[Met_help] [rt.rap.ucar.edu #90484] History for What are needed to run any executables
John Halley Gotway via RT
met_help at ucar.edu
Wed Jul 28 14:53:04 MDT 2021
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Actually, I'm a beginner. Somehow i have installed met tools & got all
executables. Now what else i need to run an executable. Reading user guide
i understand i need an input data file, a configure file. but I find these
instructions pretty complicated.
Actually, i want to calculate Skill score and rmse. i have grib2 data &
wrfout file . what else i need & how to get those things???
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: What are needed to run any executables
From: Tara Jensen
Time: Wed Jun 05 15:32:27 2019
Golam,
I see you are having trouble getting started with using MET. A good
place
to start is our online tutorial at:
https://dtcenter.org/community-code/model-evaluation-tools-met/online-
tutorial.
<https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial>
Under Tutorial Setup -> Compilation -> Tutorial Setup, there
instructions
on how to download a set of test data to demonstrate how to call the
tools
and then steps through basic calls to each tool. Once you know the
basics,
you can advance to using your own data.
Computation of traditional statistics like RMSE and skill scores would
be
done in tools like Grid-Stat (passing in 2 gridded files) or Point-
Stat
(passing in a gridded forecast file and a file containing point
observations). It's unclear to me what "Skill Score" you are trying
to
compute. There are several and might be computed using different
methods
and may require requesting different "line types" to be requested.
Please check out the online tutorial and then let us know what
questions
you have.
Thanks, Tara
On Wed, Jun 5, 2019 at 11:44 AM Golam Rabbani via RT
<met_help at ucar.edu>
wrote:
>
> Wed Jun 05 11:44:52 2019: Request 90484 was acted upon.
> Transaction: Ticket created by rabbani.du93 at gmail.com
> Queue: met_help
> Subject: What are needed to run any executables
> Owner: Nobody
> Requestors: rabbani.du93 at gmail.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
>
>
> Actually, I'm a beginner. Somehow i have installed met tools & got
all
> executables. Now what else i need to run an executable. Reading
user guide
> i understand i need an input data file, a configure file. but I find
these
> instructions pretty complicated.
>
> Actually, i want to calculate Skill score and rmse. i have grib2
data &
> wrfout file . what else i need & how to get those things???
>
>
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Tara Jensen
Project Manager II
NCAR RAL and DTC
PO Box 3000, Boulder, Colorado 80307 USA
+1 303-497-8479 jensen at ucar.edu
------------------------------------------------
Subject: What are needed to run any executables
From: Golam Rabbani
Time: Wed Oct 16 14:36:34 2019
Hello Sir, I have just installed MET Tools after spending many days.
But I
can't get output like Probability of Detection, False Alarm ratio or
Gilbert Skill Score. Can you please help me to calculate or visualize
these
things?
In contrast, I found it difficult to modify the config script. I have
WRF
model output for rain (rf) attached here as remappedmodel.nc and IMD
Gridded Rainfall as observation attached here as remapped.nc. For
remapping
I used CDO. I used GrADS sdfwrite command for getting rf as rainfall
from
the model output. Both are for 24 hours. It is also hard for me to
determine which tool (maybe grid stat) I should use for it.
In my country, I found none who used MET TOOLS. So I don't get any
support,
So, if you please support me, it would be very
helpful for my study and research.
Thanks
Golam Rabbani
Dhaka University, Bangladesh.
On Thu, Jun 6, 2019 at 3:32 AM Tara Jensen via RT <met_help at ucar.edu>
wrote:
> Golam,
>
> I see you are having trouble getting started with using MET. A good
place
> to start is our online tutorial at:
>
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> .
> <
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> >
>
> Under Tutorial Setup -> Compilation -> Tutorial Setup, there
instructions
> on how to download a set of test data to demonstrate how to call the
tools
> and then steps through basic calls to each tool. Once you know the
basics,
> you can advance to using your own data.
>
> Computation of traditional statistics like RMSE and skill scores
would be
> done in tools like Grid-Stat (passing in 2 gridded files) or Point-
Stat
> (passing in a gridded forecast file and a file containing point
> observations). It's unclear to me what "Skill Score" you are trying
to
> compute. There are several and might be computed using different
methods
> and may require requesting different "line types" to be requested.
>
> Please check out the online tutorial and then let us know what
questions
> you have.
>
> Thanks, Tara
>
>
>
> On Wed, Jun 5, 2019 at 11:44 AM Golam Rabbani via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Wed Jun 05 11:44:52 2019: Request 90484 was acted upon.
> > Transaction: Ticket created by rabbani.du93 at gmail.com
> > Queue: met_help
> > Subject: What are needed to run any executables
> > Owner: Nobody
> > Requestors: rabbani.du93 at gmail.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> >
> >
> > Actually, I'm a beginner. Somehow i have installed met tools & got
all
> > executables. Now what else i need to run an executable. Reading
user
> guide
> > i understand i need an input data file, a configure file. but I
find
> these
> > instructions pretty complicated.
> >
> > Actually, i want to calculate Skill score and rmse. i have grib2
data &
> > wrfout file . what else i need & how to get those things???
> >
> >
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Tara Jensen
> Project Manager II
> NCAR RAL and DTC
> PO Box 3000, Boulder, Colorado 80307 USA
> +1 303-497-8479 jensen at ucar.edu
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: Julie Prestopnik
Time: Wed Oct 16 14:52:16 2019
Hi.
I'm glad that you were able to get MET installed, however, I'm sorry
to see
that you are having problems getting some desired output. I am going
to
assign this ticket to John Halley Gotway, who should be able to help
you
further. Please allow a few business days for a response. Thank you
for
your interest in MET!
Julie
On Wed, Oct 16, 2019 at 2:37 PM Golam Rabbani via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
>
> Hello Sir, I have just installed MET Tools after spending many days.
But I
> can't get output like Probability of Detection, False Alarm ratio or
> Gilbert Skill Score. Can you please help me to calculate or
visualize these
> things?
> In contrast, I found it difficult to modify the config script. I
have WRF
> model output for rain (rf) attached here as remappedmodel.nc and IMD
> Gridded Rainfall as observation attached here as remapped.nc. For
> remapping
> I used CDO. I used GrADS sdfwrite command for getting rf as rainfall
from
> the model output. Both are for 24 hours. It is also hard for me to
> determine which tool (maybe grid stat) I should use for it.
> In my country, I found none who used MET TOOLS. So I don't get any
support,
> So, if you please support me, it would be very
> helpful for my study and research.
>
> Thanks
> Golam Rabbani
> Dhaka University, Bangladesh.
>
> On Thu, Jun 6, 2019 at 3:32 AM Tara Jensen via RT
<met_help at ucar.edu>
> wrote:
>
> > Golam,
> >
> > I see you are having trouble getting started with using MET. A
good
> place
> > to start is our online tutorial at:
> >
> >
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> > .
> > <
> >
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> > >
> >
> > Under Tutorial Setup -> Compilation -> Tutorial Setup, there
instructions
> > on how to download a set of test data to demonstrate how to call
the
> tools
> > and then steps through basic calls to each tool. Once you know
the
> basics,
> > you can advance to using your own data.
> >
> > Computation of traditional statistics like RMSE and skill scores
would be
> > done in tools like Grid-Stat (passing in 2 gridded files) or
Point-Stat
> > (passing in a gridded forecast file and a file containing point
> > observations). It's unclear to me what "Skill Score" you are
trying to
> > compute. There are several and might be computed using different
methods
> > and may require requesting different "line types" to be requested.
> >
> > Please check out the online tutorial and then let us know what
questions
> > you have.
> >
> > Thanks, Tara
> >
> >
> >
> > On Wed, Jun 5, 2019 at 11:44 AM Golam Rabbani via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Wed Jun 05 11:44:52 2019: Request 90484 was acted upon.
> > > Transaction: Ticket created by rabbani.du93 at gmail.com
> > > Queue: met_help
> > > Subject: What are needed to run any executables
> > > Owner: Nobody
> > > Requestors: rabbani.du93 at gmail.com
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> >
> > >
> > >
> > > Actually, I'm a beginner. Somehow i have installed met tools &
got all
> > > executables. Now what else i need to run an executable. Reading
user
> > guide
> > > i understand i need an input data file, a configure file. but I
find
> > these
> > > instructions pretty complicated.
> > >
> > > Actually, i want to calculate Skill score and rmse. i have grib2
data &
> > > wrfout file . what else i need & how to get those things???
> > >
> > >
> >
> > --
> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > Tara Jensen
> > Project Manager II
> > NCAR RAL and DTC
> > PO Box 3000, Boulder, Colorado 80307 USA
> > +1 303-497-8479 jensen at ucar.edu
> >
> >
>
>
--
Julie Prestopnik
Software Engineer
National Center for Atmospheric Research
Research Applications Laboratory
Phone: 303.497.8399
Email: jpresto at ucar.edu
My working day may not be your working day. Please do not feel
obliged to
reply to this email outside of your normal working hours.
------------------------------------------------
Subject: What are needed to run any executables
From: Golam Rabbani
Time: Wed Oct 16 14:53:27 2019
Thank you for your response.
On Thu, 17 Oct, 2019 at 2:52 AM, Julie Prestopnik via RT
<met_help at ucar.edu>
wrote:
> Hi.
>
> I'm glad that you were able to get MET installed, however, I'm sorry
to see
> that you are having problems getting some desired output. I am
going to
> assign this ticket to John Halley Gotway, who should be able to help
you
> further. Please allow a few business days for a response. Thank you
for
> your interest in MET!
>
> Julie
>
>
>
> On Wed, Oct 16, 2019 at 2:37 PM Golam Rabbani via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> >
> > Hello Sir, I have just installed MET Tools after spending many
days. But
> I
> > can't get output like Probability of Detection, False Alarm ratio
or
> > Gilbert Skill Score. Can you please help me to calculate or
visualize
> these
> > things?
> > In contrast, I found it difficult to modify the config script. I
have WRF
> > model output for rain (rf) attached here as remappedmodel.nc and
IMD
> > Gridded Rainfall as observation attached here as remapped.nc. For
> > remapping
> > I used CDO. I used GrADS sdfwrite command for getting rf as
rainfall from
> > the model output. Both are for 24 hours. It is also hard for me to
> > determine which tool (maybe grid stat) I should use for it.
> > In my country, I found none who used MET TOOLS. So I don't get any
> support,
> > So, if you please support me, it would be very
> > helpful for my study and research.
> >
> > Thanks
> > Golam Rabbani
> > Dhaka University, Bangladesh.
> >
> > On Thu, Jun 6, 2019 at 3:32 AM Tara Jensen via RT
<met_help at ucar.edu>
> > wrote:
> >
> > > Golam,
> > >
> > > I see you are having trouble getting started with using MET. A
good
> > place
> > > to start is our online tutorial at:
> > >
> > >
> >
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> > > .
> > > <
> > >
> >
> https://dtcenter.org/community-code/model-evaluation-tools-
met/online-tutorial
> > > >
> > >
> > > Under Tutorial Setup -> Compilation -> Tutorial Setup, there
> instructions
> > > on how to download a set of test data to demonstrate how to call
the
> > tools
> > > and then steps through basic calls to each tool. Once you know
the
> > basics,
> > > you can advance to using your own data.
> > >
> > > Computation of traditional statistics like RMSE and skill scores
would
> be
> > > done in tools like Grid-Stat (passing in 2 gridded files) or
Point-Stat
> > > (passing in a gridded forecast file and a file containing point
> > > observations). It's unclear to me what "Skill Score" you are
trying to
> > > compute. There are several and might be computed using
different
> methods
> > > and may require requesting different "line types" to be
requested.
> > >
> > > Please check out the online tutorial and then let us know what
> questions
> > > you have.
> > >
> > > Thanks, Tara
> > >
> > >
> > >
> > > On Wed, Jun 5, 2019 at 11:44 AM Golam Rabbani via RT <
> met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > Wed Jun 05 11:44:52 2019: Request 90484 was acted upon.
> > > > Transaction: Ticket created by rabbani.du93 at gmail.com
> > > > Queue: met_help
> > > > Subject: What are needed to run any executables
> > > > Owner: Nobody
> > > > Requestors: rabbani.du93 at gmail.com
> > > > Status: new
> > > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> > >
> > > >
> > > >
> > > > Actually, I'm a beginner. Somehow i have installed met tools &
got
> all
> > > > executables. Now what else i need to run an executable.
Reading user
> > > guide
> > > > i understand i need an input data file, a configure file. but
I find
> > > these
> > > > instructions pretty complicated.
> > > >
> > > > Actually, i want to calculate Skill score and rmse. i have
grib2
> data &
> > > > wrfout file . what else i need & how to get those things???
> > > >
> > > >
> > >
> > > --
> > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > > Tara Jensen
> > > Project Manager II
> > > NCAR RAL and DTC
> > > PO Box 3000, Boulder, Colorado 80307 USA
> > > +1 303-497-8479 jensen at ucar.edu
> > >
> > >
> >
> >
>
> --
> Julie Prestopnik
> Software Engineer
> National Center for Atmospheric Research
> Research Applications Laboratory
> Phone: 303.497.8399
> Email: jpresto at ucar.edu
>
> My working day may not be your working day. Please do not feel
obliged to
> reply to this email outside of your normal working hours.
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: John Halley Gotway
Time: Wed Oct 16 15:35:05 2019
Hi Golam,
I see you're having trouble running your NetCDF output files through
MET.
Whenever getting started with a new gridded dataset, I always start by
running them through MET's plot_data_plane tool. For the file named
remappedmodel.nc, I had to explicitly tell MET to interpret it as
being a
CF-compliant NetCDF file because the standard global
"Conventions" attribute was not present, like it is in the one named
remapped.nc:
:Conventions = "CF-1.4" ;
Here are the 2 plot_data_plane commands I ran, and I've attached png
versions of the output:
/usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
remappedmodel.ps
'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
/usr/local/met-8.1.1/bin/plot_data_plane remapped.nc remapped.ps
'name="rf"; level="(0,*,*)";'
Note also that the first command produces a warning because no timing
information is present in the file:
WARNING: NcCfFile::open() -> could not determine the valid time, using
0.
Next, I ran the grid_stat utility to compute categorical statistics,
as
follows:
mkdir out
/usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc remapped.nc
GridStatConfig -outdir out -v 3
And I've attached the GridStatConfig file I used in which I requested
output for 2x2 contingency tables, NxN multi-category contingency
tables,
and continuous statistics. Looking at the range of values present in
your
data, I somewhat arbitrarily chose thresholds of:
cat_thresh = [ >10, >20, >30, >40, >50 ];
But here's the important part of it which tells Grid-Stat what data to
process:
fcst = {
file_type=NETCDF_NCCF;
field = [
{ name = "rf"; level = [ "(*,*)" ]; }
];
}
obs = {
field = [
{ name = "rf"; level = [ "(0,*,*)" ]; }
];
}
I've also attached a sample resulting output file containing the 2x2
contingency table counts for the 5 thresholds I chose:
out/grid_stat_000000L_19700101_000000V_ctc.txt
Notice that the timestamp indicates a time of 19700101, which is the
same
as unixtime = 0. This happens because the remappedmodel.nc file lacks
timing information.
Hope this helps get you going.
Thanks,
John Halley Gotway
On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT
<met_help at ucar.edu>
wrote:
>
> Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> Transaction: Owner forcibly changed from jensen (Tara Jensen) to
johnhg
> (John Halley Gotway) by jpresto
> Queue: met_help
> Subject: What are needed to run any executables
> Owner: johnhg
> Requestors: rabbani.du93 at gmail.com
> Status: open
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
>
>
> This transaction appears to have no content
>
------------------------------------------------
Subject: What are needed to run any executables
From: John Halley Gotway
Time: Wed Oct 16 15:35:05 2019
VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD
OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR
OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH
OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL FY_OY FY_ON FN_OY FN_ON
V8.1.1 WRF NA 000000 19700101_000000 19700101_000000 000000
20190401_000000 20190401_000000 rf NA *,* rf
NA 0,*,* ANALYS FULL NEAREST 1 >10
>10 NA NA CTC 11118 2734 952 1495 5937
V8.1.1 WRF NA 000000 19700101_000000 19700101_000000 000000
20190401_000000 20190401_000000 rf NA *,* rf
NA 0,*,* ANALYS FULL NEAREST 1 >20
>20 NA NA CTC 11118 1456 622 1932 7108
V8.1.1 WRF NA 000000 19700101_000000 19700101_000000 000000
20190401_000000 20190401_000000 rf NA *,* rf
NA 0,*,* ANALYS FULL NEAREST 1 >30
>30 NA NA CTC 11118 706 402 2021 7989
V8.1.1 WRF NA 000000 19700101_000000 19700101_000000 000000
20190401_000000 20190401_000000 rf NA *,* rf
NA 0,*,* ANALYS FULL NEAREST 1 >40
>40 NA NA CTC 11118 271 289 1664 8894
V8.1.1 WRF NA 000000 19700101_000000 19700101_000000 000000
20190401_000000 20190401_000000 rf NA *,* rf
NA 0,*,* ANALYS FULL NEAREST 1 >50
>50 NA NA CTC 11118 135 144 1207 9632
------------------------------------------------
Subject: What are needed to run any executables
From: Golam Rabbani
Time: Tue Jul 13 13:50:03 2021
Dear altruists,
I am trying to plot a time series of the FSS and ETS for selected
three
neighborhood radii (1*dx, 3*dx and 5*dx where dx=horizontal grid
spacing)
and for two to three thresholds (2.5, 5 and 10 mmm / hour for
rainfall. I
am using WRF-ARW model rainfall against NASA's hourly GPM IMERG Final
Precipitation datasets. Previously, I could calculate skill scores
with
your prescribed script for daily accumulated rainfall.
Now I am confused how I can calculate the time series of the ETS and
FSS as
wrf model simulated rainfall in accumulated form in every time step
whereas
GPM IMERG is just in hourly observation. Can you please provide me a
guideline with a script to do my task using the Grid_stat tool in MET?
Greetings and thank you in advance.
Sincerely,
Golam Rabbani
Department of Meteorology
University of Dhaka, Bangladesh
On Thu, Oct 17, 2019 at 3:35 AM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:
> Hi Golam,
>
> I see you're having trouble running your NetCDF output files through
MET.
>
> Whenever getting started with a new gridded dataset, I always start
by
> running them through MET's plot_data_plane tool. For the file named
> remappedmodel.nc, I had to explicitly tell MET to interpret it as
being a
> CF-compliant NetCDF file because the standard global
> "Conventions" attribute was not present, like it is in the one named
> remapped.nc:
> :Conventions = "CF-1.4" ;
>
> Here are the 2 plot_data_plane commands I ran, and I've attached png
> versions of the output:
> /usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
remappedmodel.ps
> 'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
> /usr/local/met-8.1.1/bin/plot_data_plane remapped.nc remapped.ps
> 'name="rf"; level="(0,*,*)";'
>
> Note also that the first command produces a warning because no
timing
> information is present in the file:
> WARNING: NcCfFile::open() -> could not determine the valid time,
using 0.
>
> Next, I ran the grid_stat utility to compute categorical statistics,
as
> follows:
>
> mkdir out
> /usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc remapped.nc
> GridStatConfig -outdir out -v 3
>
> And I've attached the GridStatConfig file I used in which I
requested
> output for 2x2 contingency tables, NxN multi-category contingency
tables,
> and continuous statistics. Looking at the range of values present
in your
> data, I somewhat arbitrarily chose thresholds of:
> cat_thresh = [ >10, >20, >30, >40, >50 ];
>
> But here's the important part of it which tells Grid-Stat what data
to
> process:
> fcst = {
> file_type=NETCDF_NCCF;
> field = [
> { name = "rf"; level = [ "(*,*)" ]; }
> ];
> }
> obs = {
> field = [
> { name = "rf"; level = [ "(0,*,*)" ]; }
> ];
> }
>
> I've also attached a sample resulting output file containing the 2x2
> contingency table counts for the 5 thresholds I chose:
> out/grid_stat_000000L_19700101_000000V_ctc.txt
>
> Notice that the timestamp indicates a time of 19700101, which is the
same
> as unixtime = 0. This happens because the remappedmodel.nc file
lacks
> timing information.
>
> Hope this helps get you going.
>
> Thanks,
> John Halley Gotway
>
> On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT
<met_help at ucar.edu
> >
> wrote:
>
> >
> > Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> > Transaction: Owner forcibly changed from jensen (Tara Jensen) to
johnhg
> > (John Halley Gotway) by jpresto
> > Queue: met_help
> > Subject: What are needed to run any executables
> > Owner: johnhg
> > Requestors: rabbani.du93 at gmail.com
> > Status: open
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> >
> >
> > This transaction appears to have no content
> >
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: John Halley Gotway
Time: Wed Jul 14 10:04:43 2021
Golam,
I see that you have a question about precipitation accumulation
intervals.
Your verifying IMERG analysis files contain hourly accumulations of
precip
while your WRF-ARW output contains a "runtime" accumulation of precip,
where forecast hour N contains all precip between forecast hour 0 and
N.
Yes, it's definitely important to match up the forecast and analysis
precipitation accumulations before comparing them. You have a few
options
here.
(1) Technically, you could reconfigure WRF-ARW to write hourly
precipitation accumulations to its output instead of runtime
accumulations.
If you have full control of the WRF configuration then this could be a
good
option and make comparisons easier. If you need more info on this, I
could
check on the corresponding WRF namelist option and get back to you.
(2) You could run MET's pcp_combine tool in "subtract" mode to convert
your
WRF runtime accumulations into hourly output. Of course, you're
limited by
the frequency of your WRF output times. If writing output hourly, then
you
could get hourly precip by running pcp_combine to subtract forecast
hour N
minus forecast hour N - 1. If writing 3-hourly output, you could do
forecast hour N minus forecast hour N - 3.
(3) You could run MET's pcp_combine tool in "add" mode to sum up the
hourly
IMERG data into larger bins.
(4) You could do some combination of (2) and (3) to put the forecast
and
analysis data into a common accumulation interval.
Hopefully that helps clarify some options.
Thanks,
John
On Tue, Jul 13, 2021 at 1:50 PM Golam Rabbani via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
>
> Dear altruists,
>
> I am trying to plot a time series of the FSS and ETS for selected
three
> neighborhood radii (1*dx, 3*dx and 5*dx where dx=horizontal grid
spacing)
> and for two to three thresholds (2.5, 5 and 10 mmm / hour for
rainfall. I
> am using WRF-ARW model rainfall against NASA's hourly GPM IMERG
Final
> Precipitation datasets. Previously, I could calculate skill scores
with
> your prescribed script for daily accumulated rainfall.
> Now I am confused how I can calculate the time series of the ETS and
FSS as
> wrf model simulated rainfall in accumulated form in every time step
whereas
> GPM IMERG is just in hourly observation. Can you please provide me a
> guideline with a script to do my task using the Grid_stat tool in
MET?
>
> Greetings and thank you in advance.
>
> Sincerely,
> Golam Rabbani
> Department of Meteorology
> University of Dhaka, Bangladesh
>
> On Thu, Oct 17, 2019 at 3:35 AM John Halley Gotway via RT <
> met_help at ucar.edu>
> wrote:
>
> > Hi Golam,
> >
> > I see you're having trouble running your NetCDF output files
through MET.
> >
> > Whenever getting started with a new gridded dataset, I always
start by
> > running them through MET's plot_data_plane tool. For the file
named
> > remappedmodel.nc, I had to explicitly tell MET to interpret it as
being
> a
> > CF-compliant NetCDF file because the standard global
> > "Conventions" attribute was not present, like it is in the one
named
> > remapped.nc:
> > :Conventions = "CF-1.4" ;
> >
> > Here are the 2 plot_data_plane commands I ran, and I've attached
png
> > versions of the output:
> > /usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
> remappedmodel.ps
> > 'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
> > /usr/local/met-8.1.1/bin/plot_data_plane remapped.nc remapped.ps
> > 'name="rf"; level="(0,*,*)";'
> >
> > Note also that the first command produces a warning because no
timing
> > information is present in the file:
> > WARNING: NcCfFile::open() -> could not determine the valid time,
using 0.
> >
> > Next, I ran the grid_stat utility to compute categorical
statistics, as
> > follows:
> >
> > mkdir out
> > /usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc remapped.nc
> > GridStatConfig -outdir out -v 3
> >
> > And I've attached the GridStatConfig file I used in which I
requested
> > output for 2x2 contingency tables, NxN multi-category contingency
tables,
> > and continuous statistics. Looking at the range of values present
in
> your
> > data, I somewhat arbitrarily chose thresholds of:
> > cat_thresh = [ >10, >20, >30, >40, >50 ];
> >
> > But here's the important part of it which tells Grid-Stat what
data to
> > process:
> > fcst = {
> > file_type=NETCDF_NCCF;
> > field = [
> > { name = "rf"; level = [ "(*,*)" ]; }
> > ];
> > }
> > obs = {
> > field = [
> > { name = "rf"; level = [ "(0,*,*)" ]; }
> > ];
> > }
> >
> > I've also attached a sample resulting output file containing the
2x2
> > contingency table counts for the 5 thresholds I chose:
> > out/grid_stat_000000L_19700101_000000V_ctc.txt
> >
> > Notice that the timestamp indicates a time of 19700101, which is
the same
> > as unixtime = 0. This happens because the remappedmodel.nc file
lacks
> > timing information.
> >
> > Hope this helps get you going.
> >
> > Thanks,
> > John Halley Gotway
> >
> > On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT <
> met_help at ucar.edu
> > >
> > wrote:
> >
> > >
> > > Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> > > Transaction: Owner forcibly changed from jensen (Tara Jensen) to
johnhg
> > > (John Halley Gotway) by jpresto
> > > Queue: met_help
> > > Subject: What are needed to run any executables
> > > Owner: johnhg
> > > Requestors: rabbani.du93 at gmail.com
> > > Status: open
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> >
> > >
> > >
> > > This transaction appears to have no content
> > >
> >
> >
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: Golam Rabbani
Time: Wed Jul 14 11:33:12 2021
Dear Sir,
Thank you very much for your reply and your valuable information.
Sir, i am really sorry to say that it is hard for me with my little
knowledge to configure wrf-arw in your prescribed way.
I need to plot a time series of ETS and FSS for three neighborhood
radii
and three thresholds (2.5, 5 and 10 mmm / hour for rainfall). Any
global
rainfall source is ok to do this task. You can suggest any other
rainfall
source for it. I have attached my namelist.input of WRF-ARW. Please
check
it and tell me what i can do in this regard.
*Sir, for calculating FSS and ETS using grid stat tool where i should
set
neighborhood radius such as 1*dx, 3*dx and 5*dx (where dx= 3 km
horizontal
grid spacing in my case). Should I set width = [ 1 ] for 1*dx
and
width = [ 2 ] for 2*dx and so on? I have attached my previous
grid
stat config script.
Greetings and thank you again.
Sincerely,
Golam Rabbani
On Wed, Jul 14, 2021 at 10:04 PM John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Golam,
>
> I see that you have a question about precipitation accumulation
intervals.
> Your verifying IMERG analysis files contain hourly accumulations of
precip
> while your WRF-ARW output contains a "runtime" accumulation of
precip,
> where forecast hour N contains all precip between forecast hour 0
and N.
>
> Yes, it's definitely important to match up the forecast and analysis
> precipitation accumulations before comparing them. You have a few
options
> here.
>
> (1) Technically, you could reconfigure WRF-ARW to write hourly
> precipitation accumulations to its output instead of runtime
accumulations.
> If you have full control of the WRF configuration then this could be
a good
> option and make comparisons easier. If you need more info on this, I
could
> check on the corresponding WRF namelist option and get back to you.
>
> (2) You could run MET's pcp_combine tool in "subtract" mode to
convert your
> WRF runtime accumulations into hourly output. Of course, you're
limited by
> the frequency of your WRF output times. If writing output hourly,
then you
> could get hourly precip by running pcp_combine to subtract forecast
hour N
> minus forecast hour N - 1. If writing 3-hourly output, you could do
> forecast hour N minus forecast hour N - 3.
>
> (3) You could run MET's pcp_combine tool in "add" mode to sum up the
hourly
> IMERG data into larger bins.
>
> (4) You could do some combination of (2) and (3) to put the forecast
and
> analysis data into a common accumulation interval.
>
> Hopefully that helps clarify some options.
>
> Thanks,
> John
>
> On Tue, Jul 13, 2021 at 1:50 PM Golam Rabbani via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> >
> > Dear altruists,
> >
> > I am trying to plot a time series of the FSS and ETS for selected
three
> > neighborhood radii (1*dx, 3*dx and 5*dx where dx=horizontal grid
spacing)
> > and for two to three thresholds (2.5, 5 and 10 mmm / hour for
rainfall. I
> > am using WRF-ARW model rainfall against NASA's hourly GPM IMERG
Final
> > Precipitation datasets. Previously, I could calculate skill scores
with
> > your prescribed script for daily accumulated rainfall.
> > Now I am confused how I can calculate the time series of the ETS
and FSS
> as
> > wrf model simulated rainfall in accumulated form in every time
step
> whereas
> > GPM IMERG is just in hourly observation. Can you please provide me
a
> > guideline with a script to do my task using the Grid_stat tool in
MET?
> >
> > Greetings and thank you in advance.
> >
> > Sincerely,
> > Golam Rabbani
> > Department of Meteorology
> > University of Dhaka, Bangladesh
> >
> > On Thu, Oct 17, 2019 at 3:35 AM John Halley Gotway via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > > Hi Golam,
> > >
> > > I see you're having trouble running your NetCDF output files
through
> MET.
> > >
> > > Whenever getting started with a new gridded dataset, I always
start by
> > > running them through MET's plot_data_plane tool. For the file
named
> > > remappedmodel.nc, I had to explicitly tell MET to interpret it
as
> being
> > a
> > > CF-compliant NetCDF file because the standard global
> > > "Conventions" attribute was not present, like it is in the one
named
> > > remapped.nc:
> > > :Conventions = "CF-1.4" ;
> > >
> > > Here are the 2 plot_data_plane commands I ran, and I've attached
png
> > > versions of the output:
> > > /usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
> > remappedmodel.ps
> > > 'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
> > > /usr/local/met-8.1.1/bin/plot_data_plane remapped.nc remapped.ps
> > > 'name="rf"; level="(0,*,*)";'
> > >
> > > Note also that the first command produces a warning because no
timing
> > > information is present in the file:
> > > WARNING: NcCfFile::open() -> could not determine the valid time,
using
> 0.
> > >
> > > Next, I ran the grid_stat utility to compute categorical
statistics, as
> > > follows:
> > >
> > > mkdir out
> > > /usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc remapped.nc
> > > GridStatConfig -outdir out -v 3
> > >
> > > And I've attached the GridStatConfig file I used in which I
requested
> > > output for 2x2 contingency tables, NxN multi-category
contingency
> tables,
> > > and continuous statistics. Looking at the range of values
present in
> > your
> > > data, I somewhat arbitrarily chose thresholds of:
> > > cat_thresh = [ >10, >20, >30, >40, >50 ];
> > >
> > > But here's the important part of it which tells Grid-Stat what
data to
> > > process:
> > > fcst = {
> > > file_type=NETCDF_NCCF;
> > > field = [
> > > { name = "rf"; level = [ "(*,*)" ]; }
> > > ];
> > > }
> > > obs = {
> > > field = [
> > > { name = "rf"; level = [ "(0,*,*)" ]; }
> > > ];
> > > }
> > >
> > > I've also attached a sample resulting output file containing the
2x2
> > > contingency table counts for the 5 thresholds I chose:
> > > out/grid_stat_000000L_19700101_000000V_ctc.txt
> > >
> > > Notice that the timestamp indicates a time of 19700101, which is
the
> same
> > > as unixtime = 0. This happens because the remappedmodel.nc file
lacks
> > > timing information.
> > >
> > > Hope this helps get you going.
> > >
> > > Thanks,
> > > John Halley Gotway
> > >
> > > On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT <
> > met_help at ucar.edu
> > > >
> > > wrote:
> > >
> > > >
> > > > Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> > > > Transaction: Owner forcibly changed from jensen (Tara Jensen)
to
> johnhg
> > > > (John Halley Gotway) by jpresto
> > > > Queue: met_help
> > > > Subject: What are needed to run any executables
> > > > Owner: johnhg
> > > > Requestors: rabbani.du93 at gmail.com
> > > > Status: open
> > > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> > >
> > > >
> > > >
> > > > This transaction appears to have no content
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: John Halley Gotway
Time: Wed Jul 14 12:04:14 2021
Golam,
The WRF-ARW namelist option which controls the output precipitation
accumulation interval is named "prec_acc_dt" in the physics section.
You
can read about it here:
https://www2.mmm.ucar.edu/wrf/users/docs/user_guide_v4/v4.3/users_guide_chap5.html
As for the Grid-Stat config file, you'd edit the "nbrhd" dictionary to
define what neighborhood widths should be processed. I've listed 3, 5,
7,
and 9 grid points below.
nbrhd = {
field = BOTH;
vld_thresh = 1.0;
shape = SQUARE;
width = [ 3, 5, 7, 9 ];
cov_thresh = [ >=0.5 ];
}
Hopefully that'll help.
John
On Wed, Jul 14, 2021 at 11:33 AM Golam Rabbani via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
>
> Dear Sir,
> Thank you very much for your reply and your valuable information.
>
> Sir, i am really sorry to say that it is hard for me with my little
> knowledge to configure wrf-arw in your prescribed way.
> I need to plot a time series of ETS and FSS for three neighborhood
radii
> and three thresholds (2.5, 5 and 10 mmm / hour for rainfall). Any
global
> rainfall source is ok to do this task. You can suggest any other
rainfall
> source for it. I have attached my namelist.input of WRF-ARW. Please
check
> it and tell me what i can do in this regard.
>
> *Sir, for calculating FSS and ETS using grid stat tool where i
should set
> neighborhood radius such as 1*dx, 3*dx and 5*dx (where dx= 3 km
horizontal
> grid spacing in my case). Should I set width = [ 1 ] for
1*dx and
> width = [ 2 ] for 2*dx and so on? I have attached my previous
grid
> stat config script.
>
> Greetings and thank you again.
>
> Sincerely,
> Golam Rabbani
>
>
>
> On Wed, Jul 14, 2021 at 10:04 PM John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Golam,
> >
> > I see that you have a question about precipitation accumulation
> intervals.
> > Your verifying IMERG analysis files contain hourly accumulations
of
> precip
> > while your WRF-ARW output contains a "runtime" accumulation of
precip,
> > where forecast hour N contains all precip between forecast hour 0
and N.
> >
> > Yes, it's definitely important to match up the forecast and
analysis
> > precipitation accumulations before comparing them. You have a few
options
> > here.
> >
> > (1) Technically, you could reconfigure WRF-ARW to write hourly
> > precipitation accumulations to its output instead of runtime
> accumulations.
> > If you have full control of the WRF configuration then this could
be a
> good
> > option and make comparisons easier. If you need more info on this,
I
> could
> > check on the corresponding WRF namelist option and get back to
you.
> >
> > (2) You could run MET's pcp_combine tool in "subtract" mode to
convert
> your
> > WRF runtime accumulations into hourly output. Of course, you're
limited
> by
> > the frequency of your WRF output times. If writing output hourly,
then
> you
> > could get hourly precip by running pcp_combine to subtract
forecast hour
> N
> > minus forecast hour N - 1. If writing 3-hourly output, you could
do
> > forecast hour N minus forecast hour N - 3.
> >
> > (3) You could run MET's pcp_combine tool in "add" mode to sum up
the
> hourly
> > IMERG data into larger bins.
> >
> > (4) You could do some combination of (2) and (3) to put the
forecast and
> > analysis data into a common accumulation interval.
> >
> > Hopefully that helps clarify some options.
> >
> > Thanks,
> > John
> >
> > On Tue, Jul 13, 2021 at 1:50 PM Golam Rabbani via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> > >
> > > Dear altruists,
> > >
> > > I am trying to plot a time series of the FSS and ETS for
selected three
> > > neighborhood radii (1*dx, 3*dx and 5*dx where dx=horizontal grid
> spacing)
> > > and for two to three thresholds (2.5, 5 and 10 mmm / hour for
> rainfall. I
> > > am using WRF-ARW model rainfall against NASA's hourly GPM IMERG
Final
> > > Precipitation datasets. Previously, I could calculate skill
scores with
> > > your prescribed script for daily accumulated rainfall.
> > > Now I am confused how I can calculate the time series of the ETS
and
> FSS
> > as
> > > wrf model simulated rainfall in accumulated form in every time
step
> > whereas
> > > GPM IMERG is just in hourly observation. Can you please provide
me a
> > > guideline with a script to do my task using the Grid_stat tool
in MET?
> > >
> > > Greetings and thank you in advance.
> > >
> > > Sincerely,
> > > Golam Rabbani
> > > Department of Meteorology
> > > University of Dhaka, Bangladesh
> > >
> > > On Thu, Oct 17, 2019 at 3:35 AM John Halley Gotway via RT <
> > > met_help at ucar.edu>
> > > wrote:
> > >
> > > > Hi Golam,
> > > >
> > > > I see you're having trouble running your NetCDF output files
through
> > MET.
> > > >
> > > > Whenever getting started with a new gridded dataset, I always
start
> by
> > > > running them through MET's plot_data_plane tool. For the file
named
> > > > remappedmodel.nc, I had to explicitly tell MET to interpret it
as
> > being
> > > a
> > > > CF-compliant NetCDF file because the standard global
> > > > "Conventions" attribute was not present, like it is in the one
named
> > > > remapped.nc:
> > > > :Conventions = "CF-1.4" ;
> > > >
> > > > Here are the 2 plot_data_plane commands I ran, and I've
attached png
> > > > versions of the output:
> > > > /usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
> > > remappedmodel.ps
> > > > 'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
> > > > /usr/local/met-8.1.1/bin/plot_data_plane remapped.nc
remapped.ps
> > > > 'name="rf"; level="(0,*,*)";'
> > > >
> > > > Note also that the first command produces a warning because no
timing
> > > > information is present in the file:
> > > > WARNING: NcCfFile::open() -> could not determine the valid
time,
> using
> > 0.
> > > >
> > > > Next, I ran the grid_stat utility to compute categorical
statistics,
> as
> > > > follows:
> > > >
> > > > mkdir out
> > > > /usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc
remapped.nc
> > > > GridStatConfig -outdir out -v 3
> > > >
> > > > And I've attached the GridStatConfig file I used in which I
requested
> > > > output for 2x2 contingency tables, NxN multi-category
contingency
> > tables,
> > > > and continuous statistics. Looking at the range of values
present in
> > > your
> > > > data, I somewhat arbitrarily chose thresholds of:
> > > > cat_thresh = [ >10, >20, >30, >40, >50 ];
> > > >
> > > > But here's the important part of it which tells Grid-Stat what
data
> to
> > > > process:
> > > > fcst = {
> > > > file_type=NETCDF_NCCF;
> > > > field = [
> > > > { name = "rf"; level = [ "(*,*)" ]; }
> > > > ];
> > > > }
> > > > obs = {
> > > > field = [
> > > > { name = "rf"; level = [ "(0,*,*)" ]; }
> > > > ];
> > > > }
> > > >
> > > > I've also attached a sample resulting output file containing
the 2x2
> > > > contingency table counts for the 5 thresholds I chose:
> > > > out/grid_stat_000000L_19700101_000000V_ctc.txt
> > > >
> > > > Notice that the timestamp indicates a time of 19700101, which
is the
> > same
> > > > as unixtime = 0. This happens because the remappedmodel.nc
file
> lacks
> > > > timing information.
> > > >
> > > > Hope this helps get you going.
> > > >
> > > > Thanks,
> > > > John Halley Gotway
> > > >
> > > > On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT <
> > > met_help at ucar.edu
> > > > >
> > > > wrote:
> > > >
> > > > >
> > > > > Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> > > > > Transaction: Owner forcibly changed from jensen (Tara
Jensen) to
> > johnhg
> > > > > (John Halley Gotway) by jpresto
> > > > > Queue: met_help
> > > > > Subject: What are needed to run any executables
> > > > > Owner: johnhg
> > > > > Requestors: rabbani.du93 at gmail.com
> > > > > Status: open
> > > > > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> > > >
> > > > >
> > > > >
> > > > > This transaction appears to have no content
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: What are needed to run any executables
From: Golam Rabbani
Time: Wed Jul 28 14:45:55 2021
Thank you very much. Finally I have plotted my desired graph with your
assistance.
On Thu, Jul 15, 2021 at 12:04 AM John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Golam,
>
> The WRF-ARW namelist option which controls the output precipitation
> accumulation interval is named "prec_acc_dt" in the physics section.
You
> can read about it here:
>
>
>
https://www2.mmm.ucar.edu/wrf/users/docs/user_guide_v4/v4.3/users_guide_chap5.html
>
> As for the Grid-Stat config file, you'd edit the "nbrhd" dictionary
to
> define what neighborhood widths should be processed. I've listed 3,
5, 7,
> and 9 grid points below.
>
> nbrhd = {
> field = BOTH;
> vld_thresh = 1.0;
> shape = SQUARE;
> width = [ 3, 5, 7, 9 ];
> cov_thresh = [ >=0.5 ];
> }
>
> Hopefully that'll help.
>
> John
>
> On Wed, Jul 14, 2021 at 11:33 AM Golam Rabbani via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484 >
> >
> > Dear Sir,
> > Thank you very much for your reply and your valuable information.
> >
> > Sir, i am really sorry to say that it is hard for me with my
little
> > knowledge to configure wrf-arw in your prescribed way.
> > I need to plot a time series of ETS and FSS for three neighborhood
radii
> > and three thresholds (2.5, 5 and 10 mmm / hour for rainfall). Any
global
> > rainfall source is ok to do this task. You can suggest any other
rainfall
> > source for it. I have attached my namelist.input of WRF-ARW.
Please check
> > it and tell me what i can do in this regard.
> >
> > *Sir, for calculating FSS and ETS using grid stat tool where i
should set
> > neighborhood radius such as 1*dx, 3*dx and 5*dx (where dx= 3 km
> horizontal
> > grid spacing in my case). Should I set width = [ 1 ] for
1*dx and
> > width = [ 2 ] for 2*dx and so on? I have attached my previous
grid
> > stat config script.
> >
> > Greetings and thank you again.
> >
> > Sincerely,
> > Golam Rabbani
> >
> >
> >
> > On Wed, Jul 14, 2021 at 10:04 PM John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Golam,
> > >
> > > I see that you have a question about precipitation accumulation
> > intervals.
> > > Your verifying IMERG analysis files contain hourly accumulations
of
> > precip
> > > while your WRF-ARW output contains a "runtime" accumulation of
precip,
> > > where forecast hour N contains all precip between forecast hour
0 and
> N.
> > >
> > > Yes, it's definitely important to match up the forecast and
analysis
> > > precipitation accumulations before comparing them. You have a
few
> options
> > > here.
> > >
> > > (1) Technically, you could reconfigure WRF-ARW to write hourly
> > > precipitation accumulations to its output instead of runtime
> > accumulations.
> > > If you have full control of the WRF configuration then this
could be a
> > good
> > > option and make comparisons easier. If you need more info on
this, I
> > could
> > > check on the corresponding WRF namelist option and get back to
you.
> > >
> > > (2) You could run MET's pcp_combine tool in "subtract" mode to
convert
> > your
> > > WRF runtime accumulations into hourly output. Of course, you're
limited
> > by
> > > the frequency of your WRF output times. If writing output
hourly, then
> > you
> > > could get hourly precip by running pcp_combine to subtract
forecast
> hour
> > N
> > > minus forecast hour N - 1. If writing 3-hourly output, you could
do
> > > forecast hour N minus forecast hour N - 3.
> > >
> > > (3) You could run MET's pcp_combine tool in "add" mode to sum up
the
> > hourly
> > > IMERG data into larger bins.
> > >
> > > (4) You could do some combination of (2) and (3) to put the
forecast
> and
> > > analysis data into a common accumulation interval.
> > >
> > > Hopefully that helps clarify some options.
> > >
> > > Thanks,
> > > John
> > >
> > > On Tue, Jul 13, 2021 at 1:50 PM Golam Rabbani via RT <
> met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
>
> > > >
> > > > Dear altruists,
> > > >
> > > > I am trying to plot a time series of the FSS and ETS for
selected
> three
> > > > neighborhood radii (1*dx, 3*dx and 5*dx where dx=horizontal
grid
> > spacing)
> > > > and for two to three thresholds (2.5, 5 and 10 mmm / hour for
> > rainfall. I
> > > > am using WRF-ARW model rainfall against NASA's hourly GPM
IMERG Final
> > > > Precipitation datasets. Previously, I could calculate skill
scores
> with
> > > > your prescribed script for daily accumulated rainfall.
> > > > Now I am confused how I can calculate the time series of the
ETS and
> > FSS
> > > as
> > > > wrf model simulated rainfall in accumulated form in every time
step
> > > whereas
> > > > GPM IMERG is just in hourly observation. Can you please
provide me a
> > > > guideline with a script to do my task using the Grid_stat tool
in
> MET?
> > > >
> > > > Greetings and thank you in advance.
> > > >
> > > > Sincerely,
> > > > Golam Rabbani
> > > > Department of Meteorology
> > > > University of Dhaka, Bangladesh
> > > >
> > > > On Thu, Oct 17, 2019 at 3:35 AM John Halley Gotway via RT <
> > > > met_help at ucar.edu>
> > > > wrote:
> > > >
> > > > > Hi Golam,
> > > > >
> > > > > I see you're having trouble running your NetCDF output files
> through
> > > MET.
> > > > >
> > > > > Whenever getting started with a new gridded dataset, I
always start
> > by
> > > > > running them through MET's plot_data_plane tool. For the
file
> named
> > > > > remappedmodel.nc, I had to explicitly tell MET to interpret
it as
> > > being
> > > > a
> > > > > CF-compliant NetCDF file because the standard global
> > > > > "Conventions" attribute was not present, like it is in the
one
> named
> > > > > remapped.nc:
> > > > > :Conventions = "CF-1.4" ;
> > > > >
> > > > > Here are the 2 plot_data_plane commands I ran, and I've
attached
> png
> > > > > versions of the output:
> > > > > /usr/local/met-8.1.1/bin/plot_data_plane remappedmodel.nc
> > > > remappedmodel.ps
> > > > > 'name="rf"; level="(*,*)"; file_type=NETCDF_NCCF;'
> > > > > /usr/local/met-8.1.1/bin/plot_data_plane remapped.nc
remapped.ps
> > > > > 'name="rf"; level="(0,*,*)";'
> > > > >
> > > > > Note also that the first command produces a warning because
no
> timing
> > > > > information is present in the file:
> > > > > WARNING: NcCfFile::open() -> could not determine the valid
time,
> > using
> > > 0.
> > > > >
> > > > > Next, I ran the grid_stat utility to compute categorical
> statistics,
> > as
> > > > > follows:
> > > > >
> > > > > mkdir out
> > > > > /usr/local/met-8.1.1/bin/grid_stat remappedmodel.nc
remapped.nc
> > > > > GridStatConfig -outdir out -v 3
> > > > >
> > > > > And I've attached the GridStatConfig file I used in which I
> requested
> > > > > output for 2x2 contingency tables, NxN multi-category
contingency
> > > tables,
> > > > > and continuous statistics. Looking at the range of values
present
> in
> > > > your
> > > > > data, I somewhat arbitrarily chose thresholds of:
> > > > > cat_thresh = [ >10, >20, >30, >40, >50 ];
> > > > >
> > > > > But here's the important part of it which tells Grid-Stat
what data
> > to
> > > > > process:
> > > > > fcst = {
> > > > > file_type=NETCDF_NCCF;
> > > > > field = [
> > > > > { name = "rf"; level = [ "(*,*)" ]; }
> > > > > ];
> > > > > }
> > > > > obs = {
> > > > > field = [
> > > > > { name = "rf"; level = [ "(0,*,*)" ]; }
> > > > > ];
> > > > > }
> > > > >
> > > > > I've also attached a sample resulting output file containing
the
> 2x2
> > > > > contingency table counts for the 5 thresholds I chose:
> > > > > out/grid_stat_000000L_19700101_000000V_ctc.txt
> > > > >
> > > > > Notice that the timestamp indicates a time of 19700101,
which is
> the
> > > same
> > > > > as unixtime = 0. This happens because the remappedmodel.nc
file
> > lacks
> > > > > timing information.
> > > > >
> > > > > Hope this helps get you going.
> > > > >
> > > > > Thanks,
> > > > > John Halley Gotway
> > > > >
> > > > > On Wed, Oct 16, 2019 at 2:53 PM Julie Prestopnik via RT <
> > > > met_help at ucar.edu
> > > > > >
> > > > > wrote:
> > > > >
> > > > > >
> > > > > > Wed Oct 16 14:52:42 2019: Request 90484 was acted upon.
> > > > > > Transaction: Owner forcibly changed from jensen (Tara
Jensen) to
> > > johnhg
> > > > > > (John Halley Gotway) by jpresto
> > > > > > Queue: met_help
> > > > > > Subject: What are needed to run any executables
> > > > > > Owner: johnhg
> > > > > > Requestors: rabbani.du93 at gmail.com
> > > > > > Status: open
> > > > > > Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90484
> > > > >
> > > > > >
> > > > > >
> > > > > > This transaction appears to have no content
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
More information about the Met_help
mailing list