > > > > > > >>> I see that you'd like some help running the
>> series-analysis
>> > > tool
>> > > > on
>> > > > > > > your
>> > > > > > > >>> wrf-arw climate runs. One thing to note is that
MET can't
>> > read
>> > > > the
>> > > > > > raw
>> > > > > > > >>> wrfout files directly. Instead you first need to
post
>> > process
>> > > > them
>> > > > > > > using
>> > > > > > > >>> either UPP (produces GRIB output) or the p_interp
utility
>> > > > (produced
>> > > > > > > >>> NetCDF
>> > > > > > > >>> output). After you've post processed some files,
please
>> send
>> > > us
>> > > > > some
>> > > > > > > >>> sample data by posting it to our anonymous ftp
site. Also
>> > > send a
>> > > > > > > >>> description of the comparisons you'd like to make
between
>> > the 2
>> > > > > > > datasets.
>> > > > > > > >>>
>> > > > > > > >>> Here's instructions for sending us data:
>> > > > > > > >>>
http://www.dtcenter.org/met/users/support/met_help.php
>> > > > > > > >>>
>> > > > > > > >>> Thanks
>> > > > > > > >>> John Halley Gotway
>> > > > > > > >>> met_help at ucar.edu <javascript:;>
>> > > > > > > >>>
>> > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via
RT <
>> > > > > > > met_help at ucar.edu <javascript:;>
>> > > > > > > >>> >
>> > > > > > > >>> wrote:
>> > > > > > > >>>
>> > > > > > > >>> >
>> > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was acted
upon.
>> > > > > > > >>> > Transaction: Ticket created by
jinwoong.yoo at gmail.com
>> > > > > <javascript:;>
>> > > > > > <javascript:;>
>> > > > > > > >>> > Queue: met_help
>> > > > > > > >>> > Subject: Comparing WRF-ARW outputs and CCSM4
>> outputs
>> > > > > > > >>> > Owner: Nobody
>> > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
>> > > > > <javascript:;>
>> > > > > > > >>> > Status: new
>> > > > > > > >>> > Ticket <URL:
>> > > > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>> > > > > > > >>> >
>> > > > > > > >>> >
>> > > > > > > >>> >
>> > > > > > > >>> > Dear Met Help,
>> > > > > > > >>> >
>> > > > > > > >>> > Hi.
>> > > > > > > >>> > I am looking for a way to verify WRF model
performance
>> as a
>> > > > > climate
>> > > > > > > >>> model.
>> > > > > > > >>> > I'm completed 10-year tropical channel climate
model
>> > > > simulations
>> > > > > > > using
>> > > > > > > >>> the
>> > > > > > > >>> > CCSM4 data as input data (dynamical downscaling)
for two
>> > > > > different
>> > > > > > > time
>> > > > > > > >>> > periods: one for the LGM and the other for the
modern
>> as a
>> > > > > control
>> > > > > > > run.
>> > > > > > > >>> > The WRF simulation produced 6 hourly output for
ten
>> years
>> > > > periods
>> > > > > > > each.
>> > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
outputs.
>> > > > > > > >>> >
>> > > > > > > >>> > Basically, I'd like to compare the two simulation
>> results,
>> > > > > > verifying
>> > > > > > > >>> the
>> > > > > > > >>> > WRF model simulations.
>> > > > > > > >>> > (WRF ARW was implemented for the experiments.)
>> > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
similar
>> > jobs
>> > > > but
>> > > > > I
>> > > > > > > need
>> > > > > > > >>> > much more help from you than the MET Online
Tutorial to
>> > make
>> > > a
>> > > > > step
>> > > > > > > >>> > forward.
>> > > > > > > >>> > Any help will be highly appreciated.
>> > > > > > > >>> > Thank you.
>> > > > > > > >>> >
>> > > > > > > >>> > Regards,
>> > > > > > > >>> >
>> > > > > > > >>> > Jinwoong Yoo
>> > > > > > > >>> > Postdoc
>> > > > > > > >>> > Department of Earth and Planetary Sciences
>> > > > > > > >>> > UNM
>> > > > > > > >>> >
>> > > > > > > >>> >
>> > > > > > > >>>
>> > > > > > > >>>
>> > > > > > > >>
>> > > > > > > >
>> > > > > > >
>> > > > > > >
>> > > > > >
>> > > > > >
>> > > > >
>> > > > >
>> > > >
>> > > >
>> > >
>> > >
>> >
>> >
>> > --
>> > Julie Prestopnik
>> > National Center for Atmospheric Research
>> > Research Applications Laboratory
>> > Phone: 303.497.8399
>> > Email: jpresto at ucar.edu
>> >
>> >
>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Oct 02 14:18:56 2014
Dear John,
Actually I was looking over copygb when I found your email received.
It seems that I have to convert the UPP output again using copygb.exe
to
match up the grid of WRF results to the CCSM4's lat/lon grid.
Isn't there any way to do that in one step? No?
Along with the precipitation, I'd like to compare variables (paired
for WRF
and CCSM4) below for example:
Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
Varwrfclm = (/"T2", "TSLB","SMOIS",
"SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
"FSDS","SNOWICE","Q2M","RH2M"/)
Variables in the lists are the actual name of variables in the CCSM4
(including CAM and CLM monthly files) files and corresponding to those
in
the WRF-ARW.
I'm trying to run Grid-Stat for a test and I'd like to run Series-
Analysis
for the entire dataset of the WRF and the CCSM4.
However, I'm concerning about how to run the Series-Analysis for the
monthly or annual data using MET?
At the same time, I am also thinking about using the NCL to do the
same
statistical analyses. I don't know which one will be easier for me to
execute for now.
We'll see.
Will let you know as I make progress, John.
>>> FYI - I just talked to the person who provides support for UPP
through
>>> wrfhelp, and she's expecting an email from you.
I cc'ed this email to wrfhelp.
Thank you.
Regards,
Jinwoong Yoo
UNM
On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Jinwoong,
>
> I'm able to access your data on Yellowstone. I see that you'd like
to
> compare your UPP output in GRIB format to some data that follows the
NetCDF
> CF-1.0 convention.
>
> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> CCSM4 output: /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>
> Hopefully you're able to look at those files and identify which GRIB
> records you'd like to compare to which NetCDF variables. When you
run the
> MET tools, you'll define the forecast and observation fields to be
used.
>
> Looking at your data, I see 2 immediate problems:
>
> (1) Gridded forecast and observation data must be on the same grid
before
> MET can compare them.
> Here's a description from wgrib of the grid for your GRIB data:
> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> But it looks like the CCSM4 data is on a lat/lon grid.
>
> I'd suggest using the copygb utility to regrid the GRIB1 output from
UPP to
> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>
>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>
> (2) There's an issue in the times in the accumulated precipitation.
Try
> running the following 2 commands:
> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
> grep APCP
> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
> grep APCP
>
> For the first, APCP is an accumulation of precip from hours 0 to
252. The
> second should be 0 to 258, but it's stored as 0 to 2! That's
because UPP
> is failing to realize that the precipitation timestamps are
overflowing the
> 1 byte allocated for that second time unit. This problem was
addressed in
> an earlier version of UPP but the problem is still showing up in the
> version you're using. If you plan to verify precip, I'd suggest
writing
> wrfhelp at ucar.edu, tell them the version of UPP you're using, and ask
them
> if this should be fixed.
>
> Hope that helps.
>
> Thanks,
> John
>
>
>
>
>
>
> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> > Hi, Julie.
> > Thank you for heads-up.
> > I will wait for John to reply then.
> > Thank you.
> >
> > Jinwoong Yoo
> > UNM
> >
> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > > Hello. John is out of the office, but will be back later this
week. I
> am
> > > unable to help you with this task. Hopefully, John will be able
to
> > respond
> > > to this ticket later this week.
> > >
> > > Julie
> > >
> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
> met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
> > > >
> > > > Dear John,
> > > >
> > > > I ran UPP successfully on the UCAR Yellowstone.
> > > > I assume you can access to my glade scratch directory.
> > > > My UPP ouputs are located
> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> > > >
> > > > Against those files, I'd like to compare CCSM4 simulation
outputs
> some
> > of
> > > > which are located also under
> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> > > >
> > > > Those are six hourly files all together. If you cannot access
to
> those
> > > > files, please let me know.
> > > >
> > > > However, there are some variables among the CCSM4 outputs that
are
> > > > available only monthly or annually. In this case, how can I
> > post-process
> > > my
> > > > WRF outputs for the usage of MET through UPP?
> > > >
> > > > Thank you.
> > > >
> > > > Regards,
> > > >
> > > > Jinwoong Yoo
> > > > UNM
> > > >
> > > >
> > > >
> > > >
> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > > Jinwoo,
> > > > >
> > > > > I believe UPP will still run, just using a single time value
as if
> it
> > > > were
> > > > > an instantaneous field rather than an accumulation interval.
The
> > > default
> > > > > for ARW is to store precip as a runtime accumulation. For
> forecasts
> > > > hours
> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
> accumulation
> > > > from 0
> > > > > to 12. For hours > 255, like 300, I suspect it'll be stored
as an
> > > > > instantaneous 300 hour forecast. It's a subtle difference,
but
> > > something
> > > > > to be aware of when you set up MET config files to verify
the
> output.
> > > > >
> > > > > Give UPP a try and let me know how it goes.
> > > > >
> > > > > John
> > > > >
> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
> > met_help at ucar.edu>
> > > > > wrote:
> > > > >
> > > > > >
> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > > > > >
> > > > > > Dear John,
> > > > > >
> > > > > > I see. Let me change "export lastfhr=18" to "export
lastfhr=744".
> > > > > > Thank you for your comments.
> > > > > > Yes, I'd like to evaluate precipitation, too, and my 10
year
> > > simulation
> > > > > was
> > > > > > integrated from a single initialization. Then, will the
UPP
> > > run_unipost
> > > > > > crash after the 255 forecast hours or just will
precipitation not
> > be
> > > > > stored
> > > > > > any more aster the 255 forecast hours?
> > > > > >
> > > > > > Thank you.
> > > > > >
> > > > > > Regards,
> > > > > >
> > > > > > Jinwoong Yoo
> > > > > > UNM
> > > > > >
> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via RT
<
> > > > > > met_help at ucar.edu <javascript:;>> wrote:
> > > > > >
> > > > > > > Jinwoong,
> > > > > > >
> > > > > > > UPP support is typically provided through
wrf_help at ucar.edu
> > > > > > <javascript:;>. I'll give it
> > > > > > > a shot, but if this doesn't solve it, I'll refer you to
> wrf_help.
> > > > > > >
> > > > > > > I assume you're running the script named "run_unipost".
> Looking
> > in
> > > > > > there,
> > > > > > > you'll see:
> > > > > > >
> > > > > > > export startdate=2005012300
> > > > > > >
> > > > > > > export fhr=00
> > > > > > > export lastfhr=18
> > > > > > > export incrementhr=03
> > > > > > >
> > > > > > > I'd suggest modifying the...
> > > > > > > - "startdate" to be the model initialization time to be
> > processed.
> > > > > > > - Leave "fhr" set to 00 to start at the analysis time.
> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
> > hours/day).
> > > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
> > > > > > >
> > > > > > > Give that a shot and let me know if that gets you output
past
> the
> > > > first
> > > > > > > day.
> > > > > > >
> > > > > > > Are you going to be evaluating precipitation? I ask
that
> because
> > > > > precip
> > > > > > is
> > > > > > > an accumulated interval, meaning that it has a beginning
> > > accumulation
> > > > > > time
> > > > > > > and ending accumulation time. GRIB1 is limited in the
amount
> of
> > > > space
> > > > > > > available for storing times. When you get past 255
forecast
> > hours,
> > > > you
> > > > > > can
> > > > > > > no longer store beginning/ending times for precip
because it
> runs
> > > out
> > > > > of
> > > > > > > room in the header. Do you have a 10 year simulation
computed
> > > from a
> > > > > > > single initialization? Meaning, do you have forecast
lead
> times
> > > out
> > > > to
> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
wondering
> > if
> > > > > we'll
> > > > > > > run into problems formatting those time strings.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > >
> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
> > > > > met_help at ucar.edu
> > > > > > <javascript:;>>
> > > > > > > wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >
> > > > > > > >
> > > > > > > > Dear John,
> > > > > > > >
> > > > > > > > I ran UPP with my multiple wrfout files, first with
the first
> > > month
> > > > > of
> > > > > > > the
> > > > > > > > 10 year simulation. However, after executing the
run_unipost
> > > > script,
> > > > > > UPP
> > > > > > > > produced only four files: unipost_d01.00.out,
> > unipost_d01.06.out,
> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems
that UPP
> > > > > processed
> > > > > > > > only the first day's output among the one month data.
> > > > > > > > If you are familiar with UPP, how can I make sure for
UPP to
> > > > process
> > > > > > > > multiple wrfout files more than one day?
> > > > > > > > Thank you.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > >
> > > > > > > > Jinwoong Yoo
> > > > > > > > UNM
> > > > > > > >
> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> > > > > jinwoong.yoo at gmail.com
> > > > > > <javascript:;>>
> > > > > > > > wrote:
> > > > > > > >
> > > > > > > > > Dear John,
> > > > > > > > >
> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
> > > > > > > > > Do I need to post process of the CCSM4 files also to
run
> MET
> > > > along
> > > > > > with
> > > > > > > > > the WRF-ARW files?
> > > > > > > > > Thank you.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > >
> > > > > > > > > Jinwoong Yoo
> > > > > > > > > UNM
> > > > > > > > >
> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
> > > > > > > > > wrote:
> > > > > > > > >
> > > > > > > > >> Dear John,
> > > > > > > > >>
> > > > > > > > >> Thank you for your kind email with the direction.
> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs using
> > UPP
> > > > and
> > > > > > let
> > > > > > > > >> you know.
> > > > > > > > >> Thank you very much.
> > > > > > > > >>
> > > > > > > > >> Regards,
> > > > > > > > >>
> > > > > > > > >> Jinwoong Yoo
> > > > > > > > >> UNM
> > > > > > > > >>
> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley Gotway
via
> RT <
> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > > > > > > > >>
> > > > > > > > >>> Jinwoong,
> > > > > > > > >>>
> > > > > > > > >>> I see that you'd like some help running the
> series-analysis
> > > > tool
> > > > > on
> > > > > > > > your
> > > > > > > > >>> wrf-arw climate runs. One thing to note is that
MET
> can't
> > > read
> > > > > the
> > > > > > > raw
> > > > > > > > >>> wrfout files directly. Instead you first need to
post
> > > process
> > > > > them
> > > > > > > > using
> > > > > > > > >>> either UPP (produces GRIB output) or the p_interp
utility
> > > > > (produced
> > > > > > > > >>> NetCDF
> > > > > > > > >>> output). After you've post processed some files,
please
> > send
> > > > us
> > > > > > some
> > > > > > > > >>> sample data by posting it to our anonymous ftp
site.
> Also
> > > > send a
> > > > > > > > >>> description of the comparisons you'd like to make
between
> > > the 2
> > > > > > > > datasets.
> > > > > > > > >>>
> > > > > > > > >>> Here's instructions for sending us data:
> > > > > > > > >>>
http://www.dtcenter.org/met/users/support/met_help.php
> > > > > > > > >>>
> > > > > > > > >>> Thanks
> > > > > > > > >>> John Halley Gotway
> > > > > > > > >>> met_help at ucar.edu <javascript:;>
> > > > > > > > >>>
> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo via
RT <
> > > > > > > > met_help at ucar.edu <javascript:;>
> > > > > > > > >>> >
> > > > > > > > >>> wrote:
> > > > > > > > >>>
> > > > > > > > >>> >
> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted upon.
> > > > > > > > >>> > Transaction: Ticket created by
jinwoong.yoo at gmail.com
> > > > > > <javascript:;>
> > > > > > > <javascript:;>
> > > > > > > > >>> > Queue: met_help
> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
> outputs
> > > > > > > > >>> > Owner: Nobody
> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
> > > > > > <javascript:;>
> > > > > > > > >>> > Status: new
> > > > > > > > >>> > Ticket <URL:
> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > > > > > > >>> >
> > > > > > > > >>> >
> > > > > > > > >>> >
> > > > > > > > >>> > Dear Met Help,
> > > > > > > > >>> >
> > > > > > > > >>> > Hi.
> > > > > > > > >>> > I am looking for a way to verify WRF model
performance
> > as a
> > > > > > climate
> > > > > > > > >>> model.
> > > > > > > > >>> > I'm completed 10-year tropical channel climate
model
> > > > > simulations
> > > > > > > > using
> > > > > > > > >>> the
> > > > > > > > >>> > CCSM4 data as input data (dynamical downscaling)
for
> two
> > > > > > different
> > > > > > > > time
> > > > > > > > >>> > periods: one for the LGM and the other for the
modern
> as
> > a
> > > > > > control
> > > > > > > > run.
> > > > > > > > >>> > The WRF simulation produced 6 hourly output for
ten
> years
> > > > > periods
> > > > > > > > each.
> > > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
outputs.
> > > > > > > > >>> >
> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
> > results,
> > > > > > > verifying
> > > > > > > > >>> the
> > > > > > > > >>> > WRF model simulations.
> > > > > > > > >>> > (WRF ARW was implemented for the experiments.)
> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
similar
> > > jobs
> > > > > but
> > > > > > I
> > > > > > > > need
> > > > > > > > >>> > much more help from you than the MET Online
Tutorial to
> > > make
> > > > a
> > > > > > step
> > > > > > > > >>> > forward.
> > > > > > > > >>> > Any help will be highly appreciated.
> > > > > > > > >>> > Thank you.
> > > > > > > > >>> >
> > > > > > > > >>> > Regards,
> > > > > > > > >>> >
> > > > > > > > >>> > Jinwoong Yoo
> > > > > > > > >>> > Postdoc
> > > > > > > > >>> > Department of Earth and Planetary Sciences
> > > > > > > > >>> > UNM
> > > > > > > > >>> >
> > > > > > > > >>> >
> > > > > > > > >>>
> > > > > > > > >>>
> > > > > > > > >>
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Julie Prestopnik
> > > National Center for Atmospheric Research
> > > Research Applications Laboratory
> > > Phone: 303.497.8399
> > > Email: jpresto at ucar.edu
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Oct 02 16:51:58 2014
Dear John,
Let me have a quick question about getting correct DX and DY in copygb
arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*"
-x
in.grb out.grb).
As you may know, grid info in my UPP output is as below:
Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny 349650
long 0.449000 to -0.449000 by 36.000000 km, (1110 x 315)
scan 64
mode 136 Latin 0.000000 bdsgrid 1
So my argument should be something like this:
*copygb.exe \*
* -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
However, there are something unclear in the document (
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
).
NX : Number of points on latitude circle
NY : Number of points on longitude meridian
DX : Longitudinal increment in millidegrees
DY : Latitudinal increment in millidegrees
You would notice that X,Y are not consistent as they refer to lat/lon.
I think flipping X/Y in NX/NY should work so that X represents
longitude.
That's not a big deal.
Here my own problem goes.
dx and dy in my WRF simulation are 36 km both in Lat and Lon.
However, if I convert them into millidegrees, DX is 323.515315 and DY
is
287.23175 computationally supposing the earth is a complete sphere.
copygb
takes only integers, DX is 323 and DY is 287. But, I'm not sure if
it's
right to use the different DX and DY or if I should use the same
number for DX
and DY.
Please let me know.
Thank you.
Jinwoong Yoo
UNM
On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> Dear John,
>
> Actually I was looking over copygb when I found your email received.
> It seems that I have to convert the UPP output again using
copygb.exe to
> match up the grid of WRF results to the CCSM4's lat/lon grid.
>
> Isn't there any way to do that in one step? No?
>
> Along with the precipitation, I'd like to compare variables (paired
for
> WRF and CCSM4) below for example:
>
> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>
>
> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>
>
> Varwrfclm = (/"T2", "TSLB","SMOIS",
"SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
> "FSDS","SNOWICE","Q2M","RH2M"/)
>
> Variables in the lists are the actual name of variables in the CCSM4
> (including CAM and CLM monthly files) files and corresponding to
those in
> the WRF-ARW.
>
>
> I'm trying to run Grid-Stat for a test and I'd like to run Series-
Analysis
> for the entire dataset of the WRF and the CCSM4.
> However, I'm concerning about how to run the Series-Analysis for the
> monthly or annual data using MET?
> At the same time, I am also thinking about using the NCL to do the
same
> statistical analyses. I don't know which one will be easier for me
to
> execute for now.
> We'll see.
>
> Will let you know as I make progress, John.
>
> >>> FYI - I just talked to the person who provides support for UPP
through
> >>> wrfhelp, and she's expecting an email from you.
>
> I cc'ed this email to wrfhelp.
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
>> Jinwoong,
>>
>> I'm able to access your data on Yellowstone. I see that you'd like
to
>> compare your UPP output in GRIB format to some data that follows
the
>> NetCDF
>> CF-1.0 convention.
>>
>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>
>> Hopefully you're able to look at those files and identify which
GRIB
>> records you'd like to compare to which NetCDF variables. When you
run the
>> MET tools, you'll define the forecast and observation fields to be
used.
>>
>> Looking at your data, I see 2 immediate problems:
>>
>> (1) Gridded forecast and observation data must be on the same grid
before
>> MET can compare them.
>> Here's a description from wgrib of the grid for your GRIB data:
>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>> But it looks like the CCSM4 data is on a lat/lon grid.
>>
>> I'd suggest using the copygb utility to regrid the GRIB1 output
from UPP
>> to
>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>>
>>
>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>
>> (2) There's an issue in the times in the accumulated precipitation.
Try
>> running the following 2 commands:
>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
>> grep APCP
>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
>> grep APCP
>>
>> For the first, APCP is an accumulation of precip from hours 0 to
252. The
>> second should be 0 to 258, but it's stored as 0 to 2! That's
because UPP
>> is failing to realize that the precipitation timestamps are
overflowing
>> the
>> 1 byte allocated for that second time unit. This problem was
addressed in
>> an earlier version of UPP but the problem is still showing up in
the
>> version you're using. If you plan to verify precip, I'd suggest
writing
>> wrfhelp at ucar.edu, tell them the version of UPP you're using, and
ask them
>> if this should be fixed.
>>
>> Hope that helps.
>>
>> Thanks,
>> John
>>
>>
>>
>>
>>
>>
>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
>> wrote:
>>
>> >
>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>> >
>> > Hi, Julie.
>> > Thank you for heads-up.
>> > I will wait for John to reply then.
>> > Thank you.
>> >
>> > Jinwoong Yoo
>> > UNM
>> >
>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>> > met_help at ucar.edu>
>> > wrote:
>> >
>> > > Hello. John is out of the office, but will be back later this
week.
>> I am
>> > > unable to help you with this task. Hopefully, John will be
able to
>> > respond
>> > > to this ticket later this week.
>> > >
>> > > Julie
>> > >
>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>> met_help at ucar.edu>
>> > > wrote:
>> > >
>> > > >
>> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
>> > > >
>> > > > Dear John,
>> > > >
>> > > > I ran UPP successfully on the UCAR Yellowstone.
>> > > > I assume you can access to my glade scratch directory.
>> > > > My UPP ouputs are located
>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>> > > >
>> > > > Against those files, I'd like to compare CCSM4 simulation
outputs
>> some
>> > of
>> > > > which are located also under
>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>> > > >
>> > > > Those are six hourly files all together. If you cannot access
to
>> those
>> > > > files, please let me know.
>> > > >
>> > > > However, there are some variables among the CCSM4 outputs
that are
>> > > > available only monthly or annually. In this case, how can I
>> > post-process
>> > > my
>> > > > WRF outputs for the usage of MET through UPP?
>> > > >
>> > > > Thank you.
>> > > >
>> > > > Regards,
>> > > >
>> > > > Jinwoong Yoo
>> > > > UNM
>> > > >
>> > > >
>> > > >
>> > > >
>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT <
>> > > > met_help at ucar.edu> wrote:
>> > > >
>> > > > > Jinwoo,
>> > > > >
>> > > > > I believe UPP will still run, just using a single time
value as
>> if it
>> > > > were
>> > > > > an instantaneous field rather than an accumulation
interval. The
>> > > default
>> > > > > for ARW is to store precip as a runtime accumulation. For
>> forecasts
>> > > > hours
>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
>> accumulation
>> > > > from 0
>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored as an
>> > > > > instantaneous 300 hour forecast. It's a subtle difference,
but
>> > > something
>> > > > > to be aware of when you set up MET config files to verify
the
>> output.
>> > > > >
>> > > > > Give UPP a try and let me know how it goes.
>> > > > >
>> > > > > John
>> > > > >
>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>> > met_help at ucar.edu>
>> > > > > wrote:
>> > > > >
>> > > > > >
>> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>> > > > > >
>> > > > > > Dear John,
>> > > > > >
>> > > > > > I see. Let me change "export lastfhr=18" to "export
>> lastfhr=744".
>> > > > > > Thank you for your comments.
>> > > > > > Yes, I'd like to evaluate precipitation, too, and my 10
year
>> > > simulation
>> > > > > was
>> > > > > > integrated from a single initialization. Then, will the
UPP
>> > > run_unipost
>> > > > > > crash after the 255 forecast hours or just will
precipitation
>> not
>> > be
>> > > > > stored
>> > > > > > any more aster the 255 forecast hours?
>> > > > > >
>> > > > > > Thank you.
>> > > > > >
>> > > > > > Regards,
>> > > > > >
>> > > > > > Jinwoong Yoo
>> > > > > > UNM
>> > > > > >
>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via
RT <
>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>> > > > > >
>> > > > > > > Jinwoong,
>> > > > > > >
>> > > > > > > UPP support is typically provided through
wrf_help at ucar.edu
>> > > > > > <javascript:;>. I'll give it
>> > > > > > > a shot, but if this doesn't solve it, I'll refer you to
>> wrf_help.
>> > > > > > >
>> > > > > > > I assume you're running the script named "run_unipost".
>> Looking
>> > in
>> > > > > > there,
>> > > > > > > you'll see:
>> > > > > > >
>> > > > > > > export startdate=2005012300
>> > > > > > >
>> > > > > > > export fhr=00
>> > > > > > > export lastfhr=18
>> > > > > > > export incrementhr=03
>> > > > > > >
>> > > > > > > I'd suggest modifying the...
>> > > > > > > - "startdate" to be the model initialization time to
be
>> > processed.
>> > > > > > > - Leave "fhr" set to 00 to start at the analysis time.
>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days * 24
>> > hours/day).
>> > > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
>> > > > > > >
>> > > > > > > Give that a shot and let me know if that gets you
output past
>> the
>> > > > first
>> > > > > > > day.
>> > > > > > >
>> > > > > > > Are you going to be evaluating precipitation? I ask
that
>> because
>> > > > > precip
>> > > > > > is
>> > > > > > > an accumulated interval, meaning that it has a
beginning
>> > > accumulation
>> > > > > > time
>> > > > > > > and ending accumulation time. GRIB1 is limited in the
amount
>> of
>> > > > space
>> > > > > > > available for storing times. When you get past 255
forecast
>> > hours,
>> > > > you
>> > > > > > can
>> > > > > > > no longer store beginning/ending times for precip
because it
>> runs
>> > > out
>> > > > > of
>> > > > > > > room in the header. Do you have a 10 year simulation
computed
>> > > from a
>> > > > > > > single initialization? Meaning, do you have forecast
lead
>> times
>> > > out
>> > > > to
>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
>> wondering
>> > if
>> > > > > we'll
>> > > > > > > run into problems formatting those time strings.
>> > > > > > >
>> > > > > > > Thanks,
>> > > > > > > John
>> > > > > > >
>> > > > > > >
>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
>> > > > > met_help at ucar.edu
>> > > > > > <javascript:;>>
>> > > > > > > wrote:
>> > > > > > >
>> > > > > > > >
>> > > > > > > > <URL:
>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>> > >
>> > > > > > > >
>> > > > > > > > Dear John,
>> > > > > > > >
>> > > > > > > > I ran UPP with my multiple wrfout files, first with
the
>> first
>> > > month
>> > > > > of
>> > > > > > > the
>> > > > > > > > 10 year simulation. However, after executing the
run_unipost
>> > > > script,
>> > > > > > UPP
>> > > > > > > > produced only four files: unipost_d01.00.out,
>> > unipost_d01.06.out,
>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems
that
>> UPP
>> > > > > processed
>> > > > > > > > only the first day's output among the one month data.
>> > > > > > > > If you are familiar with UPP, how can I make sure for
UPP to
>> > > > process
>> > > > > > > > multiple wrfout files more than one day?
>> > > > > > > > Thank you.
>> > > > > > > >
>> > > > > > > > Regards,
>> > > > > > > >
>> > > > > > > > Jinwoong Yoo
>> > > > > > > > UNM
>> > > > > > > >
>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>> > > > > jinwoong.yoo at gmail.com
>> > > > > > <javascript:;>>
>> > > > > > > > wrote:
>> > > > > > > >
>> > > > > > > > > Dear John,
>> > > > > > > > >
>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
>> > > > > > > > > Do I need to post process of the CCSM4 files also
to run
>> MET
>> > > > along
>> > > > > > with
>> > > > > > > > > the WRF-ARW files?
>> > > > > > > > > Thank you.
>> > > > > > > > >
>> > > > > > > > > Regards,
>> > > > > > > > >
>> > > > > > > > > Jinwoong Yoo
>> > > > > > > > > UNM
>> > > > > > > > >
>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>> > > > > > > > > wrote:
>> > > > > > > > >
>> > > > > > > > >> Dear John,
>> > > > > > > > >>
>> > > > > > > > >> Thank you for your kind email with the direction.
>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
>> using
>> > UPP
>> > > > and
>> > > > > > let
>> > > > > > > > >> you know.
>> > > > > > > > >> Thank you very much.
>> > > > > > > > >>
>> > > > > > > > >> Regards,
>> > > > > > > > >>
>> > > > > > > > >> Jinwoong Yoo
>> > > > > > > > >> UNM
>> > > > > > > > >>
>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway via
>> RT <
>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>> > > > > > > > >>
>> > > > > > > > >>> Jinwoong,
>> > > > > > > > >>>
>> > > > > > > > >>> I see that you'd like some help running the
>> series-analysis
>> > > > tool
>> > > > > on
>> > > > > > > > your
>> > > > > > > > >>> wrf-arw climate runs. One thing to note is that
MET
>> can't
>> > > read
>> > > > > the
>> > > > > > > raw
>> > > > > > > > >>> wrfout files directly. Instead you first need to
post
>> > > process
>> > > > > them
>> > > > > > > > using
>> > > > > > > > >>> either UPP (produces GRIB output) or the p_interp
>> utility
>> > > > > (produced
>> > > > > > > > >>> NetCDF
>> > > > > > > > >>> output). After you've post processed some files,
please
>> > send
>> > > > us
>> > > > > > some
>> > > > > > > > >>> sample data by posting it to our anonymous ftp
site.
>> Also
>> > > > send a
>> > > > > > > > >>> description of the comparisons you'd like to make
>> between
>> > > the 2
>> > > > > > > > datasets.
>> > > > > > > > >>>
>> > > > > > > > >>> Here's instructions for sending us data:
>> > > > > > > > >>>
http://www.dtcenter.org/met/users/support/met_help.php
>> > > > > > > > >>>
>> > > > > > > > >>> Thanks
>> > > > > > > > >>> John Halley Gotway
>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>> > > > > > > > >>>
>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo
via RT <
>> > > > > > > > met_help at ucar.edu <javascript:;>
>> > > > > > > > >>> >
>> > > > > > > > >>> wrote:
>> > > > > > > > >>>
>> > > > > > > > >>> >
>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
>> upon.
>> > > > > > > > >>> > Transaction: Ticket created by
jinwoong.yoo at gmail.com
>> > > > > > <javascript:;>
>> > > > > > > <javascript:;>
>> > > > > > > > >>> > Queue: met_help
>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
>> outputs
>> > > > > > > > >>> > Owner: Nobody
>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
>> > > > > > <javascript:;>
>> > > > > > > > >>> > Status: new
>> > > > > > > > >>> > Ticket <URL:
>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>> > > > > > > > >>> >
>> > > > > > > > >>> >
>> > > > > > > > >>> >
>> > > > > > > > >>> > Dear Met Help,
>> > > > > > > > >>> >
>> > > > > > > > >>> > Hi.
>> > > > > > > > >>> > I am looking for a way to verify WRF model
performance
>> > as a
>> > > > > > climate
>> > > > > > > > >>> model.
>> > > > > > > > >>> > I'm completed 10-year tropical channel climate
model
>> > > > > simulations
>> > > > > > > > using
>> > > > > > > > >>> the
>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling) for
>> two
>> > > > > > different
>> > > > > > > > time
>> > > > > > > > >>> > periods: one for the LGM and the other for the
modern
>> as
>> > a
>> > > > > > control
>> > > > > > > > run.
>> > > > > > > > >>> > The WRF simulation produced 6 hourly output for
ten
>> years
>> > > > > periods
>> > > > > > > > each.
>> > > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
outputs.
>> > > > > > > > >>> >
>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>> > results,
>> > > > > > > verifying
>> > > > > > > > >>> the
>> > > > > > > > >>> > WRF model simulations.
>> > > > > > > > >>> > (WRF ARW was implemented for the experiments.)
>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do the
>> similar
>> > > jobs
>> > > > > but
>> > > > > > I
>> > > > > > > > need
>> > > > > > > > >>> > much more help from you than the MET Online
Tutorial
>> to
>> > > make
>> > > > a
>> > > > > > step
>> > > > > > > > >>> > forward.
>> > > > > > > > >>> > Any help will be highly appreciated.
>> > > > > > > > >>> > Thank you.
>> > > > > > > > >>> >
>> > > > > > > > >>> > Regards,
>> > > > > > > > >>> >
>> > > > > > > > >>> > Jinwoong Yoo
>> > > > > > > > >>> > Postdoc
>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>> > > > > > > > >>> > UNM
>> > > > > > > > >>> >
>> > > > > > > > >>> >
>> > > > > > > > >>>
>> > > > > > > > >>>
>> > > > > > > > >>
>> > > > > > > > >
>> > > > > > > >
>> > > > > > > >
>> > > > > > >
>> > > > > > >
>> > > > > >
>> > > > > >
>> > > > >
>> > > > >
>> > > >
>> > > >
>> > >
>> > >
>> > > --
>> > > Julie Prestopnik
>> > > National Center for Atmospheric Research
>> > > Research Applications Laboratory
>> > > Phone: 303.497.8399
>> > > Email: jpresto at ucar.edu
>> > >
>> > >
>> >
>> >
>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Thu Oct 02 17:50:25 2014
Am I misunderstanding the arguments of copygb?
(copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x in.grb
out.grb
)
Should be those numbers for the target grid in Lat/Lon of the CCSM4 in
my
case, not for the WRF output values?
Such that
*NX (*Number of points on latitude circle) : 192
*NY (*Number of points on longitude meridian*) : 288*
*LAT0 : -90*
*LON0 : 0*
*LAT1 : 90*
*LON1 : 358.75*
*DX : 942*
*DY : 1250*
Please let me know.
Thank you.
Jinwoong Yoo
UNM
On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> Dear John,
>
> Let me have a quick question about getting correct DX and DY in
copygb
> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*"
-x
> in.grb out.grb).
>
>
> As you may know, grid info in my UPP output is as below:
> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny 349650
> long 0.449000 to -0.449000 by 36.000000 km, (1110 x 315)
scan 64
> mode 136 Latin 0.000000 bdsgrid 1
>
> So my argument should be something like this:
> *copygb.exe \*
> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
>
> However, there are something unclear in the document (
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> ).
>
> NX : Number of points on latitude circle
> NY : Number of points on longitude meridian
> DX : Longitudinal increment in millidegrees
> DY : Latitudinal increment in millidegrees
>
> You would notice that X,Y are not consistent as they refer to
lat/lon.
> I think flipping X/Y in NX/NY should work so that X represents
longitude.
>
> That's not a big deal.
> Here my own problem goes.
>
> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
> However, if I convert them into millidegrees, DX is 323.515315 and
DY is
> 287.23175 computationally supposing the earth is a complete sphere.
copygb
> takes only integers, DX is 323 and DY is 287. But, I'm not sure if
it's
> right to use the different DX and DY or if I should use the same
number for DX
> and DY.
> Please let me know.
>
> Thank you.
>
> Jinwoong Yoo
> UNM
>
>
>
>
>
>
>
>
> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> Dear John,
>>
>> Actually I was looking over copygb when I found your email
received.
>> It seems that I have to convert the UPP output again using
copygb.exe to
>> match up the grid of WRF results to the CCSM4's lat/lon grid.
>>
>> Isn't there any way to do that in one step? No?
>>
>> Along with the precipitation, I'd like to compare variables (paired
for
>> WRF and CCSM4) below for example:
>>
>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>>
>>
>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>>
>>
>> Varwrfclm = (/"T2", "TSLB","SMOIS",
>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
>> "FSDS","SNOWICE","Q2M","RH2M"/)
>>
>> Variables in the lists are the actual name of variables in the
CCSM4
>> (including CAM and CLM monthly files) files and corresponding to
those in
>> the WRF-ARW.
>>
>>
>> I'm trying to run Grid-Stat for a test and I'd like to run
>> Series-Analysis for the entire dataset of the WRF and the CCSM4.
>> However, I'm concerning about how to run the Series-Analysis for
the
>> monthly or annual data using MET?
>> At the same time, I am also thinking about using the NCL to do the
same
>> statistical analyses. I don't know which one will be easier for me
to
>> execute for now.
>> We'll see.
>>
>> Will let you know as I make progress, John.
>>
>> >>> FYI - I just talked to the person who provides support for UPP
through
>> >>> wrfhelp, and she's expecting an email from you.
>>
>> I cc'ed this email to wrfhelp.
>> Thank you.
>>
>> Regards,
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
>> met_help at ucar.edu> wrote:
>>
>>> Jinwoong,
>>>
>>> I'm able to access your data on Yellowstone. I see that you'd
like to
>>> compare your UPP output in GRIB format to some data that follows
the
>>> NetCDF
>>> CF-1.0 convention.
>>>
>>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>>
>>> Hopefully you're able to look at those files and identify which
GRIB
>>> records you'd like to compare to which NetCDF variables. When you
run
>>> the
>>> MET tools, you'll define the forecast and observation fields to be
used.
>>>
>>> Looking at your data, I see 2 immediate problems:
>>>
>>> (1) Gridded forecast and observation data must be on the same grid
before
>>> MET can compare them.
>>> Here's a description from wgrib of the grid for your GRIB data:
>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
>>> long 0.449000 to -0.449000 by 36.000000 km, (1110
x 315)
>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>> But it looks like the CCSM4 data is on a lat/lon grid.
>>>
>>> I'd suggest using the copygb utility to regrid the GRIB1 output
from UPP
>>> to
>>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>>>
>>>
>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>
>>> (2) There's an issue in the times in the accumulated
precipitation. Try
>>> running the following 2 commands:
>>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
>>> grep APCP
>>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
>>> grep APCP
>>>
>>> For the first, APCP is an accumulation of precip from hours 0 to
252.
>>> The
>>> second should be 0 to 258, but it's stored as 0 to 2! That's
because UPP
>>> is failing to realize that the precipitation timestamps are
overflowing
>>> the
>>> 1 byte allocated for that second time unit. This problem was
addressed
>>> in
>>> an earlier version of UPP but the problem is still showing up in
the
>>> version you're using. If you plan to verify precip, I'd suggest
writing
>>> wrfhelp at ucar.edu, tell them the version of UPP you're using, and
ask
>>> them
>>> if this should be fixed.
>>>
>>> Hope that helps.
>>>
>>> Thanks,
>>> John
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
>>> wrote:
>>>
>>> >
>>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>> >
>>> > Hi, Julie.
>>> > Thank you for heads-up.
>>> > I will wait for John to reply then.
>>> > Thank you.
>>> >
>>> > Jinwoong Yoo
>>> > UNM
>>> >
>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>>> > met_help at ucar.edu>
>>> > wrote:
>>> >
>>> > > Hello. John is out of the office, but will be back later this
week.
>>> I am
>>> > > unable to help you with this task. Hopefully, John will be
able to
>>> > respond
>>> > > to this ticket later this week.
>>> > >
>>> > > Julie
>>> > >
>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>>> met_help at ucar.edu>
>>> > > wrote:
>>> > >
>>> > > >
>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>> > > >
>>> > > > Dear John,
>>> > > >
>>> > > > I ran UPP successfully on the UCAR Yellowstone.
>>> > > > I assume you can access to my glade scratch directory.
>>> > > > My UPP ouputs are located
>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>>> > > >
>>> > > > Against those files, I'd like to compare CCSM4 simulation
outputs
>>> some
>>> > of
>>> > > > which are located also under
>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>>> > > >
>>> > > > Those are six hourly files all together. If you cannot
access to
>>> those
>>> > > > files, please let me know.
>>> > > >
>>> > > > However, there are some variables among the CCSM4 outputs
that are
>>> > > > available only monthly or annually. In this case, how can I
>>> > post-process
>>> > > my
>>> > > > WRF outputs for the usage of MET through UPP?
>>> > > >
>>> > > > Thank you.
>>> > > >
>>> > > > Regards,
>>> > > >
>>> > > > Jinwoong Yoo
>>> > > > UNM
>>> > > >
>>> > > >
>>> > > >
>>> > > >
>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT
<
>>> > > > met_help at ucar.edu> wrote:
>>> > > >
>>> > > > > Jinwoo,
>>> > > > >
>>> > > > > I believe UPP will still run, just using a single time
value as
>>> if it
>>> > > > were
>>> > > > > an instantaneous field rather than an accumulation
interval. The
>>> > > default
>>> > > > > for ARW is to store precip as a runtime accumulation. For
>>> forecasts
>>> > > > hours
>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
>>> accumulation
>>> > > > from 0
>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored as
>>> an
>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference, but
>>> > > something
>>> > > > > to be aware of when you set up MET config files to verify
the
>>> output.
>>> > > > >
>>> > > > > Give UPP a try and let me know how it goes.
>>> > > > >
>>> > > > > John
>>> > > > >
>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>>> > met_help at ucar.edu>
>>> > > > > wrote:
>>> > > > >
>>> > > > > >
>>> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>> >
>>> > > > > >
>>> > > > > > Dear John,
>>> > > > > >
>>> > > > > > I see. Let me change "export lastfhr=18" to "export
>>> lastfhr=744".
>>> > > > > > Thank you for your comments.
>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my 10
year
>>> > > simulation
>>> > > > > was
>>> > > > > > integrated from a single initialization. Then, will the
UPP
>>> > > run_unipost
>>> > > > > > crash after the 255 forecast hours or just will
precipitation
>>> not
>>> > be
>>> > > > > stored
>>> > > > > > any more aster the 255 forecast hours?
>>> > > > > >
>>> > > > > > Thank you.
>>> > > > > >
>>> > > > > > Regards,
>>> > > > > >
>>> > > > > > Jinwoong Yoo
>>> > > > > > UNM
>>> > > > > >
>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via
RT <
>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>>> > > > > >
>>> > > > > > > Jinwoong,
>>> > > > > > >
>>> > > > > > > UPP support is typically provided through
wrf_help at ucar.edu
>>> > > > > > <javascript:;>. I'll give it
>>> > > > > > > a shot, but if this doesn't solve it, I'll refer you
to
>>> wrf_help.
>>> > > > > > >
>>> > > > > > > I assume you're running the script named
"run_unipost".
>>> Looking
>>> > in
>>> > > > > > there,
>>> > > > > > > you'll see:
>>> > > > > > >
>>> > > > > > > export startdate=2005012300
>>> > > > > > >
>>> > > > > > > export fhr=00
>>> > > > > > > export lastfhr=18
>>> > > > > > > export incrementhr=03
>>> > > > > > >
>>> > > > > > > I'd suggest modifying the...
>>> > > > > > > - "startdate" to be the model initialization time to
be
>>> > processed.
>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days *
24
>>> > hours/day).
>>> > > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
>>> > > > > > >
>>> > > > > > > Give that a shot and let me know if that gets you
output
>>> past the
>>> > > > first
>>> > > > > > > day.
>>> > > > > > >
>>> > > > > > > Are you going to be evaluating precipitation? I ask
that
>>> because
>>> > > > > precip
>>> > > > > > is
>>> > > > > > > an accumulated interval, meaning that it has a
beginning
>>> > > accumulation
>>> > > > > > time
>>> > > > > > > and ending accumulation time. GRIB1 is limited in the
>>> amount of
>>> > > > space
>>> > > > > > > available for storing times. When you get past 255
forecast
>>> > hours,
>>> > > > you
>>> > > > > > can
>>> > > > > > > no longer store beginning/ending times for precip
because it
>>> runs
>>> > > out
>>> > > > > of
>>> > > > > > > room in the header. Do you have a 10 year simulation
>>> computed
>>> > > from a
>>> > > > > > > single initialization? Meaning, do you have forecast
lead
>>> times
>>> > > out
>>> > > > to
>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
>>> wondering
>>> > if
>>> > > > > we'll
>>> > > > > > > run into problems formatting those time strings.
>>> > > > > > >
>>> > > > > > > Thanks,
>>> > > > > > > John
>>> > > > > > >
>>> > > > > > >
>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT <
>>> > > > > met_help at ucar.edu
>>> > > > > > <javascript:;>>
>>> > > > > > > wrote:
>>> > > > > > >
>>> > > > > > > >
>>> > > > > > > > <URL:
>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>> > >
>>> > > > > > > >
>>> > > > > > > > Dear John,
>>> > > > > > > >
>>> > > > > > > > I ran UPP with my multiple wrfout files, first with
the
>>> first
>>> > > month
>>> > > > > of
>>> > > > > > > the
>>> > > > > > > > 10 year simulation. However, after executing the
>>> run_unipost
>>> > > > script,
>>> > > > > > UPP
>>> > > > > > > > produced only four files: unipost_d01.00.out,
>>> > unipost_d01.06.out,
>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It seems
that
>>> UPP
>>> > > > > processed
>>> > > > > > > > only the first day's output among the one month
data.
>>> > > > > > > > If you are familiar with UPP, how can I make sure
for UPP
>>> to
>>> > > > process
>>> > > > > > > > multiple wrfout files more than one day?
>>> > > > > > > > Thank you.
>>> > > > > > > >
>>> > > > > > > > Regards,
>>> > > > > > > >
>>> > > > > > > > Jinwoong Yoo
>>> > > > > > > > UNM
>>> > > > > > > >
>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>>> > > > > jinwoong.yoo at gmail.com
>>> > > > > > <javascript:;>>
>>> > > > > > > > wrote:
>>> > > > > > > >
>>> > > > > > > > > Dear John,
>>> > > > > > > > >
>>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
>>> > > > > > > > > Do I need to post process of the CCSM4 files also
to run
>>> MET
>>> > > > along
>>> > > > > > with
>>> > > > > > > > > the WRF-ARW files?
>>> > > > > > > > > Thank you.
>>> > > > > > > > >
>>> > > > > > > > > Regards,
>>> > > > > > > > >
>>> > > > > > > > > Jinwoong Yoo
>>> > > > > > > > > UNM
>>> > > > > > > > >
>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>>> > > > > > > > > wrote:
>>> > > > > > > > >
>>> > > > > > > > >> Dear John,
>>> > > > > > > > >>
>>> > > > > > > > >> Thank you for your kind email with the direction.
>>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
>>> using
>>> > UPP
>>> > > > and
>>> > > > > > let
>>> > > > > > > > >> you know.
>>> > > > > > > > >> Thank you very much.
>>> > > > > > > > >>
>>> > > > > > > > >> Regards,
>>> > > > > > > > >>
>>> > > > > > > > >> Jinwoong Yoo
>>> > > > > > > > >> UNM
>>> > > > > > > > >>
>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway via
>>> RT <
>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>>> > > > > > > > >>
>>> > > > > > > > >>> Jinwoong,
>>> > > > > > > > >>>
>>> > > > > > > > >>> I see that you'd like some help running the
>>> series-analysis
>>> > > > tool
>>> > > > > on
>>> > > > > > > > your
>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is that
MET
>>> can't
>>> > > read
>>> > > > > the
>>> > > > > > > raw
>>> > > > > > > > >>> wrfout files directly. Instead you first need
to post
>>> > > process
>>> > > > > them
>>> > > > > > > > using
>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
>>> utility
>>> > > > > (produced
>>> > > > > > > > >>> NetCDF
>>> > > > > > > > >>> output). After you've post processed some
files,
>>> please
>>> > send
>>> > > > us
>>> > > > > > some
>>> > > > > > > > >>> sample data by posting it to our anonymous ftp
site.
>>> Also
>>> > > > send a
>>> > > > > > > > >>> description of the comparisons you'd like to
make
>>> between
>>> > > the 2
>>> > > > > > > > datasets.
>>> > > > > > > > >>>
>>> > > > > > > > >>> Here's instructions for sending us data:
>>> > > > > > > > >>>
http://www.dtcenter.org/met/users/support/met_help.php
>>> > > > > > > > >>>
>>> > > > > > > > >>> Thanks
>>> > > > > > > > >>> John Halley Gotway
>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>>> > > > > > > > >>>
>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo
via RT <
>>> > > > > > > > met_help at ucar.edu <javascript:;>
>>> > > > > > > > >>> >
>>> > > > > > > > >>> wrote:
>>> > > > > > > > >>>
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
>>> upon.
>>> > > > > > > > >>> > Transaction: Ticket created by
>>> jinwoong.yoo at gmail.com
>>> > > > > > <javascript:;>
>>> > > > > > > <javascript:;>
>>> > > > > > > > >>> > Queue: met_help
>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
>>> outputs
>>> > > > > > > > >>> > Owner: Nobody
>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
>>> > > > > > <javascript:;>
>>> > > > > > > > >>> > Status: new
>>> > > > > > > > >>> > Ticket <URL:
>>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>> > > > > > > > >>> >
>>> > > > > > > > >>> >
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Dear Met Help,
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Hi.
>>> > > > > > > > >>> > I am looking for a way to verify WRF model
>>> performance
>>> > as a
>>> > > > > > climate
>>> > > > > > > > >>> model.
>>> > > > > > > > >>> > I'm completed 10-year tropical channel climate
model
>>> > > > > simulations
>>> > > > > > > > using
>>> > > > > > > > >>> the
>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling) for
>>> two
>>> > > > > > different
>>> > > > > > > > time
>>> > > > > > > > >>> > periods: one for the LGM and the other for the
>>> modern as
>>> > a
>>> > > > > > control
>>> > > > > > > > run.
>>> > > > > > > > >>> > The WRF simulation produced 6 hourly output
for ten
>>> years
>>> > > > > periods
>>> > > > > > > > each.
>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
>>> outputs.
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>>> > results,
>>> > > > > > > verifying
>>> > > > > > > > >>> the
>>> > > > > > > > >>> > WRF model simulations.
>>> > > > > > > > >>> > (WRF ARW was implemented for the experiments.)
>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do
the
>>> similar
>>> > > jobs
>>> > > > > but
>>> > > > > > I
>>> > > > > > > > need
>>> > > > > > > > >>> > much more help from you than the MET Online
Tutorial
>>> to
>>> > > make
>>> > > > a
>>> > > > > > step
>>> > > > > > > > >>> > forward.
>>> > > > > > > > >>> > Any help will be highly appreciated.
>>> > > > > > > > >>> > Thank you.
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Regards,
>>> > > > > > > > >>> >
>>> > > > > > > > >>> > Jinwoong Yoo
>>> > > > > > > > >>> > Postdoc
>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>>> > > > > > > > >>> > UNM
>>> > > > > > > > >>> >
>>> > > > > > > > >>> >
>>> > > > > > > > >>>
>>> > > > > > > > >>>
>>> > > > > > > > >>
>>> > > > > > > > >
>>> > > > > > > >
>>> > > > > > > >
>>> > > > > > >
>>> > > > > > >
>>> > > > > >
>>> > > > > >
>>> > > > >
>>> > > > >
>>> > > >
>>> > > >
>>> > >
>>> > >
>>> > > --
>>> > > Julie Prestopnik
>>> > > National Center for Atmospheric Research
>>> > > Research Applications Laboratory
>>> > > Phone: 303.497.8399
>>> > > Email: jpresto at ucar.edu
>>> > >
>>> > >
>>> >
>>> >
>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Fri Oct 03 11:19:47 2014
That's right. I was wrong earlier and the later arguments worked for
my
regridding of the WRF output in Mercator to Lat-Lon grid.
Thanks.
Jinwoong Yoo
UNM
On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> Am I misunderstanding the arguments of copygb?
> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x in.grb
> out.grb)
>
> Should be those numbers for the target grid in Lat/Lon of the CCSM4
in my
> case, not for the WRF output values?
> Such that
> *NX (*Number of points on latitude circle) : 192
> *NY (*Number of points on longitude meridian*) : 288*
> *LAT0 : -90*
> *LON0 : 0*
> *LAT1 : 90*
> *LON1 : 358.75*
> *DX : 942*
> *DY : 1250*
>
> Please let me know.
> Thank you.
>
> Jinwoong Yoo
> UNM
>
> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> Dear John,
>>
>> Let me have a quick question about getting correct DX and DY in
copygb
>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY
64*" -x
>> in.grb out.grb).
>>
>>
>> As you may know, grid info in my UPP output is as below:
>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny 349650
>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x 315)
scan
>> 64 mode 136 Latin 0.000000 bdsgrid 1
>>
>> So my argument should be something like this:
>> *copygb.exe \*
>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
>>
>> However, there are something unclear in the document (
>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>> ).
>>
>> NX : Number of points on latitude circle
>> NY : Number of points on longitude meridian
>> DX : Longitudinal increment in millidegrees
>> DY : Latitudinal increment in millidegrees
>>
>> You would notice that X,Y are not consistent as they refer to
lat/lon.
>> I think flipping X/Y in NX/NY should work so that X represents
longitude.
>>
>> That's not a big deal.
>> Here my own problem goes.
>>
>> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
>> However, if I convert them into millidegrees, DX is 323.515315 and
DY is
>> 287.23175 computationally supposing the earth is a complete sphere.
copygb
>> takes only integers, DX is 323 and DY is 287. But, I'm not sure if
it's
>> right to use the different DX and DY or if I should use the same
number for DX
>> and DY.
>> Please let me know.
>>
>> Thank you.
>>
>> Jinwoong Yoo
>> UNM
>>
>>
>>
>>
>>
>>
>>
>>
>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>> wrote:
>>
>>> Dear John,
>>>
>>> Actually I was looking over copygb when I found your email
received.
>>> It seems that I have to convert the UPP output again using
copygb.exe to
>>> match up the grid of WRF results to the CCSM4's lat/lon grid.
>>>
>>> Isn't there any way to do that in one step? No?
>>>
>>> Along with the precipitation, I'd like to compare variables
(paired for
>>> WRF and CCSM4) below for example:
>>>
>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>>>
>>>
>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>>>
>>>
>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
>>> "FSDS","SNOWICE","Q2M","RH2M"/)
>>>
>>> Variables in the lists are the actual name of variables in the
CCSM4
>>> (including CAM and CLM monthly files) files and corresponding to
those in
>>> the WRF-ARW.
>>>
>>>
>>> I'm trying to run Grid-Stat for a test and I'd like to run
>>> Series-Analysis for the entire dataset of the WRF and the CCSM4.
>>> However, I'm concerning about how to run the Series-Analysis for
the
>>> monthly or annual data using MET?
>>> At the same time, I am also thinking about using the NCL to do the
same
>>> statistical analyses. I don't know which one will be easier for me
to
>>> execute for now.
>>> We'll see.
>>>
>>> Will let you know as I make progress, John.
>>>
>>> >>> FYI - I just talked to the person who provides support for UPP
>>> through
>>> >>> wrfhelp, and she's expecting an email from you.
>>>
>>> I cc'ed this email to wrfhelp.
>>> Thank you.
>>>
>>> Regards,
>>>
>>> Jinwoong Yoo
>>> UNM
>>>
>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Jinwoong,
>>>>
>>>> I'm able to access your data on Yellowstone. I see that you'd
like to
>>>> compare your UPP output in GRIB format to some data that follows
the
>>>> NetCDF
>>>> CF-1.0 convention.
>>>>
>>>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>>>
>>>> Hopefully you're able to look at those files and identify which
GRIB
>>>> records you'd like to compare to which NetCDF variables. When
you run
>>>> the
>>>> MET tools, you'll define the forecast and observation fields to
be used.
>>>>
>>>> Looking at your data, I see 2 immediate problems:
>>>>
>>>> (1) Gridded forecast and observation data must be on the same
grid
>>>> before
>>>> MET can compare them.
>>>> Here's a description from wgrib of the grid for your GRIB
data:
>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
>>>> 349650
>>>> long 0.449000 to -0.449000 by 36.000000 km, (1110
x 315)
>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>>> But it looks like the CCSM4 data is on a lat/lon grid.
>>>>
>>>> I'd suggest using the copygb utility to regrid the GRIB1 output
from
>>>> UPP to
>>>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>>>>
>>>>
>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>>
>>>> (2) There's an issue in the times in the accumulated
precipitation. Try
>>>> running the following 2 commands:
>>>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252
>>>> |
>>>> grep APCP
>>>> wgrib
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258
>>>> |
>>>> grep APCP
>>>>
>>>> For the first, APCP is an accumulation of precip from hours 0 to
252.
>>>> The
>>>> second should be 0 to 258, but it's stored as 0 to 2! That's
because
>>>> UPP
>>>> is failing to realize that the precipitation timestamps are
overflowing
>>>> the
>>>> 1 byte allocated for that second time unit. This problem was
addressed
>>>> in
>>>> an earlier version of UPP but the problem is still showing up in
the
>>>> version you're using. If you plan to verify precip, I'd suggest
writing
>>>> wrfhelp at ucar.edu, tell them the version of UPP you're using, and
ask
>>>> them
>>>> if this should be fixed.
>>>>
>>>> Hope that helps.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT
<met_help at ucar.edu
>>>> >
>>>> wrote:
>>>>
>>>> >
>>>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>> >
>>>> > Hi, Julie.
>>>> > Thank you for heads-up.
>>>> > I will wait for John to reply then.
>>>> > Thank you.
>>>> >
>>>> > Jinwoong Yoo
>>>> > UNM
>>>> >
>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>>>> > met_help at ucar.edu>
>>>> > wrote:
>>>> >
>>>> > > Hello. John is out of the office, but will be back later
this
>>>> week. I am
>>>> > > unable to help you with this task. Hopefully, John will be
able to
>>>> > respond
>>>> > > to this ticket later this week.
>>>> > >
>>>> > > Julie
>>>> > >
>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>>>> met_help at ucar.edu>
>>>> > > wrote:
>>>> > >
>>>> > > >
>>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>> > > >
>>>> > > > Dear John,
>>>> > > >
>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
>>>> > > > I assume you can access to my glade scratch directory.
>>>> > > > My UPP ouputs are located
>>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>>>> > > >
>>>> > > > Against those files, I'd like to compare CCSM4 simulation
outputs
>>>> some
>>>> > of
>>>> > > > which are located also under
>>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>>>> > > >
>>>> > > > Those are six hourly files all together. If you cannot
access to
>>>> those
>>>> > > > files, please let me know.
>>>> > > >
>>>> > > > However, there are some variables among the CCSM4 outputs
that are
>>>> > > > available only monthly or annually. In this case, how can I
>>>> > post-process
>>>> > > my
>>>> > > > WRF outputs for the usage of MET through UPP?
>>>> > > >
>>>> > > > Thank you.
>>>> > > >
>>>> > > > Regards,
>>>> > > >
>>>> > > > Jinwoong Yoo
>>>> > > > UNM
>>>> > > >
>>>> > > >
>>>> > > >
>>>> > > >
>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via RT
<
>>>> > > > met_help at ucar.edu> wrote:
>>>> > > >
>>>> > > > > Jinwoo,
>>>> > > > >
>>>> > > > > I believe UPP will still run, just using a single time
value as
>>>> if it
>>>> > > > were
>>>> > > > > an instantaneous field rather than an accumulation
interval.
>>>> The
>>>> > > default
>>>> > > > > for ARW is to store precip as a runtime accumulation.
For
>>>> forecasts
>>>> > > > hours
>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
>>>> accumulation
>>>> > > > from 0
>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored as
>>>> an
>>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference, but
>>>> > > something
>>>> > > > > to be aware of when you set up MET config files to verify
the
>>>> output.
>>>> > > > >
>>>> > > > > Give UPP a try and let me know how it goes.
>>>> > > > >
>>>> > > > > John
>>>> > > > >
>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>>>> > met_help at ucar.edu>
>>>> > > > > wrote:
>>>> > > > >
>>>> > > > > >
>>>> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>> >
>>>> > > > > >
>>>> > > > > > Dear John,
>>>> > > > > >
>>>> > > > > > I see. Let me change "export lastfhr=18" to "export
>>>> lastfhr=744".
>>>> > > > > > Thank you for your comments.
>>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my 10
year
>>>> > > simulation
>>>> > > > > was
>>>> > > > > > integrated from a single initialization. Then, will the
UPP
>>>> > > run_unipost
>>>> > > > > > crash after the 255 forecast hours or just will
precipitation
>>>> not
>>>> > be
>>>> > > > > stored
>>>> > > > > > any more aster the 255 forecast hours?
>>>> > > > > >
>>>> > > > > > Thank you.
>>>> > > > > >
>>>> > > > > > Regards,
>>>> > > > > >
>>>> > > > > > Jinwoong Yoo
>>>> > > > > > UNM
>>>> > > > > >
>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway via
RT <
>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>>>> > > > > >
>>>> > > > > > > Jinwoong,
>>>> > > > > > >
>>>> > > > > > > UPP support is typically provided through
wrf_help at ucar.edu
>>>> > > > > > <javascript:;>. I'll give it
>>>> > > > > > > a shot, but if this doesn't solve it, I'll refer you
to
>>>> wrf_help.
>>>> > > > > > >
>>>> > > > > > > I assume you're running the script named
"run_unipost".
>>>> Looking
>>>> > in
>>>> > > > > > there,
>>>> > > > > > > you'll see:
>>>> > > > > > >
>>>> > > > > > > export startdate=2005012300
>>>> > > > > > >
>>>> > > > > > > export fhr=00
>>>> > > > > > > export lastfhr=18
>>>> > > > > > > export incrementhr=03
>>>> > > > > > >
>>>> > > > > > > I'd suggest modifying the...
>>>> > > > > > > - "startdate" to be the model initialization time to
be
>>>> > processed.
>>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days *
24
>>>> > hours/day).
>>>> > > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
>>>> > > > > > >
>>>> > > > > > > Give that a shot and let me know if that gets you
output
>>>> past the
>>>> > > > first
>>>> > > > > > > day.
>>>> > > > > > >
>>>> > > > > > > Are you going to be evaluating precipitation? I ask
that
>>>> because
>>>> > > > > precip
>>>> > > > > > is
>>>> > > > > > > an accumulated interval, meaning that it has a
beginning
>>>> > > accumulation
>>>> > > > > > time
>>>> > > > > > > and ending accumulation time. GRIB1 is limited in
the
>>>> amount of
>>>> > > > space
>>>> > > > > > > available for storing times. When you get past 255
forecast
>>>> > hours,
>>>> > > > you
>>>> > > > > > can
>>>> > > > > > > no longer store beginning/ending times for precip
because
>>>> it runs
>>>> > > out
>>>> > > > > of
>>>> > > > > > > room in the header. Do you have a 10 year simulation
>>>> computed
>>>> > > from a
>>>> > > > > > > single initialization? Meaning, do you have forecast
lead
>>>> times
>>>> > > out
>>>> > > > to
>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)? I'm
>>>> wondering
>>>> > if
>>>> > > > > we'll
>>>> > > > > > > run into problems formatting those time strings.
>>>> > > > > > >
>>>> > > > > > > Thanks,
>>>> > > > > > > John
>>>> > > > > > >
>>>> > > > > > >
>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT
<
>>>> > > > > met_help at ucar.edu
>>>> > > > > > <javascript:;>>
>>>> > > > > > > wrote:
>>>> > > > > > >
>>>> > > > > > > >
>>>> > > > > > > > <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>> > >
>>>> > > > > > > >
>>>> > > > > > > > Dear John,
>>>> > > > > > > >
>>>> > > > > > > > I ran UPP with my multiple wrfout files, first with
the
>>>> first
>>>> > > month
>>>> > > > > of
>>>> > > > > > > the
>>>> > > > > > > > 10 year simulation. However, after executing the
>>>> run_unipost
>>>> > > > script,
>>>> > > > > > UPP
>>>> > > > > > > > produced only four files: unipost_d01.00.out,
>>>> > unipost_d01.06.out,
>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems that
>>>> UPP
>>>> > > > > processed
>>>> > > > > > > > only the first day's output among the one month
data.
>>>> > > > > > > > If you are familiar with UPP, how can I make sure
for UPP
>>>> to
>>>> > > > process
>>>> > > > > > > > multiple wrfout files more than one day?
>>>> > > > > > > > Thank you.
>>>> > > > > > > >
>>>> > > > > > > > Regards,
>>>> > > > > > > >
>>>> > > > > > > > Jinwoong Yoo
>>>> > > > > > > > UNM
>>>> > > > > > > >
>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>>>> > > > > jinwoong.yoo at gmail.com
>>>> > > > > > <javascript:;>>
>>>> > > > > > > > wrote:
>>>> > > > > > > >
>>>> > > > > > > > > Dear John,
>>>> > > > > > > > >
>>>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
>>>> > > > > > > > > Do I need to post process of the CCSM4 files also
to
>>>> run MET
>>>> > > > along
>>>> > > > > > with
>>>> > > > > > > > > the WRF-ARW files?
>>>> > > > > > > > > Thank you.
>>>> > > > > > > > >
>>>> > > > > > > > > Regards,
>>>> > > > > > > > >
>>>> > > > > > > > > Jinwoong Yoo
>>>> > > > > > > > > UNM
>>>> > > > > > > > >
>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>>>> > > > > > > > > wrote:
>>>> > > > > > > > >
>>>> > > > > > > > >> Dear John,
>>>> > > > > > > > >>
>>>> > > > > > > > >> Thank you for your kind email with the
direction.
>>>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
>>>> using
>>>> > UPP
>>>> > > > and
>>>> > > > > > let
>>>> > > > > > > > >> you know.
>>>> > > > > > > > >> Thank you very much.
>>>> > > > > > > > >>
>>>> > > > > > > > >> Regards,
>>>> > > > > > > > >>
>>>> > > > > > > > >> Jinwoong Yoo
>>>> > > > > > > > >> UNM
>>>> > > > > > > > >>
>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway
>>>> via RT <
>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>>>> > > > > > > > >>
>>>> > > > > > > > >>> Jinwoong,
>>>> > > > > > > > >>>
>>>> > > > > > > > >>> I see that you'd like some help running the
>>>> series-analysis
>>>> > > > tool
>>>> > > > > on
>>>> > > > > > > > your
>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is
that MET
>>>> can't
>>>> > > read
>>>> > > > > the
>>>> > > > > > > raw
>>>> > > > > > > > >>> wrfout files directly. Instead you first need
to post
>>>> > > process
>>>> > > > > them
>>>> > > > > > > > using
>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
>>>> utility
>>>> > > > > (produced
>>>> > > > > > > > >>> NetCDF
>>>> > > > > > > > >>> output). After you've post processed some
files,
>>>> please
>>>> > send
>>>> > > > us
>>>> > > > > > some
>>>> > > > > > > > >>> sample data by posting it to our anonymous ftp
site.
>>>> Also
>>>> > > > send a
>>>> > > > > > > > >>> description of the comparisons you'd like to
make
>>>> between
>>>> > > the 2
>>>> > > > > > > > datasets.
>>>> > > > > > > > >>>
>>>> > > > > > > > >>> Here's instructions for sending us data:
>>>> > > > > > > > >>>
>>>> http://www.dtcenter.org/met/users/support/met_help.php
>>>> > > > > > > > >>>
>>>> > > > > > > > >>> Thanks
>>>> > > > > > > > >>> John Halley Gotway
>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>>>> > > > > > > > >>>
>>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo
via RT
>>>> <
>>>> > > > > > > > met_help at ucar.edu <javascript:;>
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> wrote:
>>>> > > > > > > > >>>
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
>>>> upon.
>>>> > > > > > > > >>> > Transaction: Ticket created by
>>>> jinwoong.yoo at gmail.com
>>>> > > > > > <javascript:;>
>>>> > > > > > > <javascript:;>
>>>> > > > > > > > >>> > Queue: met_help
>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
>>>> outputs
>>>> > > > > > > > >>> > Owner: Nobody
>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
<javascript:;>
>>>> > > > > > <javascript:;>
>>>> > > > > > > > >>> > Status: new
>>>> > > > > > > > >>> > Ticket <URL:
>>>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Dear Met Help,
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Hi.
>>>> > > > > > > > >>> > I am looking for a way to verify WRF model
>>>> performance
>>>> > as a
>>>> > > > > > climate
>>>> > > > > > > > >>> model.
>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate model
>>>> > > > > simulations
>>>> > > > > > > > using
>>>> > > > > > > > >>> the
>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling)
>>>> for two
>>>> > > > > > different
>>>> > > > > > > > time
>>>> > > > > > > > >>> > periods: one for the LGM and the other for
the
>>>> modern as
>>>> > a
>>>> > > > > > control
>>>> > > > > > > > run.
>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly output
for ten
>>>> years
>>>> > > > > periods
>>>> > > > > > > > each.
>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6 hourly
>>>> outputs.
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>>>> > results,
>>>> > > > > > > verifying
>>>> > > > > > > > >>> the
>>>> > > > > > > > >>> > WRF model simulations.
>>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do
the
>>>> similar
>>>> > > jobs
>>>> > > > > but
>>>> > > > > > I
>>>> > > > > > > > need
>>>> > > > > > > > >>> > much more help from you than the MET Online
>>>> Tutorial to
>>>> > > make
>>>> > > > a
>>>> > > > > > step
>>>> > > > > > > > >>> > forward.
>>>> > > > > > > > >>> > Any help will be highly appreciated.
>>>> > > > > > > > >>> > Thank you.
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Regards,
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> > Jinwoong Yoo
>>>> > > > > > > > >>> > Postdoc
>>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>>>> > > > > > > > >>> > UNM
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>> >
>>>> > > > > > > > >>>
>>>> > > > > > > > >>>
>>>> > > > > > > > >>
>>>> > > > > > > > >
>>>> > > > > > > >
>>>> > > > > > > >
>>>> > > > > > >
>>>> > > > > > >
>>>> > > > > >
>>>> > > > > >
>>>> > > > >
>>>> > > > >
>>>> > > >
>>>> > > >
>>>> > >
>>>> > >
>>>> > > --
>>>> > > Julie Prestopnik
>>>> > > National Center for Atmospheric Research
>>>> > > Research Applications Laboratory
>>>> > > Phone: 303.497.8399
>>>> > > Email: jpresto at ucar.edu
>>>> > >
>>>> > >
>>>> >
>>>> >
>>>>
>>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: John Halley Gotway
Time: Fri Oct 03 12:17:04 2014
Jinwoong,
Great! Glad you were able to figure it out. I'll resolve this ticket
now. Just let us know if any more questions or issues arise in your
use of
MET.
Thanks,
John
On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> That's right. I was wrong earlier and the later arguments worked for
my
> regridding of the WRF output in Mercator to Lat-Lon grid.
> Thanks.
>
> Jinwoong Yoo
> UNM
>
> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
> > Am I misunderstanding the arguments of copygb?
> > (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
> > out.grb)
> >
> > Should be those numbers for the target grid in Lat/Lon of the
CCSM4 in my
> > case, not for the WRF output values?
> > Such that
> > *NX (*Number of points on latitude circle) : 192
> > *NY (*Number of points on longitude meridian*) : 288*
> > *LAT0 : -90*
> > *LON0 : 0*
> > *LAT1 : 90*
> > *LON1 : 358.75*
> > *DX : 942*
> > *DY : 1250*
> >
> > Please let me know.
> > Thank you.
> >
> > Jinwoong Yoo
> > UNM
> >
> > On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > wrote:
> >
> >> Dear John,
> >>
> >> Let me have a quick question about getting correct DX and DY in
copygb
> >> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY
64*" -x
> >> in.grb out.grb).
> >>
> >>
> >> As you may know, grid info in my UPP output is as below:
> >> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
> >> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315) scan
> >> 64 mode 136 Latin 0.000000 bdsgrid 1
> >>
> >> So my argument should be something like this:
> >> *copygb.exe \*
> >> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
> >>
> >> However, there are something unclear in the document (
> >>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> >> ).
> >>
> >> NX : Number of points on latitude circle
> >> NY : Number of points on longitude meridian
> >> DX : Longitudinal increment in millidegrees
> >> DY : Latitudinal increment in millidegrees
> >>
> >> You would notice that X,Y are not consistent as they refer to
lat/lon.
> >> I think flipping X/Y in NX/NY should work so that X represents
> longitude.
> >>
> >> That's not a big deal.
> >> Here my own problem goes.
> >>
> >> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
> >> However, if I convert them into millidegrees, DX is 323.515315
and DY is
> >> 287.23175 computationally supposing the earth is a complete
sphere.
> copygb
> >> takes only integers, DX is 323 and DY is 287. But, I'm not sure
if it's
> >> right to use the different DX and DY or if I should use the same
number
> for DX
> >> and DY.
> >> Please let me know.
> >>
> >> Thank you.
> >>
> >> Jinwoong Yoo
> >> UNM
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> >> wrote:
> >>
> >>> Dear John,
> >>>
> >>> Actually I was looking over copygb when I found your email
received.
> >>> It seems that I have to convert the UPP output again using
copygb.exe
> to
> >>> match up the grid of WRF results to the CCSM4's lat/lon grid.
> >>>
> >>> Isn't there any way to do that in one step? No?
> >>>
> >>> Along with the precipitation, I'd like to compare variables
(paired for
> >>> WRF and CCSM4) below for example:
> >>>
> >>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
> >>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
> >>>
> >>>
> >>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
> >>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
> >>>
> >>>
> >>> Varwrfclm = (/"T2", "TSLB","SMOIS",
> >>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
> >>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
> >>> "FSDS","SNOWICE","Q2M","RH2M"/)
> >>>
> >>> Variables in the lists are the actual name of variables in the
CCSM4
> >>> (including CAM and CLM monthly files) files and corresponding to
those
> in
> >>> the WRF-ARW.
> >>>
> >>>
> >>> I'm trying to run Grid-Stat for a test and I'd like to run
> >>> Series-Analysis for the entire dataset of the WRF and the CCSM4.
> >>> However, I'm concerning about how to run the Series-Analysis for
the
> >>> monthly or annual data using MET?
> >>> At the same time, I am also thinking about using the NCL to do
the same
> >>> statistical analyses. I don't know which one will be easier for
me to
> >>> execute for now.
> >>> We'll see.
> >>>
> >>> Will let you know as I make progress, John.
> >>>
> >>> >>> FYI - I just talked to the person who provides support for
UPP
> >>> through
> >>> >>> wrfhelp, and she's expecting an email from you.
> >>>
> >>> I cc'ed this email to wrfhelp.
> >>> Thank you.
> >>>
> >>> Regards,
> >>>
> >>> Jinwoong Yoo
> >>> UNM
> >>>
> >>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Jinwoong,
> >>>>
> >>>> I'm able to access your data on Yellowstone. I see that you'd
like to
> >>>> compare your UPP output in GRIB format to some data that
follows the
> >>>> NetCDF
> >>>> CF-1.0 convention.
> >>>>
> >>>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> >>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
> >>>>
> >>>> Hopefully you're able to look at those files and identify which
GRIB
> >>>> records you'd like to compare to which NetCDF variables. When
you run
> >>>> the
> >>>> MET tools, you'll define the forecast and observation fields to
be
> used.
> >>>>
> >>>> Looking at your data, I see 2 immediate problems:
> >>>>
> >>>> (1) Gridded forecast and observation data must be on the same
grid
> >>>> before
> >>>> MET can compare them.
> >>>> Here's a description from wgrib of the grid for your GRIB
data:
> >>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
> >>>> 349650
> >>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
> 315)
> >>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> >>>> But it looks like the CCSM4 data is on a lat/lon grid.
> >>>>
> >>>> I'd suggest using the copygb utility to regrid the GRIB1 output
from
> >>>> UPP to
> >>>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
> >>>>
> >>>>
> >>>>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> >>>>
> >>>> (2) There's an issue in the times in the accumulated
precipitation.
> Try
> >>>> running the following 2 commands:
> >>>> wgrib
> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252
> >>>> |
> >>>> grep APCP
> >>>> wgrib
> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258
> >>>> |
> >>>> grep APCP
> >>>>
> >>>> For the first, APCP is an accumulation of precip from hours 0
to 252.
> >>>> The
> >>>> second should be 0 to 258, but it's stored as 0 to 2! That's
because
> >>>> UPP
> >>>> is failing to realize that the precipitation timestamps are
> overflowing
> >>>> the
> >>>> 1 byte allocated for that second time unit. This problem was
> addressed
> >>>> in
> >>>> an earlier version of UPP but the problem is still showing up
in the
> >>>> version you're using. If you plan to verify precip, I'd
suggest
> writing
> >>>> wrfhelp at ucar.edu, tell them the version of UPP you're using,
and ask
> >>>> them
> >>>> if this should be fixed.
> >>>>
> >>>> Hope that helps.
> >>>>
> >>>> Thanks,
> >>>> John
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
> met_help at ucar.edu
> >>>> >
> >>>> wrote:
> >>>>
> >>>> >
> >>>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
> >>>> >
> >>>> > Hi, Julie.
> >>>> > Thank you for heads-up.
> >>>> > I will wait for John to reply then.
> >>>> > Thank you.
> >>>> >
> >>>> > Jinwoong Yoo
> >>>> > UNM
> >>>> >
> >>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
> >>>> > met_help at ucar.edu>
> >>>> > wrote:
> >>>> >
> >>>> > > Hello. John is out of the office, but will be back later
this
> >>>> week. I am
> >>>> > > unable to help you with this task. Hopefully, John will be
able
> to
> >>>> > respond
> >>>> > > to this ticket later this week.
> >>>> > >
> >>>> > > Julie
> >>>> > >
> >>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
> >>>> met_help at ucar.edu>
> >>>> > > wrote:
> >>>> > >
> >>>> > > >
> >>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >>>> > > >
> >>>> > > > Dear John,
> >>>> > > >
> >>>> > > > I ran UPP successfully on the UCAR Yellowstone.
> >>>> > > > I assume you can access to my glade scratch directory.
> >>>> > > > My UPP ouputs are located
> >>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> >>>> > > >
> >>>> > > > Against those files, I'd like to compare CCSM4 simulation
> outputs
> >>>> some
> >>>> > of
> >>>> > > > which are located also under
> >>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> >>>> > > >
> >>>> > > > Those are six hourly files all together. If you cannot
access to
> >>>> those
> >>>> > > > files, please let me know.
> >>>> > > >
> >>>> > > > However, there are some variables among the CCSM4 outputs
that
> are
> >>>> > > > available only monthly or annually. In this case, how can
I
> >>>> > post-process
> >>>> > > my
> >>>> > > > WRF outputs for the usage of MET through UPP?
> >>>> > > >
> >>>> > > > Thank you.
> >>>> > > >
> >>>> > > > Regards,
> >>>> > > >
> >>>> > > > Jinwoong Yoo
> >>>> > > > UNM
> >>>> > > >
> >>>> > > >
> >>>> > > >
> >>>> > > >
> >>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via
RT <
> >>>> > > > met_help at ucar.edu> wrote:
> >>>> > > >
> >>>> > > > > Jinwoo,
> >>>> > > > >
> >>>> > > > > I believe UPP will still run, just using a single time
value
> as
> >>>> if it
> >>>> > > > were
> >>>> > > > > an instantaneous field rather than an accumulation
interval.
> >>>> The
> >>>> > > default
> >>>> > > > > for ARW is to store precip as a runtime accumulation.
For
> >>>> forecasts
> >>>> > > > hours
> >>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
> >>>> accumulation
> >>>> > > > from 0
> >>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored
> as
> >>>> an
> >>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference,
> but
> >>>> > > something
> >>>> > > > > to be aware of when you set up MET config files to
verify the
> >>>> output.
> >>>> > > > >
> >>>> > > > > Give UPP a try and let me know how it goes.
> >>>> > > > >
> >>>> > > > > John
> >>>> > > > >
> >>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
> >>>> > met_help at ucar.edu>
> >>>> > > > > wrote:
> >>>> > > > >
> >>>> > > > > >
> >>>> > > > > > <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>> >
> >>>> > > > > >
> >>>> > > > > > Dear John,
> >>>> > > > > >
> >>>> > > > > > I see. Let me change "export lastfhr=18" to "export
> >>>> lastfhr=744".
> >>>> > > > > > Thank you for your comments.
> >>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my
10 year
> >>>> > > simulation
> >>>> > > > > was
> >>>> > > > > > integrated from a single initialization. Then, will
the UPP
> >>>> > > run_unipost
> >>>> > > > > > crash after the 255 forecast hours or just will
> precipitation
> >>>> not
> >>>> > be
> >>>> > > > > stored
> >>>> > > > > > any more aster the 255 forecast hours?
> >>>> > > > > >
> >>>> > > > > > Thank you.
> >>>> > > > > >
> >>>> > > > > > Regards,
> >>>> > > > > >
> >>>> > > > > > Jinwoong Yoo
> >>>> > > > > > UNM
> >>>> > > > > >
> >>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway
via RT <
> >>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
> >>>> > > > > >
> >>>> > > > > > > Jinwoong,
> >>>> > > > > > >
> >>>> > > > > > > UPP support is typically provided through
> wrf_help at ucar.edu
> >>>> > > > > > <javascript:;>. I'll give it
> >>>> > > > > > > a shot, but if this doesn't solve it, I'll refer
you to
> >>>> wrf_help.
> >>>> > > > > > >
> >>>> > > > > > > I assume you're running the script named
"run_unipost".
> >>>> Looking
> >>>> > in
> >>>> > > > > > there,
> >>>> > > > > > > you'll see:
> >>>> > > > > > >
> >>>> > > > > > > export startdate=2005012300
> >>>> > > > > > >
> >>>> > > > > > > export fhr=00
> >>>> > > > > > > export lastfhr=18
> >>>> > > > > > > export incrementhr=03
> >>>> > > > > > >
> >>>> > > > > > > I'd suggest modifying the...
> >>>> > > > > > > - "startdate" to be the model initialization time
to be
> >>>> > processed.
> >>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
> >>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days
* 24
> >>>> > hours/day).
> >>>> > > > > > > - Set "incrementhr" to 06 to process output every
6
> hours.
> >>>> > > > > > >
> >>>> > > > > > > Give that a shot and let me know if that gets you
output
> >>>> past the
> >>>> > > > first
> >>>> > > > > > > day.
> >>>> > > > > > >
> >>>> > > > > > > Are you going to be evaluating precipitation? I
ask that
> >>>> because
> >>>> > > > > precip
> >>>> > > > > > is
> >>>> > > > > > > an accumulated interval, meaning that it has a
beginning
> >>>> > > accumulation
> >>>> > > > > > time
> >>>> > > > > > > and ending accumulation time. GRIB1 is limited in
the
> >>>> amount of
> >>>> > > > space
> >>>> > > > > > > available for storing times. When you get past 255
> forecast
> >>>> > hours,
> >>>> > > > you
> >>>> > > > > > can
> >>>> > > > > > > no longer store beginning/ending times for precip
because
> >>>> it runs
> >>>> > > out
> >>>> > > > > of
> >>>> > > > > > > room in the header. Do you have a 10 year
simulation
> >>>> computed
> >>>> > > from a
> >>>> > > > > > > single initialization? Meaning, do you have
forecast lead
> >>>> times
> >>>> > > out
> >>>> > > > to
> >>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)?
I'm
> >>>> wondering
> >>>> > if
> >>>> > > > > we'll
> >>>> > > > > > > run into problems formatting those time strings.
> >>>> > > > > > >
> >>>> > > > > > > Thanks,
> >>>> > > > > > > John
> >>>> > > > > > >
> >>>> > > > > > >
> >>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via
RT <
> >>>> > > > > met_help at ucar.edu
> >>>> > > > > > <javascript:;>>
> >>>> > > > > > > wrote:
> >>>> > > > > > >
> >>>> > > > > > > >
> >>>> > > > > > > > <URL:
> >>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>> > >
> >>>> > > > > > > >
> >>>> > > > > > > > Dear John,
> >>>> > > > > > > >
> >>>> > > > > > > > I ran UPP with my multiple wrfout files, first
with the
> >>>> first
> >>>> > > month
> >>>> > > > > of
> >>>> > > > > > > the
> >>>> > > > > > > > 10 year simulation. However, after executing the
> >>>> run_unipost
> >>>> > > > script,
> >>>> > > > > > UPP
> >>>> > > > > > > > produced only four files: unipost_d01.00.out,
> >>>> > unipost_d01.06.out,
> >>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems
> that
> >>>> UPP
> >>>> > > > > processed
> >>>> > > > > > > > only the first day's output among the one month
data.
> >>>> > > > > > > > If you are familiar with UPP, how can I make sure
for
> UPP
> >>>> to
> >>>> > > > process
> >>>> > > > > > > > multiple wrfout files more than one day?
> >>>> > > > > > > > Thank you.
> >>>> > > > > > > >
> >>>> > > > > > > > Regards,
> >>>> > > > > > > >
> >>>> > > > > > > > Jinwoong Yoo
> >>>> > > > > > > > UNM
> >>>> > > > > > > >
> >>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
> >>>> > > > > jinwoong.yoo at gmail.com
> >>>> > > > > > <javascript:;>>
> >>>> > > > > > > > wrote:
> >>>> > > > > > > >
> >>>> > > > > > > > > Dear John,
> >>>> > > > > > > > >
> >>>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
> format?
> >>>> > > > > > > > > Do I need to post process of the CCSM4 files
also to
> >>>> run MET
> >>>> > > > along
> >>>> > > > > > with
> >>>> > > > > > > > > the WRF-ARW files?
> >>>> > > > > > > > > Thank you.
> >>>> > > > > > > > >
> >>>> > > > > > > > > Regards,
> >>>> > > > > > > > >
> >>>> > > > > > > > > Jinwoong Yoo
> >>>> > > > > > > > > UNM
> >>>> > > > > > > > >
> >>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
> >>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
> >>>> > > > > > > > > wrote:
> >>>> > > > > > > > >
> >>>> > > > > > > > >> Dear John,
> >>>> > > > > > > > >>
> >>>> > > > > > > > >> Thank you for your kind email with the
direction.
> >>>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
> >>>> using
> >>>> > UPP
> >>>> > > > and
> >>>> > > > > > let
> >>>> > > > > > > > >> you know.
> >>>> > > > > > > > >> Thank you very much.
> >>>> > > > > > > > >>
> >>>> > > > > > > > >> Regards,
> >>>> > > > > > > > >>
> >>>> > > > > > > > >> Jinwoong Yoo
> >>>> > > > > > > > >> UNM
> >>>> > > > > > > > >>
> >>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway
> >>>> via RT <
> >>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> >>>> > > > > > > > >>
> >>>> > > > > > > > >>> Jinwoong,
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>> I see that you'd like some help running the
> >>>> series-analysis
> >>>> > > > tool
> >>>> > > > > on
> >>>> > > > > > > > your
> >>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is
that MET
> >>>> can't
> >>>> > > read
> >>>> > > > > the
> >>>> > > > > > > raw
> >>>> > > > > > > > >>> wrfout files directly. Instead you first
need to
> post
> >>>> > > process
> >>>> > > > > them
> >>>> > > > > > > > using
> >>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
> >>>> utility
> >>>> > > > > (produced
> >>>> > > > > > > > >>> NetCDF
> >>>> > > > > > > > >>> output). After you've post processed some
files,
> >>>> please
> >>>> > send
> >>>> > > > us
> >>>> > > > > > some
> >>>> > > > > > > > >>> sample data by posting it to our anonymous
ftp site.
> >>>> Also
> >>>> > > > send a
> >>>> > > > > > > > >>> description of the comparisons you'd like to
make
> >>>> between
> >>>> > > the 2
> >>>> > > > > > > > datasets.
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>> Here's instructions for sending us data:
> >>>> > > > > > > > >>>
> >>>> http://www.dtcenter.org/met/users/support/met_help.php
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>> Thanks
> >>>> > > > > > > > >>> John Halley Gotway
> >>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong
Yoo via
> RT
> >>>> <
> >>>> > > > > > > > met_help at ucar.edu <javascript:;>
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> wrote:
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
> >>>> upon.
> >>>> > > > > > > > >>> > Transaction: Ticket created by
> >>>> jinwoong.yoo at gmail.com
> >>>> > > > > > <javascript:;>
> >>>> > > > > > > <javascript:;>
> >>>> > > > > > > > >>> > Queue: met_help
> >>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
> >>>> outputs
> >>>> > > > > > > > >>> > Owner: Nobody
> >>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
> <javascript:;>
> >>>> > > > > > <javascript:;>
> >>>> > > > > > > > >>> > Status: new
> >>>> > > > > > > > >>> > Ticket <URL:
> >>>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Dear Met Help,
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Hi.
> >>>> > > > > > > > >>> > I am looking for a way to verify WRF model
> >>>> performance
> >>>> > as a
> >>>> > > > > > climate
> >>>> > > > > > > > >>> model.
> >>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate
> model
> >>>> > > > > simulations
> >>>> > > > > > > > using
> >>>> > > > > > > > >>> the
> >>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling)
> >>>> for two
> >>>> > > > > > different
> >>>> > > > > > > > time
> >>>> > > > > > > > >>> > periods: one for the LGM and the other for
the
> >>>> modern as
> >>>> > a
> >>>> > > > > > control
> >>>> > > > > > > > run.
> >>>> > > > > > > > >>> > The WRF simulation produced 6 hourly output
for
> ten
> >>>> years
> >>>> > > > > periods
> >>>> > > > > > > > each.
> >>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
> >>>> outputs.
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
> >>>> > results,
> >>>> > > > > > > verifying
> >>>> > > > > > > > >>> the
> >>>> > > > > > > > >>> > WRF model simulations.
> >>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
> >>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do
the
> >>>> similar
> >>>> > > jobs
> >>>> > > > > but
> >>>> > > > > > I
> >>>> > > > > > > > need
> >>>> > > > > > > > >>> > much more help from you than the MET Online
> >>>> Tutorial to
> >>>> > > make
> >>>> > > > a
> >>>> > > > > > step
> >>>> > > > > > > > >>> > forward.
> >>>> > > > > > > > >>> > Any help will be highly appreciated.
> >>>> > > > > > > > >>> > Thank you.
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Regards,
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> > Jinwoong Yoo
> >>>> > > > > > > > >>> > Postdoc
> >>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
> >>>> > > > > > > > >>> > UNM
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>> >
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>>
> >>>> > > > > > > > >>
> >>>> > > > > > > > >
> >>>> > > > > > > >
> >>>> > > > > > > >
> >>>> > > > > > >
> >>>> > > > > > >
> >>>> > > > > >
> >>>> > > > > >
> >>>> > > > >
> >>>> > > > >
> >>>> > > >
> >>>> > > >
> >>>> > >
> >>>> > >
> >>>> > > --
> >>>> > > Julie Prestopnik
> >>>> > > National Center for Atmospheric Research
> >>>> > > Research Applications Laboratory
> >>>> > > Phone: 303.497.8399
> >>>> > > Email: jpresto at ucar.edu
> >>>> > >
> >>>> > >
> >>>> >
> >>>> >
> >>>>
> >>>>
> >>>
> >>
> >
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Mon Oct 06 13:45:26 2014
Hi.
I encounter errors while I'm running Grid_Stat on Yellowstone as
below:
[jyoo at yslogin4 analysis]$
/glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm -outdir
/glade/scratch/jyoo/domain/mean/analysis/gridstatout
DEBUG 1: Default Config File:
/glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
DEBUG 1: User Config File: GridStatConfig_ccsm
WARNING:
WARNING: NcCfFile::open() -> could not extract init time from file
name
WARNING: Using init time of 0
WARNING:
ERROR :
ERROR : process_command_line() -> The forecast and observation grids
do
not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000 lon_ll:
0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx: 288
Ny:
191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
WRF output that was postprocessed using UPP and copygb is compared
with
CCSM4 result in NetCDF format here for a test.
First,
WARNING: NcCfFile::open() -> could not extract init time from file
name
WARNING: Using init time of 0
Since the CCSM4 file contains 6 hourly data for a month, I set " level
= [ "(1,*,*)" ];"
for my desired level (surface). But it seems it didn't work. Setting
"(*,*)" didn't work, neither.
fcst = {
wind_thresh = [ NA ];
field = [
{
name = "CPRAT";
level = [ "R277" ];
cat_thresh = [ >0.0, >=5.0 ];
}
];
};
obs = {
wind_thresh = [ NA ];
field = [
{
name = "PRECC";
level = [ "(1,*,*)" ];
cat_thresh = [ >0.0, >=5.0 ];
}
];
};
Second, obviously it seems my previous copygb was not correct.
ERROR : process_command_line() -> The forecast and observation grids
do
not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000 lon_ll:
0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx: 288
Ny:
191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
My copygy commend was like this:
*/glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
* -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250 64" \*
* /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
Output grid information after executing the copygb looks like this:
latlon: lat -90.000000 to 90.000000 by 1.250000 nxny 55296
long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan 64 mode
128
bdsgrid 1
However, there are something strange.
According to what Grid-Stat reads in for the CCSM4 file (Projection:
Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942
delta_lon: 1.250), lat and lon information were flipped compared to
what I
get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ; slat = 191
;slon
= 288) and the Ny: 191 matches to slat(staggered latitude) not
latitude.
One thing to note is that my WRF simulation is a tropical channel
simulation covering 45S to 45N while the CCSM4 is global.
What am I doing wrong with copygb and GridStatConfig file?
Thank you in advance.
Regards,
Jinwoong Yoo
UNM
On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> That's right. I was wrong earlier and the later arguments worked for
my
> regridding of the WRF output in Mercator to Lat-Lon grid.
> Thanks.
>
> Jinwoong Yoo
> UNM
>
> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> Am I misunderstanding the arguments of copygb?
>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
>> out.grb)
>>
>> Should be those numbers for the target grid in Lat/Lon of the CCSM4
in my
>> case, not for the WRF output values?
>> Such that
>> *NX (*Number of points on latitude circle) : 192
>> *NY (*Number of points on longitude meridian*) : 288*
>> *LAT0 : -90*
>> *LON0 : 0*
>> *LAT1 : 90*
>> *LON1 : 358.75*
>> *DX : 942*
>> *DY : 1250*
>>
>> Please let me know.
>> Thank you.
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>> wrote:
>>
>>> Dear John,
>>>
>>> Let me have a quick question about getting correct DX and DY in
copygb
>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY
64*" -x
>>> in.grb out.grb).
>>>
>>>
>>> As you may know, grid info in my UPP output is as below:
>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
>>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x 315)
scan
>>> 64 mode 136 Latin 0.000000 bdsgrid 1
>>>
>>> So my argument should be something like this:
>>> *copygb.exe \*
>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
>>>
>>> However, there are something unclear in the document (
>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>> ).
>>>
>>> NX : Number of points on latitude circle
>>> NY : Number of points on longitude meridian
>>> DX : Longitudinal increment in millidegrees
>>> DY : Latitudinal increment in millidegrees
>>>
>>> You would notice that X,Y are not consistent as they refer to
lat/lon.
>>> I think flipping X/Y in NX/NY should work so that X represents
longitude.
>>>
>>> That's not a big deal.
>>> Here my own problem goes.
>>>
>>> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
>>> However, if I convert them into millidegrees, DX is 323.515315 and
DY is
>>> 287.23175 computationally supposing the earth is a complete
sphere. copygb
>>> takes only integers, DX is 323 and DY is 287. But, I'm not sure if
it's
>>> right to use the different DX and DY or if I should use the same
number for DX
>>> and DY.
>>> Please let me know.
>>>
>>> Thank you.
>>>
>>> Jinwoong Yoo
>>> UNM
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>> wrote:
>>>
>>>> Dear John,
>>>>
>>>> Actually I was looking over copygb when I found your email
received.
>>>> It seems that I have to convert the UPP output again using
copygb.exe
>>>> to match up the grid of WRF results to the CCSM4's lat/lon grid.
>>>>
>>>> Isn't there any way to do that in one step? No?
>>>>
>>>> Along with the precipitation, I'd like to compare variables
(paired for
>>>> WRF and CCSM4) below for example:
>>>>
>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>>>>
>>>>
>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>>>>
>>>>
>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
>>>>
>>>> Variables in the lists are the actual name of variables in the
CCSM4
>>>> (including CAM and CLM monthly files) files and corresponding to
those in
>>>> the WRF-ARW.
>>>>
>>>>
>>>> I'm trying to run Grid-Stat for a test and I'd like to run
>>>> Series-Analysis for the entire dataset of the WRF and the CCSM4.
>>>> However, I'm concerning about how to run the Series-Analysis for
the
>>>> monthly or annual data using MET?
>>>> At the same time, I am also thinking about using the NCL to do
the same
>>>> statistical analyses. I don't know which one will be easier for
me to
>>>> execute for now.
>>>> We'll see.
>>>>
>>>> Will let you know as I make progress, John.
>>>>
>>>> >>> FYI - I just talked to the person who provides support for
UPP
>>>> through
>>>> >>> wrfhelp, and she's expecting an email from you.
>>>>
>>>> I cc'ed this email to wrfhelp.
>>>> Thank you.
>>>>
>>>> Regards,
>>>>
>>>> Jinwoong Yoo
>>>> UNM
>>>>
>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
>>>> met_help at ucar.edu> wrote:
>>>>
>>>>> Jinwoong,
>>>>>
>>>>> I'm able to access your data on Yellowstone. I see that you'd
like to
>>>>> compare your UPP output in GRIB format to some data that follows
the
>>>>> NetCDF
>>>>> CF-1.0 convention.
>>>>>
>>>>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>>>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>>>>
>>>>> Hopefully you're able to look at those files and identify which
GRIB
>>>>> records you'd like to compare to which NetCDF variables. When
you run
>>>>> the
>>>>> MET tools, you'll define the forecast and observation fields to
be
>>>>> used.
>>>>>
>>>>> Looking at your data, I see 2 immediate problems:
>>>>>
>>>>> (1) Gridded forecast and observation data must be on the same
grid
>>>>> before
>>>>> MET can compare them.
>>>>> Here's a description from wgrib of the grid for your GRIB
data:
>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
>>>>> 349650
>>>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
>>>>> 315)
>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>>>> But it looks like the CCSM4 data is on a lat/lon grid.
>>>>>
>>>>> I'd suggest using the copygb utility to regrid the GRIB1 output
from
>>>>> UPP to
>>>>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>>>>>
>>>>>
>>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>>>
>>>>> (2) There's an issue in the times in the accumulated
precipitation.
>>>>> Try
>>>>> running the following 2 commands:
>>>>> wgrib
>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
>>>>> grep APCP
>>>>> wgrib
>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
>>>>> grep APCP
>>>>>
>>>>> For the first, APCP is an accumulation of precip from hours 0 to
252.
>>>>> The
>>>>> second should be 0 to 258, but it's stored as 0 to 2! That's
because
>>>>> UPP
>>>>> is failing to realize that the precipitation timestamps are
>>>>> overflowing the
>>>>> 1 byte allocated for that second time unit. This problem was
>>>>> addressed in
>>>>> an earlier version of UPP but the problem is still showing up in
the
>>>>> version you're using. If you plan to verify precip, I'd suggest
>>>>> writing
>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're using, and
ask
>>>>> them
>>>>> if this should be fixed.
>>>>>
>>>>> Hope that helps.
>>>>>
>>>>> Thanks,
>>>>> John
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
>>>>> met_help at ucar.edu>
>>>>> wrote:
>>>>>
>>>>> >
>>>>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
>>>>> >
>>>>> > Hi, Julie.
>>>>> > Thank you for heads-up.
>>>>> > I will wait for John to reply then.
>>>>> > Thank you.
>>>>> >
>>>>> > Jinwoong Yoo
>>>>> > UNM
>>>>> >
>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>>>>> > met_help at ucar.edu>
>>>>> > wrote:
>>>>> >
>>>>> > > Hello. John is out of the office, but will be back later
this
>>>>> week. I am
>>>>> > > unable to help you with this task. Hopefully, John will be
able to
>>>>> > respond
>>>>> > > to this ticket later this week.
>>>>> > >
>>>>> > > Julie
>>>>> > >
>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>>>>> met_help at ucar.edu>
>>>>> > > wrote:
>>>>> > >
>>>>> > > >
>>>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>> > > >
>>>>> > > > Dear John,
>>>>> > > >
>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
>>>>> > > > I assume you can access to my glade scratch directory.
>>>>> > > > My UPP ouputs are located
>>>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>>>>> > > >
>>>>> > > > Against those files, I'd like to compare CCSM4 simulation
>>>>> outputs some
>>>>> > of
>>>>> > > > which are located also under
>>>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>>>>> > > >
>>>>> > > > Those are six hourly files all together. If you cannot
access to
>>>>> those
>>>>> > > > files, please let me know.
>>>>> > > >
>>>>> > > > However, there are some variables among the CCSM4 outputs
that
>>>>> are
>>>>> > > > available only monthly or annually. In this case, how can
I
>>>>> > post-process
>>>>> > > my
>>>>> > > > WRF outputs for the usage of MET through UPP?
>>>>> > > >
>>>>> > > > Thank you.
>>>>> > > >
>>>>> > > > Regards,
>>>>> > > >
>>>>> > > > Jinwoong Yoo
>>>>> > > > UNM
>>>>> > > >
>>>>> > > >
>>>>> > > >
>>>>> > > >
>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via
RT <
>>>>> > > > met_help at ucar.edu> wrote:
>>>>> > > >
>>>>> > > > > Jinwoo,
>>>>> > > > >
>>>>> > > > > I believe UPP will still run, just using a single time
value
>>>>> as if it
>>>>> > > > were
>>>>> > > > > an instantaneous field rather than an accumulation
interval.
>>>>> The
>>>>> > > default
>>>>> > > > > for ARW is to store precip as a runtime accumulation.
For
>>>>> forecasts
>>>>> > > > hours
>>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
>>>>> accumulation
>>>>> > > > from 0
>>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored
>>>>> as an
>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference, but
>>>>> > > something
>>>>> > > > > to be aware of when you set up MET config files to
verify the
>>>>> output.
>>>>> > > > >
>>>>> > > > > Give UPP a try and let me know how it goes.
>>>>> > > > >
>>>>> > > > > John
>>>>> > > > >
>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>>>>> > met_help at ucar.edu>
>>>>> > > > > wrote:
>>>>> > > > >
>>>>> > > > > >
>>>>> > > > > > <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>> > > > > >
>>>>> > > > > > Dear John,
>>>>> > > > > >
>>>>> > > > > > I see. Let me change "export lastfhr=18" to "export
>>>>> lastfhr=744".
>>>>> > > > > > Thank you for your comments.
>>>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my
10 year
>>>>> > > simulation
>>>>> > > > > was
>>>>> > > > > > integrated from a single initialization. Then, will
the UPP
>>>>> > > run_unipost
>>>>> > > > > > crash after the 255 forecast hours or just will
>>>>> precipitation not
>>>>> > be
>>>>> > > > > stored
>>>>> > > > > > any more aster the 255 forecast hours?
>>>>> > > > > >
>>>>> > > > > > Thank you.
>>>>> > > > > >
>>>>> > > > > > Regards,
>>>>> > > > > >
>>>>> > > > > > Jinwoong Yoo
>>>>> > > > > > UNM
>>>>> > > > > >
>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway
via RT <
>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>>>>> > > > > >
>>>>> > > > > > > Jinwoong,
>>>>> > > > > > >
>>>>> > > > > > > UPP support is typically provided through
>>>>> wrf_help at ucar.edu
>>>>> > > > > > <javascript:;>. I'll give it
>>>>> > > > > > > a shot, but if this doesn't solve it, I'll refer you
to
>>>>> wrf_help.
>>>>> > > > > > >
>>>>> > > > > > > I assume you're running the script named
"run_unipost".
>>>>> Looking
>>>>> > in
>>>>> > > > > > there,
>>>>> > > > > > > you'll see:
>>>>> > > > > > >
>>>>> > > > > > > export startdate=2005012300
>>>>> > > > > > >
>>>>> > > > > > > export fhr=00
>>>>> > > > > > > export lastfhr=18
>>>>> > > > > > > export incrementhr=03
>>>>> > > > > > >
>>>>> > > > > > > I'd suggest modifying the...
>>>>> > > > > > > - "startdate" to be the model initialization time
to be
>>>>> > processed.
>>>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days *
24
>>>>> > hours/day).
>>>>> > > > > > > - Set "incrementhr" to 06 to process output every 6
hours.
>>>>> > > > > > >
>>>>> > > > > > > Give that a shot and let me know if that gets you
output
>>>>> past the
>>>>> > > > first
>>>>> > > > > > > day.
>>>>> > > > > > >
>>>>> > > > > > > Are you going to be evaluating precipitation? I ask
that
>>>>> because
>>>>> > > > > precip
>>>>> > > > > > is
>>>>> > > > > > > an accumulated interval, meaning that it has a
beginning
>>>>> > > accumulation
>>>>> > > > > > time
>>>>> > > > > > > and ending accumulation time. GRIB1 is limited in
the
>>>>> amount of
>>>>> > > > space
>>>>> > > > > > > available for storing times. When you get past 255
>>>>> forecast
>>>>> > hours,
>>>>> > > > you
>>>>> > > > > > can
>>>>> > > > > > > no longer store beginning/ending times for precip
because
>>>>> it runs
>>>>> > > out
>>>>> > > > > of
>>>>> > > > > > > room in the header. Do you have a 10 year
simulation
>>>>> computed
>>>>> > > from a
>>>>> > > > > > > single initialization? Meaning, do you have
forecast lead
>>>>> times
>>>>> > > out
>>>>> > > > to
>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)?
I'm
>>>>> wondering
>>>>> > if
>>>>> > > > > we'll
>>>>> > > > > > > run into problems formatting those time strings.
>>>>> > > > > > >
>>>>> > > > > > > Thanks,
>>>>> > > > > > > John
>>>>> > > > > > >
>>>>> > > > > > >
>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via RT
<
>>>>> > > > > met_help at ucar.edu
>>>>> > > > > > <javascript:;>>
>>>>> > > > > > > wrote:
>>>>> > > > > > >
>>>>> > > > > > > >
>>>>> > > > > > > > <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>> > >
>>>>> > > > > > > >
>>>>> > > > > > > > Dear John,
>>>>> > > > > > > >
>>>>> > > > > > > > I ran UPP with my multiple wrfout files, first
with the
>>>>> first
>>>>> > > month
>>>>> > > > > of
>>>>> > > > > > > the
>>>>> > > > > > > > 10 year simulation. However, after executing the
>>>>> run_unipost
>>>>> > > > script,
>>>>> > > > > > UPP
>>>>> > > > > > > > produced only four files: unipost_d01.00.out,
>>>>> > unipost_d01.06.out,
>>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems
>>>>> that UPP
>>>>> > > > > processed
>>>>> > > > > > > > only the first day's output among the one month
data.
>>>>> > > > > > > > If you are familiar with UPP, how can I make sure
for
>>>>> UPP to
>>>>> > > > process
>>>>> > > > > > > > multiple wrfout files more than one day?
>>>>> > > > > > > > Thank you.
>>>>> > > > > > > >
>>>>> > > > > > > > Regards,
>>>>> > > > > > > >
>>>>> > > > > > > > Jinwoong Yoo
>>>>> > > > > > > > UNM
>>>>> > > > > > > >
>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>>>>> > > > > jinwoong.yoo at gmail.com
>>>>> > > > > > <javascript:;>>
>>>>> > > > > > > > wrote:
>>>>> > > > > > > >
>>>>> > > > > > > > > Dear John,
>>>>> > > > > > > > >
>>>>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
format?
>>>>> > > > > > > > > Do I need to post process of the CCSM4 files
also to
>>>>> run MET
>>>>> > > > along
>>>>> > > > > > with
>>>>> > > > > > > > > the WRF-ARW files?
>>>>> > > > > > > > > Thank you.
>>>>> > > > > > > > >
>>>>> > > > > > > > > Regards,
>>>>> > > > > > > > >
>>>>> > > > > > > > > Jinwoong Yoo
>>>>> > > > > > > > > UNM
>>>>> > > > > > > > >
>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>>>>> > > > > > > > > wrote:
>>>>> > > > > > > > >
>>>>> > > > > > > > >> Dear John,
>>>>> > > > > > > > >>
>>>>> > > > > > > > >> Thank you for your kind email with the
direction.
>>>>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
>>>>> using
>>>>> > UPP
>>>>> > > > and
>>>>> > > > > > let
>>>>> > > > > > > > >> you know.
>>>>> > > > > > > > >> Thank you very much.
>>>>> > > > > > > > >>
>>>>> > > > > > > > >> Regards,
>>>>> > > > > > > > >>
>>>>> > > > > > > > >> Jinwoong Yoo
>>>>> > > > > > > > >> UNM
>>>>> > > > > > > > >>
>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway
>>>>> via RT <
>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>>>>> > > > > > > > >>
>>>>> > > > > > > > >>> Jinwoong,
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>> I see that you'd like some help running the
>>>>> series-analysis
>>>>> > > > tool
>>>>> > > > > on
>>>>> > > > > > > > your
>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is
that MET
>>>>> can't
>>>>> > > read
>>>>> > > > > the
>>>>> > > > > > > raw
>>>>> > > > > > > > >>> wrfout files directly. Instead you first need
to
>>>>> post
>>>>> > > process
>>>>> > > > > them
>>>>> > > > > > > > using
>>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
>>>>> utility
>>>>> > > > > (produced
>>>>> > > > > > > > >>> NetCDF
>>>>> > > > > > > > >>> output). After you've post processed some
files,
>>>>> please
>>>>> > send
>>>>> > > > us
>>>>> > > > > > some
>>>>> > > > > > > > >>> sample data by posting it to our anonymous ftp
>>>>> site. Also
>>>>> > > > send a
>>>>> > > > > > > > >>> description of the comparisons you'd like to
make
>>>>> between
>>>>> > > the 2
>>>>> > > > > > > > datasets.
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>> Here's instructions for sending us data:
>>>>> > > > > > > > >>>
>>>>> http://www.dtcenter.org/met/users/support/met_help.php
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>> Thanks
>>>>> > > > > > > > >>> John Halley Gotway
>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong Yoo
via
>>>>> RT <
>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> wrote:
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
>>>>> upon.
>>>>> > > > > > > > >>> > Transaction: Ticket created by
>>>>> jinwoong.yoo at gmail.com
>>>>> > > > > > <javascript:;>
>>>>> > > > > > > <javascript:;>
>>>>> > > > > > > > >>> > Queue: met_help
>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
>>>>> outputs
>>>>> > > > > > > > >>> > Owner: Nobody
>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
>>>>> <javascript:;>
>>>>> > > > > > <javascript:;>
>>>>> > > > > > > > >>> > Status: new
>>>>> > > > > > > > >>> > Ticket <URL:
>>>>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Dear Met Help,
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Hi.
>>>>> > > > > > > > >>> > I am looking for a way to verify WRF model
>>>>> performance
>>>>> > as a
>>>>> > > > > > climate
>>>>> > > > > > > > >>> model.
>>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate
>>>>> model
>>>>> > > > > simulations
>>>>> > > > > > > > using
>>>>> > > > > > > > >>> the
>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling)
>>>>> for two
>>>>> > > > > > different
>>>>> > > > > > > > time
>>>>> > > > > > > > >>> > periods: one for the LGM and the other for
the
>>>>> modern as
>>>>> > a
>>>>> > > > > > control
>>>>> > > > > > > > run.
>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly output
for
>>>>> ten years
>>>>> > > > > periods
>>>>> > > > > > > > each.
>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
>>>>> outputs.
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>>>>> > results,
>>>>> > > > > > > verifying
>>>>> > > > > > > > >>> the
>>>>> > > > > > > > >>> > WRF model simulations.
>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do
the
>>>>> similar
>>>>> > > jobs
>>>>> > > > > but
>>>>> > > > > > I
>>>>> > > > > > > > need
>>>>> > > > > > > > >>> > much more help from you than the MET Online
>>>>> Tutorial to
>>>>> > > make
>>>>> > > > a
>>>>> > > > > > step
>>>>> > > > > > > > >>> > forward.
>>>>> > > > > > > > >>> > Any help will be highly appreciated.
>>>>> > > > > > > > >>> > Thank you.
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Regards,
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> > Jinwoong Yoo
>>>>> > > > > > > > >>> > Postdoc
>>>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>>>>> > > > > > > > >>> > UNM
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>> >
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>>
>>>>> > > > > > > > >>
>>>>> > > > > > > > >
>>>>> > > > > > > >
>>>>> > > > > > > >
>>>>> > > > > > >
>>>>> > > > > > >
>>>>> > > > > >
>>>>> > > > > >
>>>>> > > > >
>>>>> > > > >
>>>>> > > >
>>>>> > > >
>>>>> > >
>>>>> > >
>>>>> > > --
>>>>> > > Julie Prestopnik
>>>>> > > National Center for Atmospheric Research
>>>>> > > Research Applications Laboratory
>>>>> > > Phone: 303.497.8399
>>>>> > > Email: jpresto at ucar.edu
>>>>> > >
>>>>> > >
>>>>> >
>>>>> >
>>>>>
>>>>>
>>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Mon Oct 06 16:42:30 2014
I'm getting closer to the point of the problem.
I have corrected my copygb argument so close as to get this minor(?)
error:
ERROR : process_command_line() -> The forecast and observation grids
do
not match:
Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: -0.625
delta_lat: 0.942 delta_lon: 1.250 !=
Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625
delta_lat:
0.942 delta_lon: 1.250
My copygb argument was:
copygb.exe -g"255 0 288 191 -89529 625 128 89529 359375 1250 942
64" -x
$infile $outfile
As you may know, 625 in this case is Longitude of origin * 1000 (lon =
-180
... 180). But I don't know how the negative sign got into the
execution of
the program, yielding lon_ll: -0.625.
Do you know how I should solve this problem?
Is this a bug in the program?
By the way, as I noted in my previous email,
I found and confirmed the fault statements in the MET Online Tutorial
for
the copygb Tool: Lat/Lon Grid (
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
)
NX should be Number of points on longitude meridian,
while NY should be Number of points on latitude circle, definitely.
Thanks.
Jinwoong Yoo
UNM
On Mon, Oct 6, 2014 at 1:45 PM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> Hi.
>
> I encounter errors while I'm running Grid_Stat on Yellowstone as
below:
>
> [jyoo at yslogin4 analysis]$
> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
> b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm -outdir
> /glade/scratch/jyoo/domain/mean/analysis/gridstatout
> DEBUG 1: Default Config File:
> /glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
> DEBUG 1: User Config File: GridStatConfig_ccsm
> WARNING:
> WARNING: NcCfFile::open() -> could not extract init time from file
name
> WARNING: Using init time of 0
> WARNING:
> ERROR :
> ERROR : process_command_line() -> The forecast and observation
grids do
> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx:
288 Ny:
> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
>
> WRF output that was postprocessed using UPP and copygb is compared
with
> CCSM4 result in NetCDF format here for a test.
>
> First,
> WARNING: NcCfFile::open() -> could not extract init time from file
name
> WARNING: Using init time of 0
>
> Since the CCSM4 file contains 6 hourly data for a month, I set "
level
> = [ "(1,*,*)" ];"
> for my desired level (surface). But it seems it didn't work.
Setting
> "(*,*)" didn't work, neither.
>
> fcst = {
> wind_thresh = [ NA ];
>
> field = [
> {
> name = "CPRAT";
> level = [ "R277" ];
> cat_thresh = [ >0.0, >=5.0 ];
> }
> ];
>
> };
>
> obs = {
> wind_thresh = [ NA ];
>
> field = [
> {
> name = "PRECC";
> level = [ "(1,*,*)" ];
> cat_thresh = [ >0.0, >=5.0 ];
> }
> ];
>
> };
>
> Second, obviously it seems my previous copygb was not correct.
> ERROR : process_command_line() -> The forecast and observation
grids do
> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx:
288 Ny:
> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
>
>
> My copygy commend was like this:
>
> */glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
> * -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250 64" \*
>
> * /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
>
> Output grid information after executing the copygb looks like this:
> latlon: lat -90.000000 to 90.000000 by 1.250000 nxny 55296
> long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan 64 mode
128
> bdsgrid 1
>
> However, there are something strange.
> According to what Grid-Stat reads in for the CCSM4 file (Projection:
> Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat:
0.942
> delta_lon: 1.250), lat and lon information were flipped compared to
what I
> get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ; slat =
191 ;slon
> = 288) and the Ny: 191 matches to slat(staggered latitude) not
latitude.
>
> One thing to note is that my WRF simulation is a tropical channel
> simulation covering 45S to 45N while the CCSM4 is global.
>
> What am I doing wrong with copygb and GridStatConfig file?
> Thank you in advance.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> That's right. I was wrong earlier and the later arguments worked
for my
>> regridding of the WRF output in Mercator to Lat-Lon grid.
>> Thanks.
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>> wrote:
>>
>>> Am I misunderstanding the arguments of copygb?
>>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
>>> out.grb)
>>>
>>> Should be those numbers for the target grid in Lat/Lon of the
CCSM4 in
>>> my case, not for the WRF output values?
>>> Such that
>>> *NX (*Number of points on latitude circle) : 192
>>> *NY (*Number of points on longitude meridian*) : 288*
>>> *LAT0 : -90*
>>> *LON0 : 0*
>>> *LAT1 : 90*
>>> *LON1 : 358.75*
>>> *DX : 942*
>>> *DY : 1250*
>>>
>>> Please let me know.
>>> Thank you.
>>>
>>> Jinwoong Yoo
>>> UNM
>>>
>>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>> wrote:
>>>
>>>> Dear John,
>>>>
>>>> Let me have a quick question about getting correct DX and DY in
copygb
>>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY
64*"
>>>> -x in.grb out.grb).
>>>>
>>>>
>>>> As you may know, grid info in my UPP output is as below:
>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
>>>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315) scan
>>>> 64 mode 136 Latin 0.000000 bdsgrid 1
>>>>
>>>> So my argument should be something like this:
>>>> *copygb.exe \*
>>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
>>>>
>>>> However, there are something unclear in the document (
>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>> ).
>>>>
>>>> NX : Number of points on latitude circle
>>>> NY : Number of points on longitude meridian
>>>> DX : Longitudinal increment in millidegrees
>>>> DY : Latitudinal increment in millidegrees
>>>>
>>>> You would notice that X,Y are not consistent as they refer to
lat/lon.
>>>> I think flipping X/Y in NX/NY should work so that X represents
>>>> longitude.
>>>>
>>>> That's not a big deal.
>>>> Here my own problem goes.
>>>>
>>>> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
>>>> However, if I convert them into millidegrees, DX is 323.515315
and DY
>>>> is 287.23175 computationally supposing the earth is a complete
sphere.
>>>> copygb takes only integers, DX is 323 and DY is 287. But, I'm not
sure
>>>> if it's right to use the different DX and DY or if I should use
the same
>>>> number for DX and DY.
>>>> Please let me know.
>>>>
>>>> Thank you.
>>>>
>>>> Jinwoong Yoo
>>>> UNM
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>>> wrote:
>>>>
>>>>> Dear John,
>>>>>
>>>>> Actually I was looking over copygb when I found your email
received.
>>>>> It seems that I have to convert the UPP output again using
copygb.exe
>>>>> to match up the grid of WRF results to the CCSM4's lat/lon grid.
>>>>>
>>>>> Isn't there any way to do that in one step? No?
>>>>>
>>>>> Along with the precipitation, I'd like to compare variables
(paired
>>>>> for WRF and CCSM4) below for example:
>>>>>
>>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
>>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>>>>>
>>>>>
>>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
>>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>>>>>
>>>>>
>>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
>>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
>>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
>>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
>>>>>
>>>>> Variables in the lists are the actual name of variables in the
CCSM4
>>>>> (including CAM and CLM monthly files) files and corresponding to
those in
>>>>> the WRF-ARW.
>>>>>
>>>>>
>>>>> I'm trying to run Grid-Stat for a test and I'd like to run
>>>>> Series-Analysis for the entire dataset of the WRF and the CCSM4.
>>>>> However, I'm concerning about how to run the Series-Analysis for
the
>>>>> monthly or annual data using MET?
>>>>> At the same time, I am also thinking about using the NCL to do
the
>>>>> same statistical analyses. I don't know which one will be easier
for me to
>>>>> execute for now.
>>>>> We'll see.
>>>>>
>>>>> Will let you know as I make progress, John.
>>>>>
>>>>> >>> FYI - I just talked to the person who provides support for
UPP
>>>>> through
>>>>> >>> wrfhelp, and she's expecting an email from you.
>>>>>
>>>>> I cc'ed this email to wrfhelp.
>>>>> Thank you.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Jinwoong Yoo
>>>>> UNM
>>>>>
>>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Jinwoong,
>>>>>>
>>>>>> I'm able to access your data on Yellowstone. I see that you'd
like to
>>>>>> compare your UPP output in GRIB format to some data that
follows the
>>>>>> NetCDF
>>>>>> CF-1.0 convention.
>>>>>>
>>>>>> UPP output: /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>>>>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>>>>>
>>>>>> Hopefully you're able to look at those files and identify which
GRIB
>>>>>> records you'd like to compare to which NetCDF variables. When
you
>>>>>> run the
>>>>>> MET tools, you'll define the forecast and observation fields to
be
>>>>>> used.
>>>>>>
>>>>>> Looking at your data, I see 2 immediate problems:
>>>>>>
>>>>>> (1) Gridded forecast and observation data must be on the same
grid
>>>>>> before
>>>>>> MET can compare them.
>>>>>> Here's a description from wgrib of the grid for your GRIB
data:
>>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
>>>>>> 349650
>>>>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
>>>>>> 315)
>>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>>>>> But it looks like the CCSM4 data is on a lat/lon grid.
>>>>>>
>>>>>> I'd suggest using the copygb utility to regrid the GRIB1 output
from
>>>>>> UPP to
>>>>>> the lat/lon grid of the CCSM4 data. Here's an example of doing
that:
>>>>>>
>>>>>>
>>>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>>>>
>>>>>> (2) There's an issue in the times in the accumulated
precipitation.
>>>>>> Try
>>>>>> running the following 2 commands:
>>>>>> wgrib
>>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252
|
>>>>>> grep APCP
>>>>>> wgrib
>>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258
|
>>>>>> grep APCP
>>>>>>
>>>>>> For the first, APCP is an accumulation of precip from hours 0
to
>>>>>> 252. The
>>>>>> second should be 0 to 258, but it's stored as 0 to 2! That's
because
>>>>>> UPP
>>>>>> is failing to realize that the precipitation timestamps are
>>>>>> overflowing the
>>>>>> 1 byte allocated for that second time unit. This problem was
>>>>>> addressed in
>>>>>> an earlier version of UPP but the problem is still showing up
in the
>>>>>> version you're using. If you plan to verify precip, I'd
suggest
>>>>>> writing
>>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're using,
and ask
>>>>>> them
>>>>>> if this should be fixed.
>>>>>>
>>>>>> Hope that helps.
>>>>>>
>>>>>> Thanks,
>>>>>> John
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
>>>>>> met_help at ucar.edu>
>>>>>> wrote:
>>>>>>
>>>>>> >
>>>>>> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
>>>>>> >
>>>>>> > Hi, Julie.
>>>>>> > Thank you for heads-up.
>>>>>> > I will wait for John to reply then.
>>>>>> > Thank you.
>>>>>> >
>>>>>> > Jinwoong Yoo
>>>>>> > UNM
>>>>>> >
>>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>>>>>> > met_help at ucar.edu>
>>>>>> > wrote:
>>>>>> >
>>>>>> > > Hello. John is out of the office, but will be back later
this
>>>>>> week. I am
>>>>>> > > unable to help you with this task. Hopefully, John will be
able
>>>>>> to
>>>>>> > respond
>>>>>> > > to this ticket later this week.
>>>>>> > >
>>>>>> > > Julie
>>>>>> > >
>>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>>>>>> met_help at ucar.edu>
>>>>>> > > wrote:
>>>>>> > >
>>>>>> > > >
>>>>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>>> > > >
>>>>>> > > > Dear John,
>>>>>> > > >
>>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
>>>>>> > > > I assume you can access to my glade scratch directory.
>>>>>> > > > My UPP ouputs are located
>>>>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>>>>>> > > >
>>>>>> > > > Against those files, I'd like to compare CCSM4 simulation
>>>>>> outputs some
>>>>>> > of
>>>>>> > > > which are located also under
>>>>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>>>>>> > > >
>>>>>> > > > Those are six hourly files all together. If you cannot
access
>>>>>> to those
>>>>>> > > > files, please let me know.
>>>>>> > > >
>>>>>> > > > However, there are some variables among the CCSM4 outputs
that
>>>>>> are
>>>>>> > > > available only monthly or annually. In this case, how can
I
>>>>>> > post-process
>>>>>> > > my
>>>>>> > > > WRF outputs for the usage of MET through UPP?
>>>>>> > > >
>>>>>> > > > Thank you.
>>>>>> > > >
>>>>>> > > > Regards,
>>>>>> > > >
>>>>>> > > > Jinwoong Yoo
>>>>>> > > > UNM
>>>>>> > > >
>>>>>> > > >
>>>>>> > > >
>>>>>> > > >
>>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via
RT <
>>>>>> > > > met_help at ucar.edu> wrote:
>>>>>> > > >
>>>>>> > > > > Jinwoo,
>>>>>> > > > >
>>>>>> > > > > I believe UPP will still run, just using a single time
value
>>>>>> as if it
>>>>>> > > > were
>>>>>> > > > > an instantaneous field rather than an accumulation
interval.
>>>>>> The
>>>>>> > > default
>>>>>> > > > > for ARW is to store precip as a runtime accumulation.
For
>>>>>> forecasts
>>>>>> > > > hours
>>>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as an
>>>>>> accumulation
>>>>>> > > > from 0
>>>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored
>>>>>> as an
>>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference,
>>>>>> but
>>>>>> > > something
>>>>>> > > > > to be aware of when you set up MET config files to
verify the
>>>>>> output.
>>>>>> > > > >
>>>>>> > > > > Give UPP a try and let me know how it goes.
>>>>>> > > > >
>>>>>> > > > > John
>>>>>> > > > >
>>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>>>>>> > met_help at ucar.edu>
>>>>>> > > > > wrote:
>>>>>> > > > >
>>>>>> > > > > >
>>>>>> > > > > > <URL:
>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>>> > > > > >
>>>>>> > > > > > Dear John,
>>>>>> > > > > >
>>>>>> > > > > > I see. Let me change "export lastfhr=18" to "export
>>>>>> lastfhr=744".
>>>>>> > > > > > Thank you for your comments.
>>>>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my
10 year
>>>>>> > > simulation
>>>>>> > > > > was
>>>>>> > > > > > integrated from a single initialization. Then, will
the UPP
>>>>>> > > run_unipost
>>>>>> > > > > > crash after the 255 forecast hours or just will
>>>>>> precipitation not
>>>>>> > be
>>>>>> > > > > stored
>>>>>> > > > > > any more aster the 255 forecast hours?
>>>>>> > > > > >
>>>>>> > > > > > Thank you.
>>>>>> > > > > >
>>>>>> > > > > > Regards,
>>>>>> > > > > >
>>>>>> > > > > > Jinwoong Yoo
>>>>>> > > > > > UNM
>>>>>> > > > > >
>>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway
via RT <
>>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>>>>>> > > > > >
>>>>>> > > > > > > Jinwoong,
>>>>>> > > > > > >
>>>>>> > > > > > > UPP support is typically provided through
>>>>>> wrf_help at ucar.edu
>>>>>> > > > > > <javascript:;>. I'll give it
>>>>>> > > > > > > a shot, but if this doesn't solve it, I'll refer
you to
>>>>>> wrf_help.
>>>>>> > > > > > >
>>>>>> > > > > > > I assume you're running the script named
"run_unipost".
>>>>>> Looking
>>>>>> > in
>>>>>> > > > > > there,
>>>>>> > > > > > > you'll see:
>>>>>> > > > > > >
>>>>>> > > > > > > export startdate=2005012300
>>>>>> > > > > > >
>>>>>> > > > > > > export fhr=00
>>>>>> > > > > > > export lastfhr=18
>>>>>> > > > > > > export incrementhr=03
>>>>>> > > > > > >
>>>>>> > > > > > > I'd suggest modifying the...
>>>>>> > > > > > > - "startdate" to be the model initialization time
to be
>>>>>> > processed.
>>>>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
>>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days
* 24
>>>>>> > hours/day).
>>>>>> > > > > > > - Set "incrementhr" to 06 to process output every
6
>>>>>> hours.
>>>>>> > > > > > >
>>>>>> > > > > > > Give that a shot and let me know if that gets you
output
>>>>>> past the
>>>>>> > > > first
>>>>>> > > > > > > day.
>>>>>> > > > > > >
>>>>>> > > > > > > Are you going to be evaluating precipitation? I
ask that
>>>>>> because
>>>>>> > > > > precip
>>>>>> > > > > > is
>>>>>> > > > > > > an accumulated interval, meaning that it has a
beginning
>>>>>> > > accumulation
>>>>>> > > > > > time
>>>>>> > > > > > > and ending accumulation time. GRIB1 is limited in
the
>>>>>> amount of
>>>>>> > > > space
>>>>>> > > > > > > available for storing times. When you get past 255
>>>>>> forecast
>>>>>> > hours,
>>>>>> > > > you
>>>>>> > > > > > can
>>>>>> > > > > > > no longer store beginning/ending times for precip
because
>>>>>> it runs
>>>>>> > > out
>>>>>> > > > > of
>>>>>> > > > > > > room in the header. Do you have a 10 year
simulation
>>>>>> computed
>>>>>> > > from a
>>>>>> > > > > > > single initialization? Meaning, do you have
forecast
>>>>>> lead times
>>>>>> > > out
>>>>>> > > > to
>>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)?
I'm
>>>>>> wondering
>>>>>> > if
>>>>>> > > > > we'll
>>>>>> > > > > > > run into problems formatting those time strings.
>>>>>> > > > > > >
>>>>>> > > > > > > Thanks,
>>>>>> > > > > > > John
>>>>>> > > > > > >
>>>>>> > > > > > >
>>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via
RT <
>>>>>> > > > > met_help at ucar.edu
>>>>>> > > > > > <javascript:;>>
>>>>>> > > > > > > wrote:
>>>>>> > > > > > >
>>>>>> > > > > > > >
>>>>>> > > > > > > > <URL:
>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>>> > >
>>>>>> > > > > > > >
>>>>>> > > > > > > > Dear John,
>>>>>> > > > > > > >
>>>>>> > > > > > > > I ran UPP with my multiple wrfout files, first
with the
>>>>>> first
>>>>>> > > month
>>>>>> > > > > of
>>>>>> > > > > > > the
>>>>>> > > > > > > > 10 year simulation. However, after executing the
>>>>>> run_unipost
>>>>>> > > > script,
>>>>>> > > > > > UPP
>>>>>> > > > > > > > produced only four files: unipost_d01.00.out,
>>>>>> > unipost_d01.06.out,
>>>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems
>>>>>> that UPP
>>>>>> > > > > processed
>>>>>> > > > > > > > only the first day's output among the one month
data.
>>>>>> > > > > > > > If you are familiar with UPP, how can I make sure
for
>>>>>> UPP to
>>>>>> > > > process
>>>>>> > > > > > > > multiple wrfout files more than one day?
>>>>>> > > > > > > > Thank you.
>>>>>> > > > > > > >
>>>>>> > > > > > > > Regards,
>>>>>> > > > > > > >
>>>>>> > > > > > > > Jinwoong Yoo
>>>>>> > > > > > > > UNM
>>>>>> > > > > > > >
>>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>>>>>> > > > > jinwoong.yoo at gmail.com
>>>>>> > > > > > <javascript:;>>
>>>>>> > > > > > > > wrote:
>>>>>> > > > > > > >
>>>>>> > > > > > > > > Dear John,
>>>>>> > > > > > > > >
>>>>>> > > > > > > > > By the way, what about the CCSM4 file in NetCDF
>>>>>> format?
>>>>>> > > > > > > > > Do I need to post process of the CCSM4 files
also to
>>>>>> run MET
>>>>>> > > > along
>>>>>> > > > > > with
>>>>>> > > > > > > > > the WRF-ARW files?
>>>>>> > > > > > > > > Thank you.
>>>>>> > > > > > > > >
>>>>>> > > > > > > > > Regards,
>>>>>> > > > > > > > >
>>>>>> > > > > > > > > Jinwoong Yoo
>>>>>> > > > > > > > > UNM
>>>>>> > > > > > > > >
>>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo <
>>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>>>>>> > > > > > > > > wrote:
>>>>>> > > > > > > > >
>>>>>> > > > > > > > >> Dear John,
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >> Thank you for your kind email with the
direction.
>>>>>> > > > > > > > >> Let me first post process some of my WRF-ARW
outputs
>>>>>> using
>>>>>> > UPP
>>>>>> > > > and
>>>>>> > > > > > let
>>>>>> > > > > > > > >> you know.
>>>>>> > > > > > > > >> Thank you very much.
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >> Regards,
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >> Jinwoong Yoo
>>>>>> > > > > > > > >> UNM
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway
>>>>>> via RT <
>>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >>> Jinwoong,
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>> I see that you'd like some help running the
>>>>>> series-analysis
>>>>>> > > > tool
>>>>>> > > > > on
>>>>>> > > > > > > > your
>>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is
that
>>>>>> MET can't
>>>>>> > > read
>>>>>> > > > > the
>>>>>> > > > > > > raw
>>>>>> > > > > > > > >>> wrfout files directly. Instead you first
need to
>>>>>> post
>>>>>> > > process
>>>>>> > > > > them
>>>>>> > > > > > > > using
>>>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
>>>>>> utility
>>>>>> > > > > (produced
>>>>>> > > > > > > > >>> NetCDF
>>>>>> > > > > > > > >>> output). After you've post processed some
files,
>>>>>> please
>>>>>> > send
>>>>>> > > > us
>>>>>> > > > > > some
>>>>>> > > > > > > > >>> sample data by posting it to our anonymous
ftp
>>>>>> site. Also
>>>>>> > > > send a
>>>>>> > > > > > > > >>> description of the comparisons you'd like to
make
>>>>>> between
>>>>>> > > the 2
>>>>>> > > > > > > > datasets.
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>> Here's instructions for sending us data:
>>>>>> > > > > > > > >>>
>>>>>> http://www.dtcenter.org/met/users/support/met_help.php
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>> Thanks
>>>>>> > > > > > > > >>> John Halley Gotway
>>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong
Yoo via
>>>>>> RT <
>>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> wrote:
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099 was
acted
>>>>>> upon.
>>>>>> > > > > > > > >>> > Transaction: Ticket created by
>>>>>> jinwoong.yoo at gmail.com
>>>>>> > > > > > <javascript:;>
>>>>>> > > > > > > <javascript:;>
>>>>>> > > > > > > > >>> > Queue: met_help
>>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs and
CCSM4
>>>>>> outputs
>>>>>> > > > > > > > >>> > Owner: Nobody
>>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
>>>>>> <javascript:;>
>>>>>> > > > > > <javascript:;>
>>>>>> > > > > > > > >>> > Status: new
>>>>>> > > > > > > > >>> > Ticket <URL:
>>>>>> > > > > > > >
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Dear Met Help,
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Hi.
>>>>>> > > > > > > > >>> > I am looking for a way to verify WRF model
>>>>>> performance
>>>>>> > as a
>>>>>> > > > > > climate
>>>>>> > > > > > > > >>> model.
>>>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate
>>>>>> model
>>>>>> > > > > simulations
>>>>>> > > > > > > > using
>>>>>> > > > > > > > >>> the
>>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling)
>>>>>> for two
>>>>>> > > > > > different
>>>>>> > > > > > > > time
>>>>>> > > > > > > > >>> > periods: one for the LGM and the other for
the
>>>>>> modern as
>>>>>> > a
>>>>>> > > > > > control
>>>>>> > > > > > > > run.
>>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly output
for
>>>>>> ten years
>>>>>> > > > > periods
>>>>>> > > > > > > > each.
>>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
>>>>>> outputs.
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>>>>>> > results,
>>>>>> > > > > > > verifying
>>>>>> > > > > > > > >>> the
>>>>>> > > > > > > > >>> > WRF model simulations.
>>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
>>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to do
the
>>>>>> similar
>>>>>> > > jobs
>>>>>> > > > > but
>>>>>> > > > > > I
>>>>>> > > > > > > > need
>>>>>> > > > > > > > >>> > much more help from you than the MET Online
>>>>>> Tutorial to
>>>>>> > > make
>>>>>> > > > a
>>>>>> > > > > > step
>>>>>> > > > > > > > >>> > forward.
>>>>>> > > > > > > > >>> > Any help will be highly appreciated.
>>>>>> > > > > > > > >>> > Thank you.
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Regards,
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> > Jinwoong Yoo
>>>>>> > > > > > > > >>> > Postdoc
>>>>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>>>>>> > > > > > > > >>> > UNM
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>> >
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>>
>>>>>> > > > > > > > >>
>>>>>> > > > > > > > >
>>>>>> > > > > > > >
>>>>>> > > > > > > >
>>>>>> > > > > > >
>>>>>> > > > > > >
>>>>>> > > > > >
>>>>>> > > > > >
>>>>>> > > > >
>>>>>> > > > >
>>>>>> > > >
>>>>>> > > >
>>>>>> > >
>>>>>> > >
>>>>>> > > --
>>>>>> > > Julie Prestopnik
>>>>>> > > National Center for Atmospheric Research
>>>>>> > > Research Applications Laboratory
>>>>>> > > Phone: 303.497.8399
>>>>>> > > Email: jpresto at ucar.edu
>>>>>> > >
>>>>>> > >
>>>>>> >
>>>>>> >
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Mon Oct 06 16:58:12 2014
Supporting my arguments here,
"wgrib -V WRFPRS_d01.00_2latlon" gives grid information like:
latlon: lat -89.529000 to 89.529000 by 0.942000 nxny 55008
long 0.625000 to 359.375000 by 1.250000, (288 x 191) scan 64
mode
128
where, WRFPRS_d01.00_2latlon is copygb output file.
So, any comment on this issue with
/glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat on
Yellowstone?
Please let me know.
Thank you.
Jinwoong Yoo
UNM
On Mon, Oct 6, 2014 at 4:42 PM, Jinwoong Yoo <jinwoong.yoo at gmail.com>
wrote:
> I'm getting closer to the point of the problem.
>
> I have corrected my copygb argument so close as to get this minor(?)
error:
>
> ERROR : process_command_line() -> The forecast and observation
grids do
> not match:
> Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: -0.625
> delta_lat: 0.942 delta_lon: 1.250 !=
> Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625
delta_lat:
> 0.942 delta_lon: 1.250
>
>
> My copygb argument was:
> copygb.exe -g"255 0 288 191 -89529 625 128 89529 359375 1250 942
64"
> -x $infile $outfile
>
> As you may know, 625 in this case is Longitude of origin * 1000 (lon
=
> -180 ... 180). But I don't know how the negative sign got into the
> execution of the program, yielding lon_ll: -0.625.
>
> Do you know how I should solve this problem?
> Is this a bug in the program?
>
> By the way, as I noted in my previous email,
> I found and confirmed the fault statements in the MET Online
Tutorial for
> the copygb Tool: Lat/Lon Grid (
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> )
>
> NX should be Number of points on longitude meridian,
> while NY should be Number of points on latitude circle, definitely.
>
> Thanks.
>
> Jinwoong Yoo
> UNM
>
>
> On Mon, Oct 6, 2014 at 1:45 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
>> Hi.
>>
>> I encounter errors while I'm running Grid_Stat on Yellowstone as
below:
>>
>> [jyoo at yslogin4 analysis]$
>> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
>> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
>> b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm
-outdir
>> /glade/scratch/jyoo/domain/mean/analysis/gridstatout
>> DEBUG 1: Default Config File:
>> /glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
>> DEBUG 1: User Config File: GridStatConfig_ccsm
>> WARNING:
>> WARNING: NcCfFile::open() -> could not extract init time from file
name
>> WARNING: Using init time of 0
>> WARNING:
>> ERROR :
>> ERROR : process_command_line() -> The forecast and observation
grids do
>> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
>> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx:
288 Ny:
>> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
>>
>> WRF output that was postprocessed using UPP and copygb is compared
with
>> CCSM4 result in NetCDF format here for a test.
>>
>> First,
>> WARNING: NcCfFile::open() -> could not extract init time from file
name
>> WARNING: Using init time of 0
>>
>> Since the CCSM4 file contains 6 hourly data for a month, I set "
level
>> = [ "(1,*,*)" ];"
>> for my desired level (surface). But it seems it didn't work.
Setting
>> "(*,*)" didn't work, neither.
>>
>> fcst = {
>> wind_thresh = [ NA ];
>>
>> field = [
>> {
>> name = "CPRAT";
>> level = [ "R277" ];
>> cat_thresh = [ >0.0, >=5.0 ];
>> }
>> ];
>>
>> };
>>
>> obs = {
>> wind_thresh = [ NA ];
>>
>> field = [
>> {
>> name = "PRECC";
>> level = [ "(1,*,*)" ];
>> cat_thresh = [ >0.0, >=5.0 ];
>> }
>> ];
>>
>> };
>>
>> Second, obviously it seems my previous copygb was not correct.
>> ERROR : process_command_line() -> The forecast and observation
grids do
>> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
>> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon Nx:
288 Ny:
>> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon: 1.250
>>
>>
>> My copygy commend was like this:
>>
>> */glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
>> * -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250 64" \*
>>
>> * /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
>>
>> Output grid information after executing the copygb looks like this:
>> latlon: lat -90.000000 to 90.000000 by 1.250000 nxny 55296
>> long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan 64 mode
128
>> bdsgrid 1
>>
>> However, there are something strange.
>> According to what Grid-Stat reads in for the CCSM4 file
(Projection:
>> Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat:
0.942
>> delta_lon: 1.250), lat and lon information were flipped compared to
what I
>> get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ; slat =
191 ;slon
>> = 288) and the Ny: 191 matches to slat(staggered latitude) not
latitude.
>>
>> One thing to note is that my WRF simulation is a tropical channel
>> simulation covering 45S to 45N while the CCSM4 is global.
>>
>> What am I doing wrong with copygb and GridStatConfig file?
>> Thank you in advance.
>>
>> Regards,
>>
>> Jinwoong Yoo
>> UNM
>>
>> On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>> wrote:
>>
>>> That's right. I was wrong earlier and the later arguments worked
for my
>>> regridding of the WRF output in Mercator to Lat-Lon grid.
>>> Thanks.
>>>
>>> Jinwoong Yoo
>>> UNM
>>>
>>> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>> wrote:
>>>
>>>> Am I misunderstanding the arguments of copygb?
>>>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
>>>> out.grb)
>>>>
>>>> Should be those numbers for the target grid in Lat/Lon of the
CCSM4 in
>>>> my case, not for the WRF output values?
>>>> Such that
>>>> *NX (*Number of points on latitude circle) : 192
>>>> *NY (*Number of points on longitude meridian*) : 288*
>>>> *LAT0 : -90*
>>>> *LON0 : 0*
>>>> *LAT1 : 90*
>>>> *LON1 : 358.75*
>>>> *DX : 942*
>>>> *DY : 1250*
>>>>
>>>> Please let me know.
>>>> Thank you.
>>>>
>>>> Jinwoong Yoo
>>>> UNM
>>>>
>>>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>>> wrote:
>>>>
>>>>> Dear John,
>>>>>
>>>>> Let me have a quick question about getting correct DX and DY in
copygb
>>>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY
64*"
>>>>> -x in.grb out.grb).
>>>>>
>>>>>
>>>>> As you may know, grid info in my UPP output is as below:
>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
>>>>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>>>>
>>>>> So my argument should be something like this:
>>>>> *copygb.exe \*
>>>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
>>>>>
>>>>> However, there are something unclear in the document (
>>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>>> ).
>>>>>
>>>>> NX : Number of points on latitude circle
>>>>> NY : Number of points on longitude meridian
>>>>> DX : Longitudinal increment in millidegrees
>>>>> DY : Latitudinal increment in millidegrees
>>>>>
>>>>> You would notice that X,Y are not consistent as they refer to
lat/lon.
>>>>> I think flipping X/Y in NX/NY should work so that X represents
>>>>> longitude.
>>>>>
>>>>> That's not a big deal.
>>>>> Here my own problem goes.
>>>>>
>>>>> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
>>>>> However, if I convert them into millidegrees, DX is 323.515315
and DY
>>>>> is 287.23175 computationally supposing the earth is a complete
sphere.
>>>>> copygb takes only integers, DX is 323 and DY is 287. But, I'm
not
>>>>> sure if it's right to use the different DX and DY or if I should
use the
>>>>> same number for DX and DY.
>>>>> Please let me know.
>>>>>
>>>>> Thank you.
>>>>>
>>>>> Jinwoong Yoo
>>>>> UNM
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Dear John,
>>>>>>
>>>>>> Actually I was looking over copygb when I found your email
received.
>>>>>> It seems that I have to convert the UPP output again using
copygb.exe
>>>>>> to match up the grid of WRF results to the CCSM4's lat/lon
grid.
>>>>>>
>>>>>> Isn't there any way to do that in one step? No?
>>>>>>
>>>>>> Along with the precipitation, I'd like to compare variables
(paired
>>>>>> for WRF and CCSM4) below for example:
>>>>>>
>>>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
>>>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
>>>>>>
>>>>>>
>>>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
>>>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
>>>>>>
>>>>>>
>>>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
>>>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
>>>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
>>>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
>>>>>>
>>>>>> Variables in the lists are the actual name of variables in the
CCSM4
>>>>>> (including CAM and CLM monthly files) files and corresponding
to those in
>>>>>> the WRF-ARW.
>>>>>>
>>>>>>
>>>>>> I'm trying to run Grid-Stat for a test and I'd like to run
>>>>>> Series-Analysis for the entire dataset of the WRF and the
CCSM4.
>>>>>> However, I'm concerning about how to run the Series-Analysis
for the
>>>>>> monthly or annual data using MET?
>>>>>> At the same time, I am also thinking about using the NCL to do
the
>>>>>> same statistical analyses. I don't know which one will be
easier for me to
>>>>>> execute for now.
>>>>>> We'll see.
>>>>>>
>>>>>> Will let you know as I make progress, John.
>>>>>>
>>>>>> >>> FYI - I just talked to the person who provides support for
UPP
>>>>>> through
>>>>>> >>> wrfhelp, and she's expecting an email from you.
>>>>>>
>>>>>> I cc'ed this email to wrfhelp.
>>>>>> Thank you.
>>>>>>
>>>>>> Regards,
>>>>>>
>>>>>> Jinwoong Yoo
>>>>>> UNM
>>>>>>
>>>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
>>>>>> met_help at ucar.edu> wrote:
>>>>>>
>>>>>>> Jinwoong,
>>>>>>>
>>>>>>> I'm able to access your data on Yellowstone. I see that you'd
like
>>>>>>> to
>>>>>>> compare your UPP output in GRIB format to some data that
follows the
>>>>>>> NetCDF
>>>>>>> CF-1.0 convention.
>>>>>>>
>>>>>>> UPP output:
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
>>>>>>> CCSM4 output:
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
>>>>>>>
>>>>>>> Hopefully you're able to look at those files and identify
which GRIB
>>>>>>> records you'd like to compare to which NetCDF variables. When
you
>>>>>>> run the
>>>>>>> MET tools, you'll define the forecast and observation fields
to be
>>>>>>> used.
>>>>>>>
>>>>>>> Looking at your data, I see 2 immediate problems:
>>>>>>>
>>>>>>> (1) Gridded forecast and observation data must be on the same
grid
>>>>>>> before
>>>>>>> MET can compare them.
>>>>>>> Here's a description from wgrib of the grid for your GRIB
data:
>>>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
>>>>>>> 349650
>>>>>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
>>>>>>> 315)
>>>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
>>>>>>> But it looks like the CCSM4 data is on a lat/lon grid.
>>>>>>>
>>>>>>> I'd suggest using the copygb utility to regrid the GRIB1
output from
>>>>>>> UPP to
>>>>>>> the lat/lon grid of the CCSM4 data. Here's an example of
doing that:
>>>>>>>
>>>>>>>
>>>>>>>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
>>>>>>>
>>>>>>> (2) There's an issue in the times in the accumulated
precipitation.
>>>>>>> Try
>>>>>>> running the following 2 commands:
>>>>>>> wgrib
>>>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252
|
>>>>>>> grep APCP
>>>>>>> wgrib
>>>>>>> /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258
|
>>>>>>> grep APCP
>>>>>>>
>>>>>>> For the first, APCP is an accumulation of precip from hours 0
to
>>>>>>> 252. The
>>>>>>> second should be 0 to 258, but it's stored as 0 to 2! That's
>>>>>>> because UPP
>>>>>>> is failing to realize that the precipitation timestamps are
>>>>>>> overflowing the
>>>>>>> 1 byte allocated for that second time unit. This problem was
>>>>>>> addressed in
>>>>>>> an earlier version of UPP but the problem is still showing up
in the
>>>>>>> version you're using. If you plan to verify precip, I'd
suggest
>>>>>>> writing
>>>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're using,
and
>>>>>>> ask them
>>>>>>> if this should be fixed.
>>>>>>>
>>>>>>> Hope that helps.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> John
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
>>>>>>> met_help at ucar.edu>
>>>>>>> wrote:
>>>>>>>
>>>>>>> >
>>>>>>> > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>>>> >
>>>>>>> > Hi, Julie.
>>>>>>> > Thank you for heads-up.
>>>>>>> > I will wait for John to reply then.
>>>>>>> > Thank you.
>>>>>>> >
>>>>>>> > Jinwoong Yoo
>>>>>>> > UNM
>>>>>>> >
>>>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
>>>>>>> > met_help at ucar.edu>
>>>>>>> > wrote:
>>>>>>> >
>>>>>>> > > Hello. John is out of the office, but will be back later
this
>>>>>>> week. I am
>>>>>>> > > unable to help you with this task. Hopefully, John will
be able
>>>>>>> to
>>>>>>> > respond
>>>>>>> > > to this ticket later this week.
>>>>>>> > >
>>>>>>> > > Julie
>>>>>>> > >
>>>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
>>>>>>> met_help at ucar.edu>
>>>>>>> > > wrote:
>>>>>>> > >
>>>>>>> > > >
>>>>>>> > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>>>> >
>>>>>>> > > >
>>>>>>> > > > Dear John,
>>>>>>> > > >
>>>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
>>>>>>> > > > I assume you can access to my glade scratch directory.
>>>>>>> > > > My UPP ouputs are located
>>>>>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
>>>>>>> > > >
>>>>>>> > > > Against those files, I'd like to compare CCSM4
simulation
>>>>>>> outputs some
>>>>>>> > of
>>>>>>> > > > which are located also under
>>>>>>> > > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
>>>>>>> > > >
>>>>>>> > > > Those are six hourly files all together. If you cannot
access
>>>>>>> to those
>>>>>>> > > > files, please let me know.
>>>>>>> > > >
>>>>>>> > > > However, there are some variables among the CCSM4
outputs that
>>>>>>> are
>>>>>>> > > > available only monthly or annually. In this case, how
can I
>>>>>>> > post-process
>>>>>>> > > my
>>>>>>> > > > WRF outputs for the usage of MET through UPP?
>>>>>>> > > >
>>>>>>> > > > Thank you.
>>>>>>> > > >
>>>>>>> > > > Regards,
>>>>>>> > > >
>>>>>>> > > > Jinwoong Yoo
>>>>>>> > > > UNM
>>>>>>> > > >
>>>>>>> > > >
>>>>>>> > > >
>>>>>>> > > >
>>>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway via
RT <
>>>>>>> > > > met_help at ucar.edu> wrote:
>>>>>>> > > >
>>>>>>> > > > > Jinwoo,
>>>>>>> > > > >
>>>>>>> > > > > I believe UPP will still run, just using a single time
value
>>>>>>> as if it
>>>>>>> > > > were
>>>>>>> > > > > an instantaneous field rather than an accumulation
>>>>>>> interval. The
>>>>>>> > > default
>>>>>>> > > > > for ARW is to store precip as a runtime accumulation.
For
>>>>>>> forecasts
>>>>>>> > > > hours
>>>>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as
an
>>>>>>> accumulation
>>>>>>> > > > from 0
>>>>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll be
stored
>>>>>>> as an
>>>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference,
>>>>>>> but
>>>>>>> > > something
>>>>>>> > > > > to be aware of when you set up MET config files to
verify
>>>>>>> the output.
>>>>>>> > > > >
>>>>>>> > > > > Give UPP a try and let me know how it goes.
>>>>>>> > > > >
>>>>>>> > > > > John
>>>>>>> > > > >
>>>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
>>>>>>> > met_help at ucar.edu>
>>>>>>> > > > > wrote:
>>>>>>> > > > >
>>>>>>> > > > > >
>>>>>>> > > > > > <URL:
>>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>>>>>>> > > > > >
>>>>>>> > > > > > Dear John,
>>>>>>> > > > > >
>>>>>>> > > > > > I see. Let me change "export lastfhr=18" to "export
>>>>>>> lastfhr=744".
>>>>>>> > > > > > Thank you for your comments.
>>>>>>> > > > > > Yes, I'd like to evaluate precipitation, too, and my
10
>>>>>>> year
>>>>>>> > > simulation
>>>>>>> > > > > was
>>>>>>> > > > > > integrated from a single initialization. Then, will
the UPP
>>>>>>> > > run_unipost
>>>>>>> > > > > > crash after the 255 forecast hours or just will
>>>>>>> precipitation not
>>>>>>> > be
>>>>>>> > > > > stored
>>>>>>> > > > > > any more aster the 255 forecast hours?
>>>>>>> > > > > >
>>>>>>> > > > > > Thank you.
>>>>>>> > > > > >
>>>>>>> > > > > > Regards,
>>>>>>> > > > > >
>>>>>>> > > > > > Jinwoong Yoo
>>>>>>> > > > > > UNM
>>>>>>> > > > > >
>>>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley Gotway
via RT
>>>>>>> <
>>>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
>>>>>>> > > > > >
>>>>>>> > > > > > > Jinwoong,
>>>>>>> > > > > > >
>>>>>>> > > > > > > UPP support is typically provided through
>>>>>>> wrf_help at ucar.edu
>>>>>>> > > > > > <javascript:;>. I'll give it
>>>>>>> > > > > > > a shot, but if this doesn't solve it, I'll refer
you to
>>>>>>> wrf_help.
>>>>>>> > > > > > >
>>>>>>> > > > > > > I assume you're running the script named
"run_unipost".
>>>>>>> Looking
>>>>>>> > in
>>>>>>> > > > > > there,
>>>>>>> > > > > > > you'll see:
>>>>>>> > > > > > >
>>>>>>> > > > > > > export startdate=2005012300
>>>>>>> > > > > > >
>>>>>>> > > > > > > export fhr=00
>>>>>>> > > > > > > export lastfhr=18
>>>>>>> > > > > > > export incrementhr=03
>>>>>>> > > > > > >
>>>>>>> > > > > > > I'd suggest modifying the...
>>>>>>> > > > > > > - "startdate" to be the model initialization time
to be
>>>>>>> > processed.
>>>>>>> > > > > > > - Leave "fhr" set to 00 to start at the analysis
time.
>>>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31 days
* 24
>>>>>>> > hours/day).
>>>>>>> > > > > > > - Set "incrementhr" to 06 to process output every
6
>>>>>>> hours.
>>>>>>> > > > > > >
>>>>>>> > > > > > > Give that a shot and let me know if that gets you
output
>>>>>>> past the
>>>>>>> > > > first
>>>>>>> > > > > > > day.
>>>>>>> > > > > > >
>>>>>>> > > > > > > Are you going to be evaluating precipitation? I
ask
>>>>>>> that because
>>>>>>> > > > > precip
>>>>>>> > > > > > is
>>>>>>> > > > > > > an accumulated interval, meaning that it has a
beginning
>>>>>>> > > accumulation
>>>>>>> > > > > > time
>>>>>>> > > > > > > and ending accumulation time. GRIB1 is limited in
the
>>>>>>> amount of
>>>>>>> > > > space
>>>>>>> > > > > > > available for storing times. When you get past
255
>>>>>>> forecast
>>>>>>> > hours,
>>>>>>> > > > you
>>>>>>> > > > > > can
>>>>>>> > > > > > > no longer store beginning/ending times for precip
>>>>>>> because it runs
>>>>>>> > > out
>>>>>>> > > > > of
>>>>>>> > > > > > > room in the header. Do you have a 10 year
simulation
>>>>>>> computed
>>>>>>> > > from a
>>>>>>> > > > > > > single initialization? Meaning, do you have
forecast
>>>>>>> lead times
>>>>>>> > > out
>>>>>>> > > > to
>>>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)?
I'm
>>>>>>> wondering
>>>>>>> > if
>>>>>>> > > > > we'll
>>>>>>> > > > > > > run into problems formatting those time strings.
>>>>>>> > > > > > >
>>>>>>> > > > > > > Thanks,
>>>>>>> > > > > > > John
>>>>>>> > > > > > >
>>>>>>> > > > > > >
>>>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo via
RT <
>>>>>>> > > > > met_help at ucar.edu
>>>>>>> > > > > > <javascript:;>>
>>>>>>> > > > > > > wrote:
>>>>>>> > > > > > >
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > <URL:
>>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>>>> > >
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > Dear John,
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > I ran UPP with my multiple wrfout files, first
with
>>>>>>> the first
>>>>>>> > > month
>>>>>>> > > > > of
>>>>>>> > > > > > > the
>>>>>>> > > > > > > > 10 year simulation. However, after executing the
>>>>>>> run_unipost
>>>>>>> > > > script,
>>>>>>> > > > > > UPP
>>>>>>> > > > > > > > produced only four files: unipost_d01.00.out,
>>>>>>> > unipost_d01.06.out,
>>>>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems
>>>>>>> that UPP
>>>>>>> > > > > processed
>>>>>>> > > > > > > > only the first day's output among the one month
data.
>>>>>>> > > > > > > > If you are familiar with UPP, how can I make
sure for
>>>>>>> UPP to
>>>>>>> > > > process
>>>>>>> > > > > > > > multiple wrfout files more than one day?
>>>>>>> > > > > > > > Thank you.
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > Regards,
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > Jinwoong Yoo
>>>>>>> > > > > > > > UNM
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo <
>>>>>>> > > > > jinwoong.yoo at gmail.com
>>>>>>> > > > > > <javascript:;>>
>>>>>>> > > > > > > > wrote:
>>>>>>> > > > > > > >
>>>>>>> > > > > > > > > Dear John,
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > > > By the way, what about the CCSM4 file in
NetCDF
>>>>>>> format?
>>>>>>> > > > > > > > > Do I need to post process of the CCSM4 files
also to
>>>>>>> run MET
>>>>>>> > > > along
>>>>>>> > > > > > with
>>>>>>> > > > > > > > > the WRF-ARW files?
>>>>>>> > > > > > > > > Thank you.
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > > > Regards,
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > > > Jinwoong Yoo
>>>>>>> > > > > > > > > UNM
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong Yoo
<
>>>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
>>>>>>> > > > > > > > > wrote:
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > > >> Dear John,
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >> Thank you for your kind email with the
direction.
>>>>>>> > > > > > > > >> Let me first post process some of my WRF-ARW
>>>>>>> outputs using
>>>>>>> > UPP
>>>>>>> > > > and
>>>>>>> > > > > > let
>>>>>>> > > > > > > > >> you know.
>>>>>>> > > > > > > > >> Thank you very much.
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >> Regards,
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >> Jinwoong Yoo
>>>>>>> > > > > > > > >> UNM
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John Halley
Gotway
>>>>>>> via RT <
>>>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >>> Jinwoong,
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>> I see that you'd like some help running the
>>>>>>> series-analysis
>>>>>>> > > > tool
>>>>>>> > > > > on
>>>>>>> > > > > > > > your
>>>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note is
that
>>>>>>> MET can't
>>>>>>> > > read
>>>>>>> > > > > the
>>>>>>> > > > > > > raw
>>>>>>> > > > > > > > >>> wrfout files directly. Instead you first
need to
>>>>>>> post
>>>>>>> > > process
>>>>>>> > > > > them
>>>>>>> > > > > > > > using
>>>>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
>>>>>>> utility
>>>>>>> > > > > (produced
>>>>>>> > > > > > > > >>> NetCDF
>>>>>>> > > > > > > > >>> output). After you've post processed some
files,
>>>>>>> please
>>>>>>> > send
>>>>>>> > > > us
>>>>>>> > > > > > some
>>>>>>> > > > > > > > >>> sample data by posting it to our anonymous
ftp
>>>>>>> site. Also
>>>>>>> > > > send a
>>>>>>> > > > > > > > >>> description of the comparisons you'd like to
make
>>>>>>> between
>>>>>>> > > the 2
>>>>>>> > > > > > > > datasets.
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>> Here's instructions for sending us data:
>>>>>>> > > > > > > > >>>
>>>>>>> http://www.dtcenter.org/met/users/support/met_help.php
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>> Thanks
>>>>>>> > > > > > > > >>> John Halley Gotway
>>>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong
Yoo via
>>>>>>> RT <
>>>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> wrote:
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099
was
>>>>>>> acted upon.
>>>>>>> > > > > > > > >>> > Transaction: Ticket created by
>>>>>>> jinwoong.yoo at gmail.com
>>>>>>> > > > > > <javascript:;>
>>>>>>> > > > > > > <javascript:;>
>>>>>>> > > > > > > > >>> > Queue: met_help
>>>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs
and
>>>>>>> CCSM4 outputs
>>>>>>> > > > > > > > >>> > Owner: Nobody
>>>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
>>>>>>> <javascript:;>
>>>>>>> > > > > > <javascript:;>
>>>>>>> > > > > > > > >>> > Status: new
>>>>>>> > > > > > > > >>> > Ticket <URL:
>>>>>>> > > > > > > >
>>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Dear Met Help,
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Hi.
>>>>>>> > > > > > > > >>> > I am looking for a way to verify WRF model
>>>>>>> performance
>>>>>>> > as a
>>>>>>> > > > > > climate
>>>>>>> > > > > > > > >>> model.
>>>>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate
>>>>>>> model
>>>>>>> > > > > simulations
>>>>>>> > > > > > > > using
>>>>>>> > > > > > > > >>> the
>>>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
downscaling)
>>>>>>> for two
>>>>>>> > > > > > different
>>>>>>> > > > > > > > time
>>>>>>> > > > > > > > >>> > periods: one for the LGM and the other for
the
>>>>>>> modern as
>>>>>>> > a
>>>>>>> > > > > > control
>>>>>>> > > > > > > > run.
>>>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly
output for
>>>>>>> ten years
>>>>>>> > > > > periods
>>>>>>> > > > > > > > each.
>>>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
>>>>>>> outputs.
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Basically, I'd like to compare the two
simulation
>>>>>>> > results,
>>>>>>> > > > > > > verifying
>>>>>>> > > > > > > > >>> the
>>>>>>> > > > > > > > >>> > WRF model simulations.
>>>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
>>>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to
do the
>>>>>>> similar
>>>>>>> > > jobs
>>>>>>> > > > > but
>>>>>>> > > > > > I
>>>>>>> > > > > > > > need
>>>>>>> > > > > > > > >>> > much more help from you than the MET
Online
>>>>>>> Tutorial to
>>>>>>> > > make
>>>>>>> > > > a
>>>>>>> > > > > > step
>>>>>>> > > > > > > > >>> > forward.
>>>>>>> > > > > > > > >>> > Any help will be highly appreciated.
>>>>>>> > > > > > > > >>> > Thank you.
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Regards,
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> > Jinwoong Yoo
>>>>>>> > > > > > > > >>> > Postdoc
>>>>>>> > > > > > > > >>> > Department of Earth and Planetary Sciences
>>>>>>> > > > > > > > >>> > UNM
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>> >
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>>
>>>>>>> > > > > > > > >>
>>>>>>> > > > > > > > >
>>>>>>> > > > > > > >
>>>>>>> > > > > > > >
>>>>>>> > > > > > >
>>>>>>> > > > > > >
>>>>>>> > > > > >
>>>>>>> > > > > >
>>>>>>> > > > >
>>>>>>> > > > >
>>>>>>> > > >
>>>>>>> > > >
>>>>>>> > >
>>>>>>> > >
>>>>>>> > > --
>>>>>>> > > Julie Prestopnik
>>>>>>> > > National Center for Atmospheric Research
>>>>>>> > > Research Applications Laboratory
>>>>>>> > > Phone: 303.497.8399
>>>>>>> > > Email: jpresto at ucar.edu
>>>>>>> > >
>>>>>>> > >
>>>>>>> >
>>>>>>> >
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Julie Prestopnik
Time: Tue Oct 21 09:46:33 2014
Just wanted to check in. Is there any additional support you need
from us?
Thanks,
Julie
On Mon, Oct 6, 2014 at 4:58 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Supporting my arguments here,
> "wgrib -V WRFPRS_d01.00_2latlon" gives grid information like:
> latlon: lat -89.529000 to 89.529000 by 0.942000 nxny 55008
> long 0.625000 to 359.375000 by 1.250000, (288 x 191) scan
64 mode
> 128
> where, WRFPRS_d01.00_2latlon is copygb output file.
>
> So, any comment on this issue with
> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat on
Yellowstone?
>
> Please let me know.
> Thank you.
>
> Jinwoong Yoo
> UNM
>
>
> On Mon, Oct 6, 2014 at 4:42 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> wrote:
>
> > I'm getting closer to the point of the problem.
> >
> > I have corrected my copygb argument so close as to get this
minor(?)
> error:
> >
> > ERROR : process_command_line() -> The forecast and observation
grids do
> > not match:
> > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: -0.625
> > delta_lat: 0.942 delta_lon: 1.250 !=
> > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625
> delta_lat:
> > 0.942 delta_lon: 1.250
> >
> >
> > My copygb argument was:
> > copygb.exe -g"255 0 288 191 -89529 625 128 89529 359375 1250
942 64"
> > -x $infile $outfile
> >
> > As you may know, 625 in this case is Longitude of origin * 1000
(lon =
> > -180 ... 180). But I don't know how the negative sign got into
the
> > execution of the program, yielding lon_ll: -0.625.
> >
> > Do you know how I should solve this problem?
> > Is this a bug in the program?
> >
> > By the way, as I noted in my previous email,
> > I found and confirmed the fault statements in the MET Online
Tutorial for
> > the copygb Tool: Lat/Lon Grid (
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > )
> >
> > NX should be Number of points on longitude meridian,
> > while NY should be Number of points on latitude circle,
definitely.
> >
> > Thanks.
> >
> > Jinwoong Yoo
> > UNM
> >
> >
> > On Mon, Oct 6, 2014 at 1:45 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > wrote:
> >
> >> Hi.
> >>
> >> I encounter errors while I'm running Grid_Stat on Yellowstone as
below:
> >>
> >> [jyoo at yslogin4 analysis]$
> >> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
> >>
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
> >> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
> >> b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm
-outdir
> >> /glade/scratch/jyoo/domain/mean/analysis/gridstatout
> >> DEBUG 1: Default Config File:
> >>
> /glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
> >> DEBUG 1: User Config File: GridStatConfig_ccsm
> >> WARNING:
> >> WARNING: NcCfFile::open() -> could not extract init time from
file name
> >> WARNING: Using init time of 0
> >> WARNING:
> >> ERROR :
> >> ERROR : process_command_line() -> The forecast and observation
grids do
> >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon
Nx: 288
> Ny:
> >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> >>
> >> WRF output that was postprocessed using UPP and copygb is
compared with
> >> CCSM4 result in NetCDF format here for a test.
> >>
> >> First,
> >> WARNING: NcCfFile::open() -> could not extract init time from
file name
> >> WARNING: Using init time of 0
> >>
> >> Since the CCSM4 file contains 6 hourly data for a month, I set "
level
> >> = [ "(1,*,*)" ];"
> >> for my desired level (surface). But it seems it didn't work.
Setting
> >> "(*,*)" didn't work, neither.
> >>
> >> fcst = {
> >> wind_thresh = [ NA ];
> >>
> >> field = [
> >> {
> >> name = "CPRAT";
> >> level = [ "R277" ];
> >> cat_thresh = [ >0.0, >=5.0 ];
> >> }
> >> ];
> >>
> >> };
> >>
> >> obs = {
> >> wind_thresh = [ NA ];
> >>
> >> field = [
> >> {
> >> name = "PRECC";
> >> level = [ "(1,*,*)" ];
> >> cat_thresh = [ >0.0, >=5.0 ];
> >> }
> >> ];
> >>
> >> };
> >>
> >> Second, obviously it seems my previous copygb was not correct.
> >> ERROR : process_command_line() -> The forecast and observation
grids do
> >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon
Nx: 288
> Ny:
> >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> >>
> >>
> >> My copygy commend was like this:
> >>
> >> */glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
> >> * -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250 64"
\*
> >>
> >> * /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
> >>
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
> >>
> >> Output grid information after executing the copygb looks like
this:
> >> latlon: lat -90.000000 to 90.000000 by 1.250000 nxny 55296
> >> long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan 64
mode 128
> >> bdsgrid 1
> >>
> >> However, there are something strange.
> >> According to what Grid-Stat reads in for the CCSM4 file
(Projection:
> >> Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat:
0.942
> >> delta_lon: 1.250), lat and lon information were flipped compared
to
> what I
> >> get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ; slat =
191
> ;slon
> >> = 288) and the Ny: 191 matches to slat(staggered latitude) not
latitude.
> >>
> >> One thing to note is that my WRF simulation is a tropical channel
> >> simulation covering 45S to 45N while the CCSM4 is global.
> >>
> >> What am I doing wrong with copygb and GridStatConfig file?
> >> Thank you in advance.
> >>
> >> Regards,
> >>
> >> Jinwoong Yoo
> >> UNM
> >>
> >> On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> >> wrote:
> >>
> >>> That's right. I was wrong earlier and the later arguments worked
for my
> >>> regridding of the WRF output in Mercator to Lat-Lon grid.
> >>> Thanks.
> >>>
> >>> Jinwoong Yoo
> >>> UNM
> >>>
> >>> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> >>> wrote:
> >>>
> >>>> Am I misunderstanding the arguments of copygb?
> >>>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
> >>>> out.grb)
> >>>>
> >>>> Should be those numbers for the target grid in Lat/Lon of the
CCSM4 in
> >>>> my case, not for the WRF output values?
> >>>> Such that
> >>>> *NX (*Number of points on latitude circle) : 192
> >>>> *NY (*Number of points on longitude meridian*) : 288*
> >>>> *LAT0 : -90*
> >>>> *LON0 : 0*
> >>>> *LAT1 : 90*
> >>>> *LON1 : 358.75*
> >>>> *DX : 942*
> >>>> *DY : 1250*
> >>>>
> >>>> Please let me know.
> >>>> Thank you.
> >>>>
> >>>> Jinwoong Yoo
> >>>> UNM
> >>>>
> >>>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> >>>> wrote:
> >>>>
> >>>>> Dear John,
> >>>>>
> >>>>> Let me have a quick question about getting correct DX and DY
in
> copygb
> >>>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX
DY 64*"
> >>>>> -x in.grb out.grb).
> >>>>>
> >>>>>
> >>>>> As you may know, grid info in my UPP output is as below:
> >>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
> >>>>> long 0.449000 to -0.449000 by 36.000000 km, (1110 x
315)
> >>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> >>>>>
> >>>>> So my argument should be something like this:
> >>>>> *copygb.exe \*
> >>>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64" *
> >>>>>
> >>>>> However, there are something unclear in the document (
> >>>>>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> >>>>> ).
> >>>>>
> >>>>> NX : Number of points on latitude circle
> >>>>> NY : Number of points on longitude meridian
> >>>>> DX : Longitudinal increment in millidegrees
> >>>>> DY : Latitudinal increment in millidegrees
> >>>>>
> >>>>> You would notice that X,Y are not consistent as they refer to
> lat/lon.
> >>>>> I think flipping X/Y in NX/NY should work so that X represents
> >>>>> longitude.
> >>>>>
> >>>>> That's not a big deal.
> >>>>> Here my own problem goes.
> >>>>>
> >>>>> dx and dy in my WRF simulation are 36 km both in Lat and Lon.
> >>>>> However, if I convert them into millidegrees, DX is 323.515315
and DY
> >>>>> is 287.23175 computationally supposing the earth is a complete
> sphere.
> >>>>> copygb takes only integers, DX is 323 and DY is 287. But, I'm
not
> >>>>> sure if it's right to use the different DX and DY or if I
should use
> the
> >>>>> same number for DX and DY.
> >>>>> Please let me know.
> >>>>>
> >>>>> Thank you.
> >>>>>
> >>>>> Jinwoong Yoo
> >>>>> UNM
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com
> >
> >>>>> wrote:
> >>>>>
> >>>>>> Dear John,
> >>>>>>
> >>>>>> Actually I was looking over copygb when I found your email
received.
> >>>>>> It seems that I have to convert the UPP output again using
> copygb.exe
> >>>>>> to match up the grid of WRF results to the CCSM4's lat/lon
grid.
> >>>>>>
> >>>>>> Isn't there any way to do that in one step? No?
> >>>>>>
> >>>>>> Along with the precipitation, I'd like to compare variables
(paired
> >>>>>> for WRF and CCSM4) below for example:
> >>>>>>
> >>>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V", "QVAPOR","GEOPT"/)
> >>>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
> >>>>>>
> >>>>>>
> >>>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
> >>>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
> >>>>>>
> >>>>>>
> >>>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
> >>>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
> >>>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
> >>>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
> >>>>>>
> >>>>>> Variables in the lists are the actual name of variables in
the CCSM4
> >>>>>> (including CAM and CLM monthly files) files and corresponding
to
> those in
> >>>>>> the WRF-ARW.
> >>>>>>
> >>>>>>
> >>>>>> I'm trying to run Grid-Stat for a test and I'd like to run
> >>>>>> Series-Analysis for the entire dataset of the WRF and the
CCSM4.
> >>>>>> However, I'm concerning about how to run the Series-Analysis
for the
> >>>>>> monthly or annual data using MET?
> >>>>>> At the same time, I am also thinking about using the NCL to
do the
> >>>>>> same statistical analyses. I don't know which one will be
easier
> for me to
> >>>>>> execute for now.
> >>>>>> We'll see.
> >>>>>>
> >>>>>> Will let you know as I make progress, John.
> >>>>>>
> >>>>>> >>> FYI - I just talked to the person who provides support
for UPP
> >>>>>> through
> >>>>>> >>> wrfhelp, and she's expecting an email from you.
> >>>>>>
> >>>>>> I cc'ed this email to wrfhelp.
> >>>>>> Thank you.
> >>>>>>
> >>>>>> Regards,
> >>>>>>
> >>>>>> Jinwoong Yoo
> >>>>>> UNM
> >>>>>>
> >>>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT <
> >>>>>> met_help at ucar.edu> wrote:
> >>>>>>
> >>>>>>> Jinwoong,
> >>>>>>>
> >>>>>>> I'm able to access your data on Yellowstone. I see that
you'd like
> >>>>>>> to
> >>>>>>> compare your UPP output in GRIB format to some data that
follows
> the
> >>>>>>> NetCDF
> >>>>>>> CF-1.0 convention.
> >>>>>>>
> >>>>>>> UPP output:
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> >>>>>>> CCSM4 output:
> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
> >>>>>>>
> >>>>>>> Hopefully you're able to look at those files and identify
which
> GRIB
> >>>>>>> records you'd like to compare to which NetCDF variables.
When you
> >>>>>>> run the
> >>>>>>> MET tools, you'll define the forecast and observation fields
to be
> >>>>>>> used.
> >>>>>>>
> >>>>>>> Looking at your data, I see 2 immediate problems:
> >>>>>>>
> >>>>>>> (1) Gridded forecast and observation data must be on the
same grid
> >>>>>>> before
> >>>>>>> MET can compare them.
> >>>>>>> Here's a description from wgrib of the grid for your GRIB
data:
> >>>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
> >>>>>>> 349650
> >>>>>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
> >>>>>>> 315)
> >>>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> >>>>>>> But it looks like the CCSM4 data is on a lat/lon grid.
> >>>>>>>
> >>>>>>> I'd suggest using the copygb utility to regrid the GRIB1
output
> from
> >>>>>>> UPP to
> >>>>>>> the lat/lon grid of the CCSM4 data. Here's an example of
doing
> that:
> >>>>>>>
> >>>>>>>
> >>>>>>>
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> >>>>>>>
> >>>>>>> (2) There's an issue in the times in the accumulated
precipitation.
> >>>>>>> Try
> >>>>>>> running the following 2 commands:
> >>>>>>> wgrib
> >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
> >>>>>>> grep APCP
> >>>>>>> wgrib
> >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
> >>>>>>> grep APCP
> >>>>>>>
> >>>>>>> For the first, APCP is an accumulation of precip from hours
0 to
> >>>>>>> 252. The
> >>>>>>> second should be 0 to 258, but it's stored as 0 to 2!
That's
> >>>>>>> because UPP
> >>>>>>> is failing to realize that the precipitation timestamps are
> >>>>>>> overflowing the
> >>>>>>> 1 byte allocated for that second time unit. This problem
was
> >>>>>>> addressed in
> >>>>>>> an earlier version of UPP but the problem is still showing
up in
> the
> >>>>>>> version you're using. If you plan to verify precip, I'd
suggest
> >>>>>>> writing
> >>>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're using,
and
> >>>>>>> ask them
> >>>>>>> if this should be fixed.
> >>>>>>>
> >>>>>>> Hope that helps.
> >>>>>>>
> >>>>>>> Thanks,
> >>>>>>> John
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
> >>>>>>> met_help at ucar.edu>
> >>>>>>> wrote:
> >>>>>>>
> >>>>>>> >
> >>>>>>> > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >>>>>>> >
> >>>>>>> > Hi, Julie.
> >>>>>>> > Thank you for heads-up.
> >>>>>>> > I will wait for John to reply then.
> >>>>>>> > Thank you.
> >>>>>>> >
> >>>>>>> > Jinwoong Yoo
> >>>>>>> > UNM
> >>>>>>> >
> >>>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT <
> >>>>>>> > met_help at ucar.edu>
> >>>>>>> > wrote:
> >>>>>>> >
> >>>>>>> > > Hello. John is out of the office, but will be back
later this
> >>>>>>> week. I am
> >>>>>>> > > unable to help you with this task. Hopefully, John will
be
> able
> >>>>>>> to
> >>>>>>> > respond
> >>>>>>> > > to this ticket later this week.
> >>>>>>> > >
> >>>>>>> > > Julie
> >>>>>>> > >
> >>>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
> >>>>>>> met_help at ucar.edu>
> >>>>>>> > > wrote:
> >>>>>>> > >
> >>>>>>> > > >
> >>>>>>> > > > <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>>>>> >
> >>>>>>> > > >
> >>>>>>> > > > Dear John,
> >>>>>>> > > >
> >>>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
> >>>>>>> > > > I assume you can access to my glade scratch directory.
> >>>>>>> > > > My UPP ouputs are located
> >>>>>>> > > > under /glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> >>>>>>> > > >
> >>>>>>> > > > Against those files, I'd like to compare CCSM4
simulation
> >>>>>>> outputs some
> >>>>>>> > of
> >>>>>>> > > > which are located also under
> >>>>>>> > > >
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> >>>>>>> > > >
> >>>>>>> > > > Those are six hourly files all together. If you cannot
access
> >>>>>>> to those
> >>>>>>> > > > files, please let me know.
> >>>>>>> > > >
> >>>>>>> > > > However, there are some variables among the CCSM4
outputs
> that
> >>>>>>> are
> >>>>>>> > > > available only monthly or annually. In this case, how
can I
> >>>>>>> > post-process
> >>>>>>> > > my
> >>>>>>> > > > WRF outputs for the usage of MET through UPP?
> >>>>>>> > > >
> >>>>>>> > > > Thank you.
> >>>>>>> > > >
> >>>>>>> > > > Regards,
> >>>>>>> > > >
> >>>>>>> > > > Jinwoong Yoo
> >>>>>>> > > > UNM
> >>>>>>> > > >
> >>>>>>> > > >
> >>>>>>> > > >
> >>>>>>> > > >
> >>>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway
via RT <
> >>>>>>> > > > met_help at ucar.edu> wrote:
> >>>>>>> > > >
> >>>>>>> > > > > Jinwoo,
> >>>>>>> > > > >
> >>>>>>> > > > > I believe UPP will still run, just using a single
time
> value
> >>>>>>> as if it
> >>>>>>> > > > were
> >>>>>>> > > > > an instantaneous field rather than an accumulation
> >>>>>>> interval. The
> >>>>>>> > > default
> >>>>>>> > > > > for ARW is to store precip as a runtime
accumulation. For
> >>>>>>> forecasts
> >>>>>>> > > > hours
> >>>>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB as
an
> >>>>>>> accumulation
> >>>>>>> > > > from 0
> >>>>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll
be
> stored
> >>>>>>> as an
> >>>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
difference,
> >>>>>>> but
> >>>>>>> > > something
> >>>>>>> > > > > to be aware of when you set up MET config files to
verify
> >>>>>>> the output.
> >>>>>>> > > > >
> >>>>>>> > > > > Give UPP a try and let me know how it goes.
> >>>>>>> > > > >
> >>>>>>> > > > > John
> >>>>>>> > > > >
> >>>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT <
> >>>>>>> > met_help at ucar.edu>
> >>>>>>> > > > > wrote:
> >>>>>>> > > > >
> >>>>>>> > > > > >
> >>>>>>> > > > > > <URL:
> >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >>>>>>> > > > > >
> >>>>>>> > > > > > Dear John,
> >>>>>>> > > > > >
> >>>>>>> > > > > > I see. Let me change "export lastfhr=18" to
"export
> >>>>>>> lastfhr=744".
> >>>>>>> > > > > > Thank you for your comments.
> >>>>>>> > > > > > Yes, I'd like to evaluate precipitation, too, and
my 10
> >>>>>>> year
> >>>>>>> > > simulation
> >>>>>>> > > > > was
> >>>>>>> > > > > > integrated from a single initialization. Then,
will the
> UPP
> >>>>>>> > > run_unipost
> >>>>>>> > > > > > crash after the 255 forecast hours or just will
> >>>>>>> precipitation not
> >>>>>>> > be
> >>>>>>> > > > > stored
> >>>>>>> > > > > > any more aster the 255 forecast hours?
> >>>>>>> > > > > >
> >>>>>>> > > > > > Thank you.
> >>>>>>> > > > > >
> >>>>>>> > > > > > Regards,
> >>>>>>> > > > > >
> >>>>>>> > > > > > Jinwoong Yoo
> >>>>>>> > > > > > UNM
> >>>>>>> > > > > >
> >>>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley
Gotway via
> RT
> >>>>>>> <
> >>>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
> >>>>>>> > > > > >
> >>>>>>> > > > > > > Jinwoong,
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > UPP support is typically provided through
> >>>>>>> wrf_help at ucar.edu
> >>>>>>> > > > > > <javascript:;>. I'll give it
> >>>>>>> > > > > > > a shot, but if this doesn't solve it, I'll refer
you to
> >>>>>>> wrf_help.
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > I assume you're running the script named
"run_unipost".
> >>>>>>> Looking
> >>>>>>> > in
> >>>>>>> > > > > > there,
> >>>>>>> > > > > > > you'll see:
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > export startdate=2005012300
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > export fhr=00
> >>>>>>> > > > > > > export lastfhr=18
> >>>>>>> > > > > > > export incrementhr=03
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > I'd suggest modifying the...
> >>>>>>> > > > > > > - "startdate" to be the model initialization
time to
> be
> >>>>>>> > processed.
> >>>>>>> > > > > > > - Leave "fhr" set to 00 to start at the
analysis time.
> >>>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31
days * 24
> >>>>>>> > hours/day).
> >>>>>>> > > > > > > - Set "incrementhr" to 06 to process output
every 6
> >>>>>>> hours.
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > Give that a shot and let me know if that gets
you
> output
> >>>>>>> past the
> >>>>>>> > > > first
> >>>>>>> > > > > > > day.
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > Are you going to be evaluating precipitation? I
ask
> >>>>>>> that because
> >>>>>>> > > > > precip
> >>>>>>> > > > > > is
> >>>>>>> > > > > > > an accumulated interval, meaning that it has a
> beginning
> >>>>>>> > > accumulation
> >>>>>>> > > > > > time
> >>>>>>> > > > > > > and ending accumulation time. GRIB1 is limited
in the
> >>>>>>> amount of
> >>>>>>> > > > space
> >>>>>>> > > > > > > available for storing times. When you get past
255
> >>>>>>> forecast
> >>>>>>> > hours,
> >>>>>>> > > > you
> >>>>>>> > > > > > can
> >>>>>>> > > > > > > no longer store beginning/ending times for
precip
> >>>>>>> because it runs
> >>>>>>> > > out
> >>>>>>> > > > > of
> >>>>>>> > > > > > > room in the header. Do you have a 10 year
simulation
> >>>>>>> computed
> >>>>>>> > > from a
> >>>>>>> > > > > > > single initialization? Meaning, do you have
forecast
> >>>>>>> lead times
> >>>>>>> > > out
> >>>>>>> > > > to
> >>>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24 hours)?
I'm
> >>>>>>> wondering
> >>>>>>> > if
> >>>>>>> > > > > we'll
> >>>>>>> > > > > > > run into problems formatting those time strings.
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > Thanks,
> >>>>>>> > > > > > > John
> >>>>>>> > > > > > >
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo
via RT <
> >>>>>>> > > > > met_help at ucar.edu
> >>>>>>> > > > > > <javascript:;>>
> >>>>>>> > > > > > > wrote:
> >>>>>>> > > > > > >
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > <URL:
> >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>>>>> > >
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > Dear John,
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > I ran UPP with my multiple wrfout files, first
with
> >>>>>>> the first
> >>>>>>> > > month
> >>>>>>> > > > > of
> >>>>>>> > > > > > > the
> >>>>>>> > > > > > > > 10 year simulation. However, after executing
the
> >>>>>>> run_unipost
> >>>>>>> > > > script,
> >>>>>>> > > > > > UPP
> >>>>>>> > > > > > > > produced only four files: unipost_d01.00.out,
> >>>>>>> > unipost_d01.06.out,
> >>>>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out. It
seems
> >>>>>>> that UPP
> >>>>>>> > > > > processed
> >>>>>>> > > > > > > > only the first day's output among the one
month data.
> >>>>>>> > > > > > > > If you are familiar with UPP, how can I make
sure for
> >>>>>>> UPP to
> >>>>>>> > > > process
> >>>>>>> > > > > > > > multiple wrfout files more than one day?
> >>>>>>> > > > > > > > Thank you.
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > Regards,
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > Jinwoong Yoo
> >>>>>>> > > > > > > > UNM
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong Yoo
<
> >>>>>>> > > > > jinwoong.yoo at gmail.com
> >>>>>>> > > > > > <javascript:;>>
> >>>>>>> > > > > > > > wrote:
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > > > Dear John,
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > > > By the way, what about the CCSM4 file in
NetCDF
> >>>>>>> format?
> >>>>>>> > > > > > > > > Do I need to post process of the CCSM4 files
also
> to
> >>>>>>> run MET
> >>>>>>> > > > along
> >>>>>>> > > > > > with
> >>>>>>> > > > > > > > > the WRF-ARW files?
> >>>>>>> > > > > > > > > Thank you.
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > > > Regards,
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > > > Jinwoong Yoo
> >>>>>>> > > > > > > > > UNM
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong
Yoo <
> >>>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
> >>>>>>> > > > > > > > > wrote:
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > > >> Dear John,
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >> Thank you for your kind email with the
direction.
> >>>>>>> > > > > > > > >> Let me first post process some of my WRF-
ARW
> >>>>>>> outputs using
> >>>>>>> > UPP
> >>>>>>> > > > and
> >>>>>>> > > > > > let
> >>>>>>> > > > > > > > >> you know.
> >>>>>>> > > > > > > > >> Thank you very much.
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >> Regards,
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >> Jinwoong Yoo
> >>>>>>> > > > > > > > >> UNM
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John
Halley
> Gotway
> >>>>>>> via RT <
> >>>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >>> Jinwoong,
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>> I see that you'd like some help running
the
> >>>>>>> series-analysis
> >>>>>>> > > > tool
> >>>>>>> > > > > on
> >>>>>>> > > > > > > > your
> >>>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note
is that
> >>>>>>> MET can't
> >>>>>>> > > read
> >>>>>>> > > > > the
> >>>>>>> > > > > > > raw
> >>>>>>> > > > > > > > >>> wrfout files directly. Instead you first
need to
> >>>>>>> post
> >>>>>>> > > process
> >>>>>>> > > > > them
> >>>>>>> > > > > > > > using
> >>>>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
p_interp
> >>>>>>> utility
> >>>>>>> > > > > (produced
> >>>>>>> > > > > > > > >>> NetCDF
> >>>>>>> > > > > > > > >>> output). After you've post processed some
files,
> >>>>>>> please
> >>>>>>> > send
> >>>>>>> > > > us
> >>>>>>> > > > > > some
> >>>>>>> > > > > > > > >>> sample data by posting it to our anonymous
ftp
> >>>>>>> site. Also
> >>>>>>> > > > send a
> >>>>>>> > > > > > > > >>> description of the comparisons you'd like
to make
> >>>>>>> between
> >>>>>>> > > the 2
> >>>>>>> > > > > > > > datasets.
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>> Here's instructions for sending us data:
> >>>>>>> > > > > > > > >>>
> >>>>>>> http://www.dtcenter.org/met/users/support/met_help.php
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>> Thanks
> >>>>>>> > > > > > > > >>> John Halley Gotway
> >>>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>> On Wednesday, September 24, 2014, Jinwoong
Yoo
> via
> >>>>>>> RT <
> >>>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> wrote:
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request 69099
was
> >>>>>>> acted upon.
> >>>>>>> > > > > > > > >>> > Transaction: Ticket created by
> >>>>>>> jinwoong.yoo at gmail.com
> >>>>>>> > > > > > <javascript:;>
> >>>>>>> > > > > > > <javascript:;>
> >>>>>>> > > > > > > > >>> > Queue: met_help
> >>>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW outputs
and
> >>>>>>> CCSM4 outputs
> >>>>>>> > > > > > > > >>> > Owner: Nobody
> >>>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
> >>>>>>> <javascript:;>
> >>>>>>> > > > > > <javascript:;>
> >>>>>>> > > > > > > > >>> > Status: new
> >>>>>>> > > > > > > > >>> > Ticket <URL:
> >>>>>>> > > > > > > >
> >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Dear Met Help,
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Hi.
> >>>>>>> > > > > > > > >>> > I am looking for a way to verify WRF
model
> >>>>>>> performance
> >>>>>>> > as a
> >>>>>>> > > > > > climate
> >>>>>>> > > > > > > > >>> model.
> >>>>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
climate
> >>>>>>> model
> >>>>>>> > > > > simulations
> >>>>>>> > > > > > > > using
> >>>>>>> > > > > > > > >>> the
> >>>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
> downscaling)
> >>>>>>> for two
> >>>>>>> > > > > > different
> >>>>>>> > > > > > > > time
> >>>>>>> > > > > > > > >>> > periods: one for the LGM and the other
for the
> >>>>>>> modern as
> >>>>>>> > a
> >>>>>>> > > > > > control
> >>>>>>> > > > > > > > run.
> >>>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly
output for
> >>>>>>> ten years
> >>>>>>> > > > > periods
> >>>>>>> > > > > > > > each.
> >>>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
> >>>>>>> outputs.
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Basically, I'd like to compare the two
> simulation
> >>>>>>> > results,
> >>>>>>> > > > > > > verifying
> >>>>>>> > > > > > > > >>> the
> >>>>>>> > > > > > > > >>> > WRF model simulations.
> >>>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
experiments.)
> >>>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems to
do the
> >>>>>>> similar
> >>>>>>> > > jobs
> >>>>>>> > > > > but
> >>>>>>> > > > > > I
> >>>>>>> > > > > > > > need
> >>>>>>> > > > > > > > >>> > much more help from you than the MET
Online
> >>>>>>> Tutorial to
> >>>>>>> > > make
> >>>>>>> > > > a
> >>>>>>> > > > > > step
> >>>>>>> > > > > > > > >>> > forward.
> >>>>>>> > > > > > > > >>> > Any help will be highly appreciated.
> >>>>>>> > > > > > > > >>> > Thank you.
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Regards,
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> > Jinwoong Yoo
> >>>>>>> > > > > > > > >>> > Postdoc
> >>>>>>> > > > > > > > >>> > Department of Earth and Planetary
Sciences
> >>>>>>> > > > > > > > >>> > UNM
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>> >
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>>
> >>>>>>> > > > > > > > >>
> >>>>>>> > > > > > > > >
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > > >
> >>>>>>> > > > > > >
> >>>>>>> > > > > > >
> >>>>>>> > > > > >
> >>>>>>> > > > > >
> >>>>>>> > > > >
> >>>>>>> > > > >
> >>>>>>> > > >
> >>>>>>> > > >
> >>>>>>> > >
> >>>>>>> > >
> >>>>>>> > > --
> >>>>>>> > > Julie Prestopnik
> >>>>>>> > > National Center for Atmospheric Research
> >>>>>>> > > Research Applications Laboratory
> >>>>>>> > > Phone: 303.497.8399
> >>>>>>> > > Email: jpresto at ucar.edu
> >>>>>>> > >
> >>>>>>> > >
> >>>>>>> >
> >>>>>>> >
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>
> >>>>
> >>>
> >>
> >
>
>
--
Julie Prestopnik
National Center for Atmospheric Research
Research Applications Laboratory
Phone: 303.497.8399
Email: jpresto at ucar.edu
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Jinwoong Yoo
Time: Tue Oct 21 11:08:20 2014
Dear Julie,
Not at the present with this issue.
Thank you.
Regards,
Jinwoong Yoo
UNM
On Tue, Oct 21, 2014 at 9:46 AM, Julie Prestopnik via RT
<met_help at ucar.edu>
wrote:
> Just wanted to check in. Is there any additional support you need
from us?
>
> Thanks,
> Julie
>
> On Mon, Oct 6, 2014 at 4:58 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> >
> > Supporting my arguments here,
> > "wgrib -V WRFPRS_d01.00_2latlon" gives grid information like:
> > latlon: lat -89.529000 to 89.529000 by 0.942000 nxny 55008
> > long 0.625000 to 359.375000 by 1.250000, (288 x 191)
scan 64
> mode
> > 128
> > where, WRFPRS_d01.00_2latlon is copygb output file.
> >
> > So, any comment on this issue with
> > /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat on
Yellowstone?
> >
> > Please let me know.
> > Thank you.
> >
> > Jinwoong Yoo
> > UNM
> >
> >
> > On Mon, Oct 6, 2014 at 4:42 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > wrote:
> >
> > > I'm getting closer to the point of the problem.
> > >
> > > I have corrected my copygb argument so close as to get this
minor(?)
> > error:
> > >
> > > ERROR : process_command_line() -> The forecast and observation
grids
> do
> > > not match:
> > > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll:
-0.625
> > > delta_lat: 0.942 delta_lon: 1.250 !=
> > > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll:
0.625
> > delta_lat:
> > > 0.942 delta_lon: 1.250
> > >
> > >
> > > My copygb argument was:
> > > copygb.exe -g"255 0 288 191 -89529 625 128 89529 359375 1250
942 64"
> > > -x $infile $outfile
> > >
> > > As you may know, 625 in this case is Longitude of origin * 1000
(lon =
> > > -180 ... 180). But I don't know how the negative sign got into
the
> > > execution of the program, yielding lon_ll: -0.625.
> > >
> > > Do you know how I should solve this problem?
> > > Is this a bug in the program?
> > >
> > > By the way, as I noted in my previous email,
> > > I found and confirmed the fault statements in the MET Online
Tutorial
> for
> > > the copygb Tool: Lat/Lon Grid (
> > >
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > > )
> > >
> > > NX should be Number of points on longitude meridian,
> > > while NY should be Number of points on latitude circle,
definitely.
> > >
> > > Thanks.
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > >
> > > On Mon, Oct 6, 2014 at 1:45 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > > wrote:
> > >
> > >> Hi.
> > >>
> > >> I encounter errors while I'm running Grid_Stat on Yellowstone
as
> below:
> > >>
> > >> [jyoo at yslogin4 analysis]$
> > >> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
> > >>
> >
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
> > >> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
> > >> b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm
-outdir
> > >> /glade/scratch/jyoo/domain/mean/analysis/gridstatout
> > >> DEBUG 1: Default Config File:
> > >>
> >
> /glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
> > >> DEBUG 1: User Config File: GridStatConfig_ccsm
> > >> WARNING:
> > >> WARNING: NcCfFile::open() -> could not extract init time from
file
> name
> > >> WARNING: Using init time of 0
> > >> WARNING:
> > >> ERROR :
> > >> ERROR : process_command_line() -> The forecast and observation
grids
> do
> > >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> > >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon
Nx: 288
> > Ny:
> > >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> > >>
> > >> WRF output that was postprocessed using UPP and copygb is
compared
> with
> > >> CCSM4 result in NetCDF format here for a test.
> > >>
> > >> First,
> > >> WARNING: NcCfFile::open() -> could not extract init time from
file
> name
> > >> WARNING: Using init time of 0
> > >>
> > >> Since the CCSM4 file contains 6 hourly data for a month, I set
" level
> > >> = [ "(1,*,*)" ];"
> > >> for my desired level (surface). But it seems it didn't work.
Setting
> > >> "(*,*)" didn't work, neither.
> > >>
> > >> fcst = {
> > >> wind_thresh = [ NA ];
> > >>
> > >> field = [
> > >> {
> > >> name = "CPRAT";
> > >> level = [ "R277" ];
> > >> cat_thresh = [ >0.0, >=5.0 ];
> > >> }
> > >> ];
> > >>
> > >> };
> > >>
> > >> obs = {
> > >> wind_thresh = [ NA ];
> > >>
> > >> field = [
> > >> {
> > >> name = "PRECC";
> > >> level = [ "(1,*,*)" ];
> > >> cat_thresh = [ >0.0, >=5.0 ];
> > >> }
> > >> ];
> > >>
> > >> };
> > >>
> > >> Second, obviously it seems my previous copygb was not correct.
> > >> ERROR : process_command_line() -> The forecast and observation
grids
> do
> > >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll: -90.000
lon_ll:
> > >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection: Lat/Lon
Nx: 288
> > Ny:
> > >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> > >>
> > >>
> > >> My copygy commend was like this:
> > >>
> > >> */glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
> > >> * -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250 64"
\*
> > >>
> > >> *
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
> > >>
> >
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
> > >>
> > >> Output grid information after executing the copygb looks like
this:
> > >> latlon: lat -90.000000 to 90.000000 by 1.250000 nxny 55296
> > >> long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan 64
mode
> 128
> > >> bdsgrid 1
> > >>
> > >> However, there are something strange.
> > >> According to what Grid-Stat reads in for the CCSM4 file
(Projection:
> > >> Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625
delta_lat: 0.942
> > >> delta_lon: 1.250), lat and lon information were flipped
compared to
> > what I
> > >> get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ; slat
= 191
> > ;slon
> > >> = 288) and the Ny: 191 matches to slat(staggered latitude) not
> latitude.
> > >>
> > >> One thing to note is that my WRF simulation is a tropical
channel
> > >> simulation covering 45S to 45N while the CCSM4 is global.
> > >>
> > >> What am I doing wrong with copygb and GridStatConfig file?
> > >> Thank you in advance.
> > >>
> > >> Regards,
> > >>
> > >> Jinwoong Yoo
> > >> UNM
> > >>
> > >> On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo
<jinwoong.yoo at gmail.com
> >
> > >> wrote:
> > >>
> > >>> That's right. I was wrong earlier and the later arguments
worked for
> my
> > >>> regridding of the WRF output in Mercator to Lat-Lon grid.
> > >>> Thanks.
> > >>>
> > >>> Jinwoong Yoo
> > >>> UNM
> > >>>
> > >>> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com
> >
> > >>> wrote:
> > >>>
> > >>>> Am I misunderstanding the arguments of copygb?
> > >>>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*" -x
in.grb
> > >>>> out.grb)
> > >>>>
> > >>>> Should be those numbers for the target grid in Lat/Lon of the
CCSM4
> in
> > >>>> my case, not for the WRF output values?
> > >>>> Such that
> > >>>> *NX (*Number of points on latitude circle) : 192
> > >>>> *NY (*Number of points on longitude meridian*) : 288*
> > >>>> *LAT0 : -90*
> > >>>> *LON0 : 0*
> > >>>> *LAT1 : 90*
> > >>>> *LON1 : 358.75*
> > >>>> *DX : 942*
> > >>>> *DY : 1250*
> > >>>>
> > >>>> Please let me know.
> > >>>> Thank you.
> > >>>>
> > >>>> Jinwoong Yoo
> > >>>> UNM
> > >>>>
> > >>>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com>
> > >>>> wrote:
> > >>>>
> > >>>>> Dear John,
> > >>>>>
> > >>>>> Let me have a quick question about getting correct DX and DY
in
> > copygb
> > >>>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX
DY
> 64*"
> > >>>>> -x in.grb out.grb).
> > >>>>>
> > >>>>>
> > >>>>> As you may know, grid info in my UPP output is as below:
> > >>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km nxny
349650
> > >>>>> long 0.449000 to -0.449000 by 36.000000 km, (1110
x 315)
> > >>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> > >>>>>
> > >>>>> So my argument should be something like this:
> > >>>>> *copygb.exe \*
> > >>>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY 64"
*
> > >>>>>
> > >>>>> However, there are something unclear in the document (
> > >>>>>
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > >>>>> ).
> > >>>>>
> > >>>>> NX : Number of points on latitude circle
> > >>>>> NY : Number of points on longitude meridian
> > >>>>> DX : Longitudinal increment in millidegrees
> > >>>>> DY : Latitudinal increment in millidegrees
> > >>>>>
> > >>>>> You would notice that X,Y are not consistent as they refer
to
> > lat/lon.
> > >>>>> I think flipping X/Y in NX/NY should work so that X
represents
> > >>>>> longitude.
> > >>>>>
> > >>>>> That's not a big deal.
> > >>>>> Here my own problem goes.
> > >>>>>
> > >>>>> dx and dy in my WRF simulation are 36 km both in Lat and
Lon.
> > >>>>> However, if I convert them into millidegrees, DX is
323.515315 and
> DY
> > >>>>> is 287.23175 computationally supposing the earth is a
complete
> > sphere.
> > >>>>> copygb takes only integers, DX is 323 and DY is 287. But,
I'm not
> > >>>>> sure if it's right to use the different DX and DY or if I
should
> use
> > the
> > >>>>> same number for DX and DY.
> > >>>>> Please let me know.
> > >>>>>
> > >>>>> Thank you.
> > >>>>>
> > >>>>> Jinwoong Yoo
> > >>>>> UNM
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com
> > >
> > >>>>> wrote:
> > >>>>>
> > >>>>>> Dear John,
> > >>>>>>
> > >>>>>> Actually I was looking over copygb when I found your email
> received.
> > >>>>>> It seems that I have to convert the UPP output again using
> > copygb.exe
> > >>>>>> to match up the grid of WRF results to the CCSM4's lat/lon
grid.
> > >>>>>>
> > >>>>>> Isn't there any way to do that in one step? No?
> > >>>>>>
> > >>>>>> Along with the precipitation, I'd like to compare variables
> (paired
> > >>>>>> for WRF and CCSM4) below for example:
> > >>>>>>
> > >>>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V",
"QVAPOR","GEOPT"/)
> > >>>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
> > >>>>>>
> > >>>>>>
> > >>>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
> > >>>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
> > >>>>>>
> > >>>>>>
> > >>>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
> > >>>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
> > >>>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
> > >>>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
> > >>>>>>
> > >>>>>> Variables in the lists are the actual name of variables in
the
> CCSM4
> > >>>>>> (including CAM and CLM monthly files) files and
corresponding to
> > those in
> > >>>>>> the WRF-ARW.
> > >>>>>>
> > >>>>>>
> > >>>>>> I'm trying to run Grid-Stat for a test and I'd like to run
> > >>>>>> Series-Analysis for the entire dataset of the WRF and the
CCSM4.
> > >>>>>> However, I'm concerning about how to run the Series-
Analysis for
> the
> > >>>>>> monthly or annual data using MET?
> > >>>>>> At the same time, I am also thinking about using the NCL to
do the
> > >>>>>> same statistical analyses. I don't know which one will be
easier
> > for me to
> > >>>>>> execute for now.
> > >>>>>> We'll see.
> > >>>>>>
> > >>>>>> Will let you know as I make progress, John.
> > >>>>>>
> > >>>>>> >>> FYI - I just talked to the person who provides support
for UPP
> > >>>>>> through
> > >>>>>> >>> wrfhelp, and she's expecting an email from you.
> > >>>>>>
> > >>>>>> I cc'ed this email to wrfhelp.
> > >>>>>> Thank you.
> > >>>>>>
> > >>>>>> Regards,
> > >>>>>>
> > >>>>>> Jinwoong Yoo
> > >>>>>> UNM
> > >>>>>>
> > >>>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via RT
<
> > >>>>>> met_help at ucar.edu> wrote:
> > >>>>>>
> > >>>>>>> Jinwoong,
> > >>>>>>>
> > >>>>>>> I'm able to access your data on Yellowstone. I see that
you'd
> like
> > >>>>>>> to
> > >>>>>>> compare your UPP output in GRIB format to some data that
follows
> > the
> > >>>>>>> NetCDF
> > >>>>>>> CF-1.0 convention.
> > >>>>>>>
> > >>>>>>> UPP output:
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> > >>>>>>> CCSM4 output:
> > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
> > >>>>>>>
> > >>>>>>> Hopefully you're able to look at those files and identify
which
> > GRIB
> > >>>>>>> records you'd like to compare to which NetCDF variables.
When
> you
> > >>>>>>> run the
> > >>>>>>> MET tools, you'll define the forecast and observation
fields to
> be
> > >>>>>>> used.
> > >>>>>>>
> > >>>>>>> Looking at your data, I see 2 immediate problems:
> > >>>>>>>
> > >>>>>>> (1) Gridded forecast and observation data must be on the
same
> grid
> > >>>>>>> before
> > >>>>>>> MET can compare them.
> > >>>>>>> Here's a description from wgrib of the grid for your
GRIB
> data:
> > >>>>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000
km
> nxny
> > >>>>>>> 349650
> > >>>>>>> long 0.449000 to -0.449000 by 36.000000
km,
> (1110 x
> > >>>>>>> 315)
> > >>>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> > >>>>>>> But it looks like the CCSM4 data is on a lat/lon grid.
> > >>>>>>>
> > >>>>>>> I'd suggest using the copygb utility to regrid the GRIB1
output
> > from
> > >>>>>>> UPP to
> > >>>>>>> the lat/lon grid of the CCSM4 data. Here's an example of
doing
> > that:
> > >>>>>>>
> > >>>>>>>
> > >>>>>>>
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > >>>>>>>
> > >>>>>>> (2) There's an issue in the times in the accumulated
> precipitation.
> > >>>>>>> Try
> > >>>>>>> running the following 2 commands:
> > >>>>>>> wgrib
> > >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252 |
> > >>>>>>> grep APCP
> > >>>>>>> wgrib
> > >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258 |
> > >>>>>>> grep APCP
> > >>>>>>>
> > >>>>>>> For the first, APCP is an accumulation of precip from
hours 0 to
> > >>>>>>> 252. The
> > >>>>>>> second should be 0 to 258, but it's stored as 0 to 2!
That's
> > >>>>>>> because UPP
> > >>>>>>> is failing to realize that the precipitation timestamps
are
> > >>>>>>> overflowing the
> > >>>>>>> 1 byte allocated for that second time unit. This problem
was
> > >>>>>>> addressed in
> > >>>>>>> an earlier version of UPP but the problem is still showing
up in
> > the
> > >>>>>>> version you're using. If you plan to verify precip, I'd
suggest
> > >>>>>>> writing
> > >>>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're
using, and
> > >>>>>>> ask them
> > >>>>>>> if this should be fixed.
> > >>>>>>>
> > >>>>>>> Hope that helps.
> > >>>>>>>
> > >>>>>>> Thanks,
> > >>>>>>> John
> > >>>>>>>
> > >>>>>>>
> > >>>>>>>
> > >>>>>>>
> > >>>>>>>
> > >>>>>>>
> > >>>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
> > >>>>>>> met_help at ucar.edu>
> > >>>>>>> wrote:
> > >>>>>>>
> > >>>>>>> >
> > >>>>>>> > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> >
> > >>>>>>> >
> > >>>>>>> > Hi, Julie.
> > >>>>>>> > Thank you for heads-up.
> > >>>>>>> > I will wait for John to reply then.
> > >>>>>>> > Thank you.
> > >>>>>>> >
> > >>>>>>> > Jinwoong Yoo
> > >>>>>>> > UNM
> > >>>>>>> >
> > >>>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via RT
<
> > >>>>>>> > met_help at ucar.edu>
> > >>>>>>> > wrote:
> > >>>>>>> >
> > >>>>>>> > > Hello. John is out of the office, but will be back
later
> this
> > >>>>>>> week. I am
> > >>>>>>> > > unable to help you with this task. Hopefully, John
will be
> > able
> > >>>>>>> to
> > >>>>>>> > respond
> > >>>>>>> > > to this ticket later this week.
> > >>>>>>> > >
> > >>>>>>> > > Julie
> > >>>>>>> > >
> > >>>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT <
> > >>>>>>> met_help at ucar.edu>
> > >>>>>>> > > wrote:
> > >>>>>>> > >
> > >>>>>>> > > >
> > >>>>>>> > > > <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >>>>>>> >
> > >>>>>>> > > >
> > >>>>>>> > > > Dear John,
> > >>>>>>> > > >
> > >>>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
> > >>>>>>> > > > I assume you can access to my glade scratch
directory.
> > >>>>>>> > > > My UPP ouputs are located
> > >>>>>>> > > > under
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> > >>>>>>> > > >
> > >>>>>>> > > > Against those files, I'd like to compare CCSM4
simulation
> > >>>>>>> outputs some
> > >>>>>>> > of
> > >>>>>>> > > > which are located also under
> > >>>>>>> > > >
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> > >>>>>>> > > >
> > >>>>>>> > > > Those are six hourly files all together. If you
cannot
> access
> > >>>>>>> to those
> > >>>>>>> > > > files, please let me know.
> > >>>>>>> > > >
> > >>>>>>> > > > However, there are some variables among the CCSM4
outputs
> > that
> > >>>>>>> are
> > >>>>>>> > > > available only monthly or annually. In this case,
how can I
> > >>>>>>> > post-process
> > >>>>>>> > > my
> > >>>>>>> > > > WRF outputs for the usage of MET through UPP?
> > >>>>>>> > > >
> > >>>>>>> > > > Thank you.
> > >>>>>>> > > >
> > >>>>>>> > > > Regards,
> > >>>>>>> > > >
> > >>>>>>> > > > Jinwoong Yoo
> > >>>>>>> > > > UNM
> > >>>>>>> > > >
> > >>>>>>> > > >
> > >>>>>>> > > >
> > >>>>>>> > > >
> > >>>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley Gotway
via
> RT <
> > >>>>>>> > > > met_help at ucar.edu> wrote:
> > >>>>>>> > > >
> > >>>>>>> > > > > Jinwoo,
> > >>>>>>> > > > >
> > >>>>>>> > > > > I believe UPP will still run, just using a single
time
> > value
> > >>>>>>> as if it
> > >>>>>>> > > > were
> > >>>>>>> > > > > an instantaneous field rather than an accumulation
> > >>>>>>> interval. The
> > >>>>>>> > > default
> > >>>>>>> > > > > for ARW is to store precip as a runtime
accumulation.
> For
> > >>>>>>> forecasts
> > >>>>>>> > > > hours
> > >>>>>>> > > > > < 255, like 12, I suspect it'll be stored in GRIB
as an
> > >>>>>>> accumulation
> > >>>>>>> > > > from 0
> > >>>>>>> > > > > to 12. For hours > 255, like 300, I suspect it'll
be
> > stored
> > >>>>>>> as an
> > >>>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
> difference,
> > >>>>>>> but
> > >>>>>>> > > something
> > >>>>>>> > > > > to be aware of when you set up MET config files to
verify
> > >>>>>>> the output.
> > >>>>>>> > > > >
> > >>>>>>> > > > > Give UPP a try and let me know how it goes.
> > >>>>>>> > > > >
> > >>>>>>> > > > > John
> > >>>>>>> > > > >
> > >>>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via RT
<
> > >>>>>>> > met_help at ucar.edu>
> > >>>>>>> > > > > wrote:
> > >>>>>>> > > > >
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > <URL:
> > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > Dear John,
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > I see. Let me change "export lastfhr=18" to
"export
> > >>>>>>> lastfhr=744".
> > >>>>>>> > > > > > Thank you for your comments.
> > >>>>>>> > > > > > Yes, I'd like to evaluate precipitation, too,
and my 10
> > >>>>>>> year
> > >>>>>>> > > simulation
> > >>>>>>> > > > > was
> > >>>>>>> > > > > > integrated from a single initialization. Then,
will the
> > UPP
> > >>>>>>> > > run_unipost
> > >>>>>>> > > > > > crash after the 255 forecast hours or just will
> > >>>>>>> precipitation not
> > >>>>>>> > be
> > >>>>>>> > > > > stored
> > >>>>>>> > > > > > any more aster the 255 forecast hours?
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > Thank you.
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > Regards,
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > Jinwoong Yoo
> > >>>>>>> > > > > > UNM
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley
Gotway via
> > RT
> > >>>>>>> <
> > >>>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
> > >>>>>>> > > > > >
> > >>>>>>> > > > > > > Jinwoong,
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > UPP support is typically provided through
> > >>>>>>> wrf_help at ucar.edu
> > >>>>>>> > > > > > <javascript:;>. I'll give it
> > >>>>>>> > > > > > > a shot, but if this doesn't solve it, I'll
refer you
> to
> > >>>>>>> wrf_help.
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > I assume you're running the script named
> "run_unipost".
> > >>>>>>> Looking
> > >>>>>>> > in
> > >>>>>>> > > > > > there,
> > >>>>>>> > > > > > > you'll see:
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > export startdate=2005012300
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > export fhr=00
> > >>>>>>> > > > > > > export lastfhr=18
> > >>>>>>> > > > > > > export incrementhr=03
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > I'd suggest modifying the...
> > >>>>>>> > > > > > > - "startdate" to be the model initialization
time to
> > be
> > >>>>>>> > processed.
> > >>>>>>> > > > > > > - Leave "fhr" set to 00 to start at the
analysis
> time.
> > >>>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is 31
days *
> 24
> > >>>>>>> > hours/day).
> > >>>>>>> > > > > > > - Set "incrementhr" to 06 to process output
every 6
> > >>>>>>> hours.
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > Give that a shot and let me know if that gets
you
> > output
> > >>>>>>> past the
> > >>>>>>> > > > first
> > >>>>>>> > > > > > > day.
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > Are you going to be evaluating precipitation?
I ask
> > >>>>>>> that because
> > >>>>>>> > > > > precip
> > >>>>>>> > > > > > is
> > >>>>>>> > > > > > > an accumulated interval, meaning that it has a
> > beginning
> > >>>>>>> > > accumulation
> > >>>>>>> > > > > > time
> > >>>>>>> > > > > > > and ending accumulation time. GRIB1 is
limited in
> the
> > >>>>>>> amount of
> > >>>>>>> > > > space
> > >>>>>>> > > > > > > available for storing times. When you get
past 255
> > >>>>>>> forecast
> > >>>>>>> > hours,
> > >>>>>>> > > > you
> > >>>>>>> > > > > > can
> > >>>>>>> > > > > > > no longer store beginning/ending times for
precip
> > >>>>>>> because it runs
> > >>>>>>> > > out
> > >>>>>>> > > > > of
> > >>>>>>> > > > > > > room in the header. Do you have a 10 year
simulation
> > >>>>>>> computed
> > >>>>>>> > > from a
> > >>>>>>> > > > > > > single initialization? Meaning, do you have
forecast
> > >>>>>>> lead times
> > >>>>>>> > > out
> > >>>>>>> > > > to
> > >>>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24
hours)? I'm
> > >>>>>>> wondering
> > >>>>>>> > if
> > >>>>>>> > > > > we'll
> > >>>>>>> > > > > > > run into problems formatting those time
strings.
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > Thanks,
> > >>>>>>> > > > > > > John
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong Yoo
via RT
> <
> > >>>>>>> > > > > met_help at ucar.edu
> > >>>>>>> > > > > > <javascript:;>>
> > >>>>>>> > > > > > > wrote:
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > <URL:
> > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >>>>>>> > >
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > Dear John,
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > I ran UPP with my multiple wrfout files,
first with
> > >>>>>>> the first
> > >>>>>>> > > month
> > >>>>>>> > > > > of
> > >>>>>>> > > > > > > the
> > >>>>>>> > > > > > > > 10 year simulation. However, after executing
the
> > >>>>>>> run_unipost
> > >>>>>>> > > > script,
> > >>>>>>> > > > > > UPP
> > >>>>>>> > > > > > > > produced only four files:
unipost_d01.00.out,
> > >>>>>>> > unipost_d01.06.out,
> > >>>>>>> > > > > > > > unipost_d01.12.out, and unipost_d01.18.out.
It
> seems
> > >>>>>>> that UPP
> > >>>>>>> > > > > processed
> > >>>>>>> > > > > > > > only the first day's output among the one
month
> data.
> > >>>>>>> > > > > > > > If you are familiar with UPP, how can I make
sure
> for
> > >>>>>>> UPP to
> > >>>>>>> > > > process
> > >>>>>>> > > > > > > > multiple wrfout files more than one day?
> > >>>>>>> > > > > > > > Thank you.
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > Regards,
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > Jinwoong Yoo
> > >>>>>>> > > > > > > > UNM
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong
Yoo <
> > >>>>>>> > > > > jinwoong.yoo at gmail.com
> > >>>>>>> > > > > > <javascript:;>>
> > >>>>>>> > > > > > > > wrote:
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > > > Dear John,
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > > > By the way, what about the CCSM4 file in
NetCDF
> > >>>>>>> format?
> > >>>>>>> > > > > > > > > Do I need to post process of the CCSM4
files also
> > to
> > >>>>>>> run MET
> > >>>>>>> > > > along
> > >>>>>>> > > > > > with
> > >>>>>>> > > > > > > > > the WRF-ARW files?
> > >>>>>>> > > > > > > > > Thank you.
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > > > Regards,
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > > > Jinwoong Yoo
> > >>>>>>> > > > > > > > > UNM
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM, Jinwoong
Yoo <
> > >>>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
> > >>>>>>> > > > > > > > > wrote:
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > > >> Dear John,
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >> Thank you for your kind email with the
> direction.
> > >>>>>>> > > > > > > > >> Let me first post process some of my WRF-
ARW
> > >>>>>>> outputs using
> > >>>>>>> > UPP
> > >>>>>>> > > > and
> > >>>>>>> > > > > > let
> > >>>>>>> > > > > > > > >> you know.
> > >>>>>>> > > > > > > > >> Thank you very much.
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >> Regards,
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >> Jinwoong Yoo
> > >>>>>>> > > > > > > > >> UNM
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John
Halley
> > Gotway
> > >>>>>>> via RT <
> > >>>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>> wrote:
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >>> Jinwoong,
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>> I see that you'd like some help running
the
> > >>>>>>> series-analysis
> > >>>>>>> > > > tool
> > >>>>>>> > > > > on
> > >>>>>>> > > > > > > > your
> > >>>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to note
is
> that
> > >>>>>>> MET can't
> > >>>>>>> > > read
> > >>>>>>> > > > > the
> > >>>>>>> > > > > > > raw
> > >>>>>>> > > > > > > > >>> wrfout files directly. Instead you
first need
> to
> > >>>>>>> post
> > >>>>>>> > > process
> > >>>>>>> > > > > them
> > >>>>>>> > > > > > > > using
> > >>>>>>> > > > > > > > >>> either UPP (produces GRIB output) or the
> p_interp
> > >>>>>>> utility
> > >>>>>>> > > > > (produced
> > >>>>>>> > > > > > > > >>> NetCDF
> > >>>>>>> > > > > > > > >>> output). After you've post processed
some
> files,
> > >>>>>>> please
> > >>>>>>> > send
> > >>>>>>> > > > us
> > >>>>>>> > > > > > some
> > >>>>>>> > > > > > > > >>> sample data by posting it to our
anonymous ftp
> > >>>>>>> site. Also
> > >>>>>>> > > > send a
> > >>>>>>> > > > > > > > >>> description of the comparisons you'd
like to
> make
> > >>>>>>> between
> > >>>>>>> > > the 2
> > >>>>>>> > > > > > > > datasets.
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>> Here's instructions for sending us data:
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> http://www.dtcenter.org/met/users/support/met_help.php
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>> Thanks
> > >>>>>>> > > > > > > > >>> John Halley Gotway
> > >>>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>> On Wednesday, September 24, 2014,
Jinwoong Yoo
> > via
> > >>>>>>> RT <
> > >>>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> wrote:
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request
69099 was
> > >>>>>>> acted upon.
> > >>>>>>> > > > > > > > >>> > Transaction: Ticket created by
> > >>>>>>> jinwoong.yoo at gmail.com
> > >>>>>>> > > > > > <javascript:;>
> > >>>>>>> > > > > > > <javascript:;>
> > >>>>>>> > > > > > > > >>> > Queue: met_help
> > >>>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW
outputs and
> > >>>>>>> CCSM4 outputs
> > >>>>>>> > > > > > > > >>> > Owner: Nobody
> > >>>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
> > >>>>>>> <javascript:;>
> > >>>>>>> > > > > > <javascript:;>
> > >>>>>>> > > > > > > > >>> > Status: new
> > >>>>>>> > > > > > > > >>> > Ticket <URL:
> > >>>>>>> > > > > > > >
> > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Dear Met Help,
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Hi.
> > >>>>>>> > > > > > > > >>> > I am looking for a way to verify WRF
model
> > >>>>>>> performance
> > >>>>>>> > as a
> > >>>>>>> > > > > > climate
> > >>>>>>> > > > > > > > >>> model.
> > >>>>>>> > > > > > > > >>> > I'm completed 10-year tropical channel
> climate
> > >>>>>>> model
> > >>>>>>> > > > > simulations
> > >>>>>>> > > > > > > > using
> > >>>>>>> > > > > > > > >>> the
> > >>>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
> > downscaling)
> > >>>>>>> for two
> > >>>>>>> > > > > > different
> > >>>>>>> > > > > > > > time
> > >>>>>>> > > > > > > > >>> > periods: one for the LGM and the other
for
> the
> > >>>>>>> modern as
> > >>>>>>> > a
> > >>>>>>> > > > > > control
> > >>>>>>> > > > > > > > run.
> > >>>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly
output
> for
> > >>>>>>> ten years
> > >>>>>>> > > > > periods
> > >>>>>>> > > > > > > > each.
> > >>>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with 6
hourly
> > >>>>>>> outputs.
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Basically, I'd like to compare the two
> > simulation
> > >>>>>>> > results,
> > >>>>>>> > > > > > > verifying
> > >>>>>>> > > > > > > > >>> the
> > >>>>>>> > > > > > > > >>> > WRF model simulations.
> > >>>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
> experiments.)
> > >>>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0 seems
to do
> the
> > >>>>>>> similar
> > >>>>>>> > > jobs
> > >>>>>>> > > > > but
> > >>>>>>> > > > > > I
> > >>>>>>> > > > > > > > need
> > >>>>>>> > > > > > > > >>> > much more help from you than the MET
Online
> > >>>>>>> Tutorial to
> > >>>>>>> > > make
> > >>>>>>> > > > a
> > >>>>>>> > > > > > step
> > >>>>>>> > > > > > > > >>> > forward.
> > >>>>>>> > > > > > > > >>> > Any help will be highly appreciated.
> > >>>>>>> > > > > > > > >>> > Thank you.
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Regards,
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> > Jinwoong Yoo
> > >>>>>>> > > > > > > > >>> > Postdoc
> > >>>>>>> > > > > > > > >>> > Department of Earth and Planetary
Sciences
> > >>>>>>> > > > > > > > >>> > UNM
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>> >
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>>
> > >>>>>>> > > > > > > > >>
> > >>>>>>> > > > > > > > >
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > > >
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > > >
> > >>>>>>> > > > > >
> > >>>>>>> > > > > >
> > >>>>>>> > > > >
> > >>>>>>> > > > >
> > >>>>>>> > > >
> > >>>>>>> > > >
> > >>>>>>> > >
> > >>>>>>> > >
> > >>>>>>> > > --
> > >>>>>>> > > Julie Prestopnik
> > >>>>>>> > > National Center for Atmospheric Research
> > >>>>>>> > > Research Applications Laboratory
> > >>>>>>> > > Phone: 303.497.8399
> > >>>>>>> > > Email: jpresto at ucar.edu
> > >>>>>>> > >
> > >>>>>>> > >
> > >>>>>>> >
> > >>>>>>> >
> > >>>>>>>
> > >>>>>>>
> > >>>>>>
> > >>>>>
> > >>>>
> > >>>
> > >>
> > >
> >
> >
>
>
> --
> Julie Prestopnik
> National Center for Atmospheric Research
> Research Applications Laboratory
> Phone: 303.497.8399
> Email: jpresto at ucar.edu
>
>
------------------------------------------------
Subject: Comparing WRF-ARW outputs and CCSM4 outputs
From: Julie Prestopnik
Time: Tue Oct 21 12:08:54 2014
Great! Thank you. I will resolve this ticket.
Julie
On Tue, Oct 21, 2014 at 11:08 AM, Jinwoong Yoo via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
>
> Dear Julie,
>
> Not at the present with this issue.
> Thank you.
>
> Regards,
>
> Jinwoong Yoo
> UNM
>
> On Tue, Oct 21, 2014 at 9:46 AM, Julie Prestopnik via RT <
> met_help at ucar.edu>
> wrote:
>
> > Just wanted to check in. Is there any additional support you need
from
> us?
> >
> > Thanks,
> > Julie
> >
> > On Mon, Oct 6, 2014 at 4:58 PM, Jinwoong Yoo via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099 >
> > >
> > > Supporting my arguments here,
> > > "wgrib -V WRFPRS_d01.00_2latlon" gives grid information like:
> > > latlon: lat -89.529000 to 89.529000 by 0.942000 nxny 55008
> > > long 0.625000 to 359.375000 by 1.250000, (288 x 191)
scan 64
> > mode
> > > 128
> > > where, WRFPRS_d01.00_2latlon is copygb output file.
> > >
> > > So, any comment on this issue with
> > > /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat on
Yellowstone?
> > >
> > > Please let me know.
> > > Thank you.
> > >
> > > Jinwoong Yoo
> > > UNM
> > >
> > >
> > > On Mon, Oct 6, 2014 at 4:42 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com>
> > > wrote:
> > >
> > > > I'm getting closer to the point of the problem.
> > > >
> > > > I have corrected my copygb argument so close as to get this
minor(?)
> > > error:
> > > >
> > > > ERROR : process_command_line() -> The forecast and
observation grids
> > do
> > > > not match:
> > > > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll:
-0.625
> > > > delta_lat: 0.942 delta_lon: 1.250 !=
> > > > Projection: Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll:
0.625
> > > delta_lat:
> > > > 0.942 delta_lon: 1.250
> > > >
> > > >
> > > > My copygb argument was:
> > > > copygb.exe -g"255 0 288 191 -89529 625 128 89529 359375
1250 942
> 64"
> > > > -x $infile $outfile
> > > >
> > > > As you may know, 625 in this case is Longitude of origin *
1000 (lon
> =
> > > > -180 ... 180). But I don't know how the negative sign got
into the
> > > > execution of the program, yielding lon_ll: -0.625.
> > > >
> > > > Do you know how I should solve this problem?
> > > > Is this a bug in the program?
> > > >
> > > > By the way, as I noted in my previous email,
> > > > I found and confirmed the fault statements in the MET Online
Tutorial
> > for
> > > > the copygb Tool: Lat/Lon Grid (
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > > > )
> > > >
> > > > NX should be Number of points on longitude meridian,
> > > > while NY should be Number of points on latitude circle,
definitely.
> > > >
> > > > Thanks.
> > > >
> > > > Jinwoong Yoo
> > > > UNM
> > > >
> > > >
> > > > On Mon, Oct 6, 2014 at 1:45 PM, Jinwoong Yoo
<jinwoong.yoo at gmail.com
> >
> > > > wrote:
> > > >
> > > >> Hi.
> > > >>
> > > >> I encounter errors while I'm running Grid_Stat on Yellowstone
as
> > below:
> > > >>
> > > >> [jyoo at yslogin4 analysis]$
> > > >> /glade/p/ral/jnt/MET/MET_releases/met-5.0/bin/grid_stat
> > > >>
> > >
> >
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/outcopygb/WRFPRS_d01.00_latlon
> > > >> /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/
> > > >> b40.lgm21ka.1deg.003M.cam2.h3.1870-01.nc GridStatConfig_ccsm
> -outdir
> > > >> /glade/scratch/jyoo/domain/mean/analysis/gridstatout
> > > >> DEBUG 1: Default Config File:
> > > >>
> > >
> >
> /glade/p/ral/jnt/MET/MET_releases/met-
5.0/share/met/config/GridStatConfig_default
> > > >> DEBUG 1: User Config File: GridStatConfig_ccsm
> > > >> WARNING:
> > > >> WARNING: NcCfFile::open() -> could not extract init time from
file
> > name
> > > >> WARNING: Using init time of 0
> > > >> WARNING:
> > > >> ERROR :
> > > >> ERROR : process_command_line() -> The forecast and
observation
> grids
> > do
> > > >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll:
-90.000
> lon_ll:
> > > >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection:
Lat/Lon Nx:
> 288
> > > Ny:
> > > >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> > > >>
> > > >> WRF output that was postprocessed using UPP and copygb is
compared
> > with
> > > >> CCSM4 result in NetCDF format here for a test.
> > > >>
> > > >> First,
> > > >> WARNING: NcCfFile::open() -> could not extract init time from
file
> > name
> > > >> WARNING: Using init time of 0
> > > >>
> > > >> Since the CCSM4 file contains 6 hourly data for a month, I
set "
> level
> > > >> = [ "(1,*,*)" ];"
> > > >> for my desired level (surface). But it seems it didn't work.
> Setting
> > > >> "(*,*)" didn't work, neither.
> > > >>
> > > >> fcst = {
> > > >> wind_thresh = [ NA ];
> > > >>
> > > >> field = [
> > > >> {
> > > >> name = "CPRAT";
> > > >> level = [ "R277" ];
> > > >> cat_thresh = [ >0.0, >=5.0 ];
> > > >> }
> > > >> ];
> > > >>
> > > >> };
> > > >>
> > > >> obs = {
> > > >> wind_thresh = [ NA ];
> > > >>
> > > >> field = [
> > > >> {
> > > >> name = "PRECC";
> > > >> level = [ "(1,*,*)" ];
> > > >> cat_thresh = [ >0.0, >=5.0 ];
> > > >> }
> > > >> ];
> > > >>
> > > >> };
> > > >>
> > > >> Second, obviously it seems my previous copygb was not
correct.
> > > >> ERROR : process_command_line() -> The forecast and
observation
> grids
> > do
> > > >> not match: Projection: Lat/Lon Nx: 192 Ny: 288 lat_ll:
-90.000
> lon_ll:
> > > >> 0.000 delta_lat: 1.250 delta_lon: 0.942 != Projection:
Lat/Lon Nx:
> 288
> > > Ny:
> > > >> 191 lat_ll: -89.529 lon_ll: 0.625 delta_lat: 0.942 delta_lon:
1.250
> > > >>
> > > >>
> > > >> My copygy commend was like this:
> > > >>
> > > >> */glade/scratch/jyoo/UPPV2.2/bin/copygb.exe \*
> > > >> * -xg"255 0 *192 288 -90000 0 * 128 *90000 358750 *942 1250
64" \*
> > > >>
> > > >> *
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/WRFPRS_d01.00\
> > > >>
> > >
> >
>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/copygbout/WRFPRS_d01.00_latlon*
> > > >>
> > > >> Output grid information after executing the copygb looks like
this:
> > > >> latlon: lat -90.000000 to 90.000000 by 1.250000 nxny
55296
> > > >> long 0.000000 to 358.750000 by 0.942000, (192 x 288) scan
64 mode
> > 128
> > > >> bdsgrid 1
> > > >>
> > > >> However, there are something strange.
> > > >> According to what Grid-Stat reads in for the CCSM4 file
(Projection:
> > > >> Lat/Lon Nx: 288 Ny: 191 lat_ll: -89.529 lon_ll: 0.625
delta_lat:
> 0.942
> > > >> delta_lon: 1.250), lat and lon information were flipped
compared to
> > > what I
> > > >> get from ncdump of the CCSM4 file (lat = 192 ; lon = 288 ;
slat =
> 191
> > > ;slon
> > > >> = 288) and the Ny: 191 matches to slat(staggered latitude)
not
> > latitude.
> > > >>
> > > >> One thing to note is that my WRF simulation is a tropical
channel
> > > >> simulation covering 45S to 45N while the CCSM4 is global.
> > > >>
> > > >> What am I doing wrong with copygb and GridStatConfig file?
> > > >> Thank you in advance.
> > > >>
> > > >> Regards,
> > > >>
> > > >> Jinwoong Yoo
> > > >> UNM
> > > >>
> > > >> On Fri, Oct 3, 2014 at 11:19 AM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com
> > >
> > > >> wrote:
> > > >>
> > > >>> That's right. I was wrong earlier and the later arguments
worked
> for
> > my
> > > >>> regridding of the WRF output in Mercator to Lat-Lon grid.
> > > >>> Thanks.
> > > >>>
> > > >>> Jinwoong Yoo
> > > >>> UNM
> > > >>>
> > > >>> On Thu, Oct 2, 2014 at 5:50 PM, Jinwoong Yoo <
> jinwoong.yoo at gmail.com
> > >
> > > >>> wrote:
> > > >>>
> > > >>>> Am I misunderstanding the arguments of copygb?
> > > >>>> (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1 DX DY 64*"
-x
> in.grb
> > > >>>> out.grb)
> > > >>>>
> > > >>>> Should be those numbers for the target grid in Lat/Lon of
the
> CCSM4
> > in
> > > >>>> my case, not for the WRF output values?
> > > >>>> Such that
> > > >>>> *NX (*Number of points on latitude circle) : 192
> > > >>>> *NY (*Number of points on longitude meridian*) : 288*
> > > >>>> *LAT0 : -90*
> > > >>>> *LON0 : 0*
> > > >>>> *LAT1 : 90*
> > > >>>> *LON1 : 358.75*
> > > >>>> *DX : 942*
> > > >>>> *DY : 1250*
> > > >>>>
> > > >>>> Please let me know.
> > > >>>> Thank you.
> > > >>>>
> > > >>>> Jinwoong Yoo
> > > >>>> UNM
> > > >>>>
> > > >>>> On Thu, Oct 2, 2014 at 4:51 PM, Jinwoong Yoo <
> > jinwoong.yoo at gmail.com>
> > > >>>> wrote:
> > > >>>>
> > > >>>>> Dear John,
> > > >>>>>
> > > >>>>> Let me have a quick question about getting correct DX and
DY in
> > > copygb
> > > >>>>> arguments (copygb -g"*255 0 NX NY LAT0 LON0 128 LAT1 LON1
DX DY
> > 64*"
> > > >>>>> -x in.grb out.grb).
> > > >>>>>
> > > >>>>>
> > > >>>>> As you may know, grid info in my UPP output is as below:
> > > >>>>> Mercator: lat -45.239000 to 45.239000 by 36.000000 km
nxny
> 349650
> > > >>>>> long 0.449000 to -0.449000 by 36.000000 km,
(1110 x
> 315)
> > > >>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> > > >>>>>
> > > >>>>> So my argument should be something like this:
> > > >>>>> *copygb.exe \*
> > > >>>>> * -g"255 0 *315 1110 -45239 449 * 128 *45239 -449 *DX DY
64" *
> > > >>>>>
> > > >>>>> However, there are something unclear in the document (
> > > >>>>>
> > >
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > > >>>>> ).
> > > >>>>>
> > > >>>>> NX : Number of points on latitude circle
> > > >>>>> NY : Number of points on longitude meridian
> > > >>>>> DX : Longitudinal increment in millidegrees
> > > >>>>> DY : Latitudinal increment in millidegrees
> > > >>>>>
> > > >>>>> You would notice that X,Y are not consistent as they refer
to
> > > lat/lon.
> > > >>>>> I think flipping X/Y in NX/NY should work so that X
represents
> > > >>>>> longitude.
> > > >>>>>
> > > >>>>> That's not a big deal.
> > > >>>>> Here my own problem goes.
> > > >>>>>
> > > >>>>> dx and dy in my WRF simulation are 36 km both in Lat and
Lon.
> > > >>>>> However, if I convert them into millidegrees, DX is
323.515315
> and
> > DY
> > > >>>>> is 287.23175 computationally supposing the earth is a
complete
> > > sphere.
> > > >>>>> copygb takes only integers, DX is 323 and DY is 287. But,
I'm not
> > > >>>>> sure if it's right to use the different DX and DY or if I
should
> > use
> > > the
> > > >>>>> same number for DX and DY.
> > > >>>>> Please let me know.
> > > >>>>>
> > > >>>>> Thank you.
> > > >>>>>
> > > >>>>> Jinwoong Yoo
> > > >>>>> UNM
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>> On Thu, Oct 2, 2014 at 2:18 PM, Jinwoong Yoo <
> > jinwoong.yoo at gmail.com
> > > >
> > > >>>>> wrote:
> > > >>>>>
> > > >>>>>> Dear John,
> > > >>>>>>
> > > >>>>>> Actually I was looking over copygb when I found your
email
> > received.
> > > >>>>>> It seems that I have to convert the UPP output again
using
> > > copygb.exe
> > > >>>>>> to match up the grid of WRF results to the CCSM4's
lat/lon grid.
> > > >>>>>>
> > > >>>>>> Isn't there any way to do that in one step? No?
> > > >>>>>>
> > > >>>>>> Along with the precipitation, I'd like to compare
variables
> > (paired
> > > >>>>>> for WRF and CCSM4) below for example:
> > > >>>>>>
> > > >>>>>> Varwrfccsm = (/"PSFC","SLP","T","U", "V",
"QVAPOR","GEOPT"/)
> > > >>>>>> Varccsm = (/ "PS", "PSL","T","U", "V", "Q", "Z3"/)
> > > >>>>>>
> > > >>>>>>
> > > >>>>>> Varwrfcam = (/ "HFX","TSK","TPW","THETA","RH"/)
> > > >>>>>> Varcam = (/ "SHFLX","TS","TMQ","TH","RELHUM"/)
> > > >>>>>>
> > > >>>>>>
> > > >>>>>> Varwrfclm = (/"T2", "TSLB","SMOIS",
> > > >>>>>> "SNOWH","SWDOWN","SNOWNC","Q2","RH2"/);
> > > >>>>>> Varclm = (/ "TSA","TSOI", "H2OSOI", "SNOWDP",
> > > >>>>>> "FSDS","SNOWICE","Q2M","RH2M"/)
> > > >>>>>>
> > > >>>>>> Variables in the lists are the actual name of variables
in the
> > CCSM4
> > > >>>>>> (including CAM and CLM monthly files) files and
corresponding to
> > > those in
> > > >>>>>> the WRF-ARW.
> > > >>>>>>
> > > >>>>>>
> > > >>>>>> I'm trying to run Grid-Stat for a test and I'd like to
run
> > > >>>>>> Series-Analysis for the entire dataset of the WRF and the
CCSM4.
> > > >>>>>> However, I'm concerning about how to run the Series-
Analysis for
> > the
> > > >>>>>> monthly or annual data using MET?
> > > >>>>>> At the same time, I am also thinking about using the NCL
to do
> the
> > > >>>>>> same statistical analyses. I don't know which one will be
easier
> > > for me to
> > > >>>>>> execute for now.
> > > >>>>>> We'll see.
> > > >>>>>>
> > > >>>>>> Will let you know as I make progress, John.
> > > >>>>>>
> > > >>>>>> >>> FYI - I just talked to the person who provides
support for
> UPP
> > > >>>>>> through
> > > >>>>>> >>> wrfhelp, and she's expecting an email from you.
> > > >>>>>>
> > > >>>>>> I cc'ed this email to wrfhelp.
> > > >>>>>> Thank you.
> > > >>>>>>
> > > >>>>>> Regards,
> > > >>>>>>
> > > >>>>>> Jinwoong Yoo
> > > >>>>>> UNM
> > > >>>>>>
> > > >>>>>> On Thu, Oct 2, 2014 at 12:02 PM, John Halley Gotway via
RT <
> > > >>>>>> met_help at ucar.edu> wrote:
> > > >>>>>>
> > > >>>>>>> Jinwoong,
> > > >>>>>>>
> > > >>>>>>> I'm able to access your data on Yellowstone. I see that
you'd
> > like
> > > >>>>>>> to
> > > >>>>>>> compare your UPP output in GRIB format to some data that
> follows
> > > the
> > > >>>>>>> NetCDF
> > > >>>>>>> CF-1.0 convention.
> > > >>>>>>>
> > > >>>>>>> UPP output:
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd
> > > >>>>>>> CCSM4 output:
> > > /glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly
> > > >>>>>>>
> > > >>>>>>> Hopefully you're able to look at those files and
identify which
> > > GRIB
> > > >>>>>>> records you'd like to compare to which NetCDF variables.
When
> > you
> > > >>>>>>> run the
> > > >>>>>>> MET tools, you'll define the forecast and observation
fields to
> > be
> > > >>>>>>> used.
> > > >>>>>>>
> > > >>>>>>> Looking at your data, I see 2 immediate problems:
> > > >>>>>>>
> > > >>>>>>> (1) Gridded forecast and observation data must be on the
same
> > grid
> > > >>>>>>> before
> > > >>>>>>> MET can compare them.
> > > >>>>>>> Here's a description from wgrib of the grid for your
GRIB
> > data:
> > > >>>>>>> Mercator: lat -45.239000 to 45.239000 by
36.000000 km
> > nxny
> > > >>>>>>> 349650
> > > >>>>>>> long 0.449000 to -0.449000 by 36.000000
km,
> > (1110 x
> > > >>>>>>> 315)
> > > >>>>>>> scan 64 mode 136 Latin 0.000000 bdsgrid 1
> > > >>>>>>> But it looks like the CCSM4 data is on a lat/lon
grid.
> > > >>>>>>>
> > > >>>>>>> I'd suggest using the copygb utility to regrid the GRIB1
output
> > > from
> > > >>>>>>> UPP to
> > > >>>>>>> the lat/lon grid of the CCSM4 data. Here's an example
of doing
> > > that:
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>>
> > >
> >
>
http://www.dtcenter.org/met/users/support/online_tutorial/METv5.0/copygb/run2.php
> > > >>>>>>>
> > > >>>>>>> (2) There's an issue in the times in the accumulated
> > precipitation.
> > > >>>>>>> Try
> > > >>>>>>> running the following 2 commands:
> > > >>>>>>> wgrib
> > > >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.252
> |
> > > >>>>>>> grep APCP
> > > >>>>>>> wgrib
> > > >>>>>>>
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd/wrfprs_d01.258
> |
> > > >>>>>>> grep APCP
> > > >>>>>>>
> > > >>>>>>> For the first, APCP is an accumulation of precip from
hours 0
> to
> > > >>>>>>> 252. The
> > > >>>>>>> second should be 0 to 258, but it's stored as 0 to 2!
That's
> > > >>>>>>> because UPP
> > > >>>>>>> is failing to realize that the precipitation timestamps
are
> > > >>>>>>> overflowing the
> > > >>>>>>> 1 byte allocated for that second time unit. This
problem was
> > > >>>>>>> addressed in
> > > >>>>>>> an earlier version of UPP but the problem is still
showing up
> in
> > > the
> > > >>>>>>> version you're using. If you plan to verify precip, I'd
> suggest
> > > >>>>>>> writing
> > > >>>>>>> wrfhelp at ucar.edu, tell them the version of UPP you're
using,
> and
> > > >>>>>>> ask them
> > > >>>>>>> if this should be fixed.
> > > >>>>>>>
> > > >>>>>>> Hope that helps.
> > > >>>>>>>
> > > >>>>>>> Thanks,
> > > >>>>>>> John
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>> On Tue, Sep 30, 2014 at 4:05 PM, Jinwoong Yoo via RT <
> > > >>>>>>> met_help at ucar.edu>
> > > >>>>>>> wrote:
> > > >>>>>>>
> > > >>>>>>> >
> > > >>>>>>> > <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > >
> > > >>>>>>> >
> > > >>>>>>> > Hi, Julie.
> > > >>>>>>> > Thank you for heads-up.
> > > >>>>>>> > I will wait for John to reply then.
> > > >>>>>>> > Thank you.
> > > >>>>>>> >
> > > >>>>>>> > Jinwoong Yoo
> > > >>>>>>> > UNM
> > > >>>>>>> >
> > > >>>>>>> > On Tue, Sep 30, 2014 at 3:35 PM, Julie Prestopnik via
RT <
> > > >>>>>>> > met_help at ucar.edu>
> > > >>>>>>> > wrote:
> > > >>>>>>> >
> > > >>>>>>> > > Hello. John is out of the office, but will be back
later
> > this
> > > >>>>>>> week. I am
> > > >>>>>>> > > unable to help you with this task. Hopefully, John
will be
> > > able
> > > >>>>>>> to
> > > >>>>>>> > respond
> > > >>>>>>> > > to this ticket later this week.
> > > >>>>>>> > >
> > > >>>>>>> > > Julie
> > > >>>>>>> > >
> > > >>>>>>> > > On Tue, Sep 30, 2014 at 2:26 PM, Jinwoong Yoo via RT
<
> > > >>>>>>> met_help at ucar.edu>
> > > >>>>>>> > > wrote:
> > > >>>>>>> > >
> > > >>>>>>> > > >
> > > >>>>>>> > > > <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > >>>>>>> >
> > > >>>>>>> > > >
> > > >>>>>>> > > > Dear John,
> > > >>>>>>> > > >
> > > >>>>>>> > > > I ran UPP successfully on the UCAR Yellowstone.
> > > >>>>>>> > > > I assume you can access to my glade scratch
directory.
> > > >>>>>>> > > > My UPP ouputs are located
> > > >>>>>>> > > > under
/glade/scratch/jyoo/DOMAINS/postprd/lgm/postprd.
> > > >>>>>>> > > >
> > > >>>>>>> > > > Against those files, I'd like to compare CCSM4
simulation
> > > >>>>>>> outputs some
> > > >>>>>>> > of
> > > >>>>>>> > > > which are located also under
> > > >>>>>>> > > >
/glade/scratch/jyoo/domain/mean/analysis/LGMCCSM/6hrly/.
> > > >>>>>>> > > >
> > > >>>>>>> > > > Those are six hourly files all together. If you
cannot
> > access
> > > >>>>>>> to those
> > > >>>>>>> > > > files, please let me know.
> > > >>>>>>> > > >
> > > >>>>>>> > > > However, there are some variables among the CCSM4
outputs
> > > that
> > > >>>>>>> are
> > > >>>>>>> > > > available only monthly or annually. In this case,
how
> can I
> > > >>>>>>> > post-process
> > > >>>>>>> > > my
> > > >>>>>>> > > > WRF outputs for the usage of MET through UPP?
> > > >>>>>>> > > >
> > > >>>>>>> > > > Thank you.
> > > >>>>>>> > > >
> > > >>>>>>> > > > Regards,
> > > >>>>>>> > > >
> > > >>>>>>> > > > Jinwoong Yoo
> > > >>>>>>> > > > UNM
> > > >>>>>>> > > >
> > > >>>>>>> > > >
> > > >>>>>>> > > >
> > > >>>>>>> > > >
> > > >>>>>>> > > > On Fri, Sep 26, 2014 at 11:36 PM, John Halley
Gotway via
> > RT <
> > > >>>>>>> > > > met_help at ucar.edu> wrote:
> > > >>>>>>> > > >
> > > >>>>>>> > > > > Jinwoo,
> > > >>>>>>> > > > >
> > > >>>>>>> > > > > I believe UPP will still run, just using a
single time
> > > value
> > > >>>>>>> as if it
> > > >>>>>>> > > > were
> > > >>>>>>> > > > > an instantaneous field rather than an
accumulation
> > > >>>>>>> interval. The
> > > >>>>>>> > > default
> > > >>>>>>> > > > > for ARW is to store precip as a runtime
accumulation.
> > For
> > > >>>>>>> forecasts
> > > >>>>>>> > > > hours
> > > >>>>>>> > > > > < 255, like 12, I suspect it'll be stored in
GRIB as an
> > > >>>>>>> accumulation
> > > >>>>>>> > > > from 0
> > > >>>>>>> > > > > to 12. For hours > 255, like 300, I suspect
it'll be
> > > stored
> > > >>>>>>> as an
> > > >>>>>>> > > > > instantaneous 300 hour forecast. It's a subtle
> > difference,
> > > >>>>>>> but
> > > >>>>>>> > > something
> > > >>>>>>> > > > > to be aware of when you set up MET config files
to
> verify
> > > >>>>>>> the output.
> > > >>>>>>> > > > >
> > > >>>>>>> > > > > Give UPP a try and let me know how it goes.
> > > >>>>>>> > > > >
> > > >>>>>>> > > > > John
> > > >>>>>>> > > > >
> > > >>>>>>> > > > > On Friday, September 26, 2014, Jinwoong Yoo via
RT <
> > > >>>>>>> > met_help at ucar.edu>
> > > >>>>>>> > > > > wrote:
> > > >>>>>>> > > > >
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > <URL:
> > > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
>
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > Dear John,
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > I see. Let me change "export lastfhr=18" to
"export
> > > >>>>>>> lastfhr=744".
> > > >>>>>>> > > > > > Thank you for your comments.
> > > >>>>>>> > > > > > Yes, I'd like to evaluate precipitation, too,
and my
> 10
> > > >>>>>>> year
> > > >>>>>>> > > simulation
> > > >>>>>>> > > > > was
> > > >>>>>>> > > > > > integrated from a single initialization. Then,
will
> the
> > > UPP
> > > >>>>>>> > > run_unipost
> > > >>>>>>> > > > > > crash after the 255 forecast hours or just
will
> > > >>>>>>> precipitation not
> > > >>>>>>> > be
> > > >>>>>>> > > > > stored
> > > >>>>>>> > > > > > any more aster the 255 forecast hours?
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > Thank you.
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > Regards,
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > Jinwoong Yoo
> > > >>>>>>> > > > > > UNM
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > On Fri, Sep 26, 2014 at 8:21 PM, John Halley
Gotway
> via
> > > RT
> > > >>>>>>> <
> > > >>>>>>> > > > > > met_help at ucar.edu <javascript:;>> wrote:
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > > > Jinwoong,
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > UPP support is typically provided through
> > > >>>>>>> wrf_help at ucar.edu
> > > >>>>>>> > > > > > <javascript:;>. I'll give it
> > > >>>>>>> > > > > > > a shot, but if this doesn't solve it, I'll
refer
> you
> > to
> > > >>>>>>> wrf_help.
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > I assume you're running the script named
> > "run_unipost".
> > > >>>>>>> Looking
> > > >>>>>>> > in
> > > >>>>>>> > > > > > there,
> > > >>>>>>> > > > > > > you'll see:
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > export startdate=2005012300
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > export fhr=00
> > > >>>>>>> > > > > > > export lastfhr=18
> > > >>>>>>> > > > > > > export incrementhr=03
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > I'd suggest modifying the...
> > > >>>>>>> > > > > > > - "startdate" to be the model
initialization time
> to
> > > be
> > > >>>>>>> > processed.
> > > >>>>>>> > > > > > > - Leave "fhr" set to 00 to start at the
analysis
> > time.
> > > >>>>>>> > > > > > > - Set "lastfhr" to be maybe 744 (which is
31 days
> *
> > 24
> > > >>>>>>> > hours/day).
> > > >>>>>>> > > > > > > - Set "incrementhr" to 06 to process output
every
> 6
> > > >>>>>>> hours.
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > Give that a shot and let me know if that
gets you
> > > output
> > > >>>>>>> past the
> > > >>>>>>> > > > first
> > > >>>>>>> > > > > > > day.
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > Are you going to be evaluating
precipitation? I
> ask
> > > >>>>>>> that because
> > > >>>>>>> > > > > precip
> > > >>>>>>> > > > > > is
> > > >>>>>>> > > > > > > an accumulated interval, meaning that it has
a
> > > beginning
> > > >>>>>>> > > accumulation
> > > >>>>>>> > > > > > time
> > > >>>>>>> > > > > > > and ending accumulation time. GRIB1 is
limited in
> > the
> > > >>>>>>> amount of
> > > >>>>>>> > > > space
> > > >>>>>>> > > > > > > available for storing times. When you get
past 255
> > > >>>>>>> forecast
> > > >>>>>>> > hours,
> > > >>>>>>> > > > you
> > > >>>>>>> > > > > > can
> > > >>>>>>> > > > > > > no longer store beginning/ending times for
precip
> > > >>>>>>> because it runs
> > > >>>>>>> > > out
> > > >>>>>>> > > > > of
> > > >>>>>>> > > > > > > room in the header. Do you have a 10 year
> simulation
> > > >>>>>>> computed
> > > >>>>>>> > > from a
> > > >>>>>>> > > > > > > single initialization? Meaning, do you have
> forecast
> > > >>>>>>> lead times
> > > >>>>>>> > > out
> > > >>>>>>> > > > to
> > > >>>>>>> > > > > > > 87,600 hours (= 10 years * 365 days * 24
hours)?
> I'm
> > > >>>>>>> wondering
> > > >>>>>>> > if
> > > >>>>>>> > > > > we'll
> > > >>>>>>> > > > > > > run into problems formatting those time
strings.
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > Thanks,
> > > >>>>>>> > > > > > > John
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > On Thu, Sep 25, 2014 at 3:13 PM, Jinwoong
Yoo via
> RT
> > <
> > > >>>>>>> > > > > met_help at ucar.edu
> > > >>>>>>> > > > > > <javascript:;>>
> > > >>>>>>> > > > > > > wrote:
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > <URL:
> > > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > >>>>>>> > >
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > Dear John,
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > I ran UPP with my multiple wrfout files,
first
> with
> > > >>>>>>> the first
> > > >>>>>>> > > month
> > > >>>>>>> > > > > of
> > > >>>>>>> > > > > > > the
> > > >>>>>>> > > > > > > > 10 year simulation. However, after
executing the
> > > >>>>>>> run_unipost
> > > >>>>>>> > > > script,
> > > >>>>>>> > > > > > UPP
> > > >>>>>>> > > > > > > > produced only four files:
unipost_d01.00.out,
> > > >>>>>>> > unipost_d01.06.out,
> > > >>>>>>> > > > > > > > unipost_d01.12.out, and
unipost_d01.18.out. It
> > seems
> > > >>>>>>> that UPP
> > > >>>>>>> > > > > processed
> > > >>>>>>> > > > > > > > only the first day's output among the one
month
> > data.
> > > >>>>>>> > > > > > > > If you are familiar with UPP, how can I
make sure
> > for
> > > >>>>>>> UPP to
> > > >>>>>>> > > > process
> > > >>>>>>> > > > > > > > multiple wrfout files more than one day?
> > > >>>>>>> > > > > > > > Thank you.
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > Regards,
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > Jinwoong Yoo
> > > >>>>>>> > > > > > > > UNM
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > On Thu, Sep 25, 2014 at 10:34 AM, Jinwoong
Yoo <
> > > >>>>>>> > > > > jinwoong.yoo at gmail.com
> > > >>>>>>> > > > > > <javascript:;>>
> > > >>>>>>> > > > > > > > wrote:
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > > > Dear John,
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > > > By the way, what about the CCSM4 file in
NetCDF
> > > >>>>>>> format?
> > > >>>>>>> > > > > > > > > Do I need to post process of the CCSM4
files
> also
> > > to
> > > >>>>>>> run MET
> > > >>>>>>> > > > along
> > > >>>>>>> > > > > > with
> > > >>>>>>> > > > > > > > > the WRF-ARW files?
> > > >>>>>>> > > > > > > > > Thank you.
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > > > Regards,
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > > > Jinwoong Yoo
> > > >>>>>>> > > > > > > > > UNM
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > > > On Thu, Sep 25, 2014 at 9:44 AM,
Jinwoong Yoo <
> > > >>>>>>> > > > > > jinwoong.yoo at gmail.com <javascript:;>>
> > > >>>>>>> > > > > > > > > wrote:
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > > >> Dear John,
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >> Thank you for your kind email with the
> > direction.
> > > >>>>>>> > > > > > > > >> Let me first post process some of my
WRF-ARW
> > > >>>>>>> outputs using
> > > >>>>>>> > UPP
> > > >>>>>>> > > > and
> > > >>>>>>> > > > > > let
> > > >>>>>>> > > > > > > > >> you know.
> > > >>>>>>> > > > > > > > >> Thank you very much.
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >> Regards,
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >> Jinwoong Yoo
> > > >>>>>>> > > > > > > > >> UNM
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >> On Thu, Sep 25, 2014 at 8:52 AM, John
Halley
> > > Gotway
> > > >>>>>>> via RT <
> > > >>>>>>> > > > > > > > >> met_help at ucar.edu <javascript:;>>
wrote:
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >>> Jinwoong,
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>> I see that you'd like some help
running the
> > > >>>>>>> series-analysis
> > > >>>>>>> > > > tool
> > > >>>>>>> > > > > on
> > > >>>>>>> > > > > > > > your
> > > >>>>>>> > > > > > > > >>> wrf-arw climate runs. One thing to
note is
> > that
> > > >>>>>>> MET can't
> > > >>>>>>> > > read
> > > >>>>>>> > > > > the
> > > >>>>>>> > > > > > > raw
> > > >>>>>>> > > > > > > > >>> wrfout files directly. Instead you
first
> need
> > to
> > > >>>>>>> post
> > > >>>>>>> > > process
> > > >>>>>>> > > > > them
> > > >>>>>>> > > > > > > > using
> > > >>>>>>> > > > > > > > >>> either UPP (produces GRIB output) or
the
> > p_interp
> > > >>>>>>> utility
> > > >>>>>>> > > > > (produced
> > > >>>>>>> > > > > > > > >>> NetCDF
> > > >>>>>>> > > > > > > > >>> output). After you've post processed
some
> > files,
> > > >>>>>>> please
> > > >>>>>>> > send
> > > >>>>>>> > > > us
> > > >>>>>>> > > > > > some
> > > >>>>>>> > > > > > > > >>> sample data by posting it to our
anonymous
> ftp
> > > >>>>>>> site. Also
> > > >>>>>>> > > > send a
> > > >>>>>>> > > > > > > > >>> description of the comparisons you'd
like to
> > make
> > > >>>>>>> between
> > > >>>>>>> > > the 2
> > > >>>>>>> > > > > > > > datasets.
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>> Here's instructions for sending us
data:
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> http://www.dtcenter.org/met/users/support/met_help.php
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>> Thanks
> > > >>>>>>> > > > > > > > >>> John Halley Gotway
> > > >>>>>>> > > > > > > > >>> met_help at ucar.edu <javascript:;>
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>> On Wednesday, September 24, 2014,
Jinwoong
> Yoo
> > > via
> > > >>>>>>> RT <
> > > >>>>>>> > > > > > > > met_help at ucar.edu <javascript:;>
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> wrote:
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Wed Sep 24 17:49:47 2014: Request
69099 was
> > > >>>>>>> acted upon.
> > > >>>>>>> > > > > > > > >>> > Transaction: Ticket created by
> > > >>>>>>> jinwoong.yoo at gmail.com
> > > >>>>>>> > > > > > <javascript:;>
> > > >>>>>>> > > > > > > <javascript:;>
> > > >>>>>>> > > > > > > > >>> > Queue: met_help
> > > >>>>>>> > > > > > > > >>> > Subject: Comparing WRF-ARW
outputs and
> > > >>>>>>> CCSM4 outputs
> > > >>>>>>> > > > > > > > >>> > Owner: Nobody
> > > >>>>>>> > > > > > > > >>> > Requestors: jinwoong.yoo at gmail.com
> > > >>>>>>> <javascript:;>
> > > >>>>>>> > > > > > <javascript:;>
> > > >>>>>>> > > > > > > > >>> > Status: new
> > > >>>>>>> > > > > > > > >>> > Ticket <URL:
> > > >>>>>>> > > > > > > >
> > > >>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=69099
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Dear Met Help,
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Hi.
> > > >>>>>>> > > > > > > > >>> > I am looking for a way to verify WRF
model
> > > >>>>>>> performance
> > > >>>>>>> > as a
> > > >>>>>>> > > > > > climate
> > > >>>>>>> > > > > > > > >>> model.
> > > >>>>>>> > > > > > > > >>> > I'm completed 10-year tropical
channel
> > climate
> > > >>>>>>> model
> > > >>>>>>> > > > > simulations
> > > >>>>>>> > > > > > > > using
> > > >>>>>>> > > > > > > > >>> the
> > > >>>>>>> > > > > > > > >>> > CCSM4 data as input data (dynamical
> > > downscaling)
> > > >>>>>>> for two
> > > >>>>>>> > > > > > different
> > > >>>>>>> > > > > > > > time
> > > >>>>>>> > > > > > > > >>> > periods: one for the LGM and the
other for
> > the
> > > >>>>>>> modern as
> > > >>>>>>> > a
> > > >>>>>>> > > > > > control
> > > >>>>>>> > > > > > > > run.
> > > >>>>>>> > > > > > > > >>> > The WRF simulation produced 6 hourly
output
> > for
> > > >>>>>>> ten years
> > > >>>>>>> > > > > periods
> > > >>>>>>> > > > > > > > each.
> > > >>>>>>> > > > > > > > >>> > Also the CCSM4 simulations come with
6
> hourly
> > > >>>>>>> outputs.
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Basically, I'd like to compare the
two
> > > simulation
> > > >>>>>>> > results,
> > > >>>>>>> > > > > > > verifying
> > > >>>>>>> > > > > > > > >>> the
> > > >>>>>>> > > > > > > > >>> > WRF model simulations.
> > > >>>>>>> > > > > > > > >>> > (WRF ARW was implemented for the
> > experiments.)
> > > >>>>>>> > > > > > > > >>> > Series-Analysis in the MET v5.0
seems to do
> > the
> > > >>>>>>> similar
> > > >>>>>>> > > jobs
> > > >>>>>>> > > > > but
> > > >>>>>>> > > > > > I
> > > >>>>>>> > > > > > > > need
> > > >>>>>>> > > > > > > > >>> > much more help from you than the MET
Online
> > > >>>>>>> Tutorial to
> > > >>>>>>> > > make
> > > >>>>>>> > > > a
> > > >>>>>>> > > > > > step
> > > >>>>>>> > > > > > > > >>> > forward.
> > > >>>>>>> > > > > > > > >>> > Any help will be highly appreciated.
> > > >>>>>>> > > > > > > > >>> > Thank you.
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Regards,
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> > Jinwoong Yoo
> > > >>>>>>> > > > > > > > >>> > Postdoc
> > > >>>>>>> > > > > > > > >>> > Department of Earth and Planetary
Sciences
> > > >>>>>>> > > > > > > > >>> > UNM
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>> >
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>>
> > > >>>>>>> > > > > > > > >>
> > > >>>>>>> > > > > > > > >
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > > >
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > > >
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > > >
> > > >>>>>>> > > > >
> > > >>>>>>> > > > >
> > > >>>>>>> > > >
> > > >>>>>>> > > >
> > > >>>>>>> > >
> > > >>>>>>> > >
> > > >>>>>>> > > --
> > > >>>>>>> > > Julie Prestopnik
> > > >>>>>>> > > National Center for Atmospheric Research
> > > >>>>>>> > > Research Applications Laboratory
> > > >>>>>>> > > Phone: 303.497.8399
> > > >>>>>>> > > Email: jpresto at ucar.edu
> > > >>>>>>> > >
> > > >>>>>>> > >
> > > >>>>>>> >
> > > >>>>>>> >
> > > >>>>>>>
> > > >>>>>>>
> > > >>>>>>
> > > >>>>>
> > > >>>>
> > > >>>
> > > >>
> > > >
> > >
> > >
> >
> >
> > --
> > Julie Prestopnik
> > National Center for Atmospheric Research
> > Research Applications Laboratory
> > Phone: 303.497.8399
> > Email: jpresto at ucar.edu
> >
> >
>
>
--
Julie Prestopnik
National Center for Atmospheric Research
Research Applications Laboratory
Phone: 303.497.8399
Email: jpresto at ucar.edu
------------------------------------------------
More information about the Met_help
mailing list