[Met_help] [rt.rap.ucar.edu #76495] History for Persistence Data
John Halley Gotway via RT
met_help at ucar.edu
Tue Jun 7 11:11:14 MDT 2016
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
John or Julie, I need to use persistence data as my reference forecast for the HiRA verification I am working on. It will be used to calculated Brier Skill Scores. The documentation mentions the use of climo data for this. Is it as simple as providing a persistence forecast for the climo mean filename and field?
Also I want to put a plug in for determining how to get MODIS regrid to compile. We have a guy here who wants to use it in dust verification.
Thanks
Bob
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Persistence Data
From: John Halley Gotway
Time: Thu May 26 13:54:09 2016
Bob,
I emailed Matt this morning with info on compiling modis_regrid.
Regarding persistence...
I see that you'd like to compute BSS in the HiRA framework as the
brier
score of the model relative to persistence. Correct me if I'm wrong,
but
you've written code to generate the MPR lines to mimic the output of
Point-Stat. The second to last column of the MPR line type is named
"CLIMO". I'd suggest that you fill that column with that persistence
value.
*BUT* this has an implication in the logic of processing this data
through
STAT-Analysis.
When we run MPR -> PCT, the PCT line type no longer contains
information
about persistence. Instead, you'd need to go directly from MPR ->
PSTD...
since MPR has climo values, it should use that in the computation of
BSS
for the output PSTD line type.
However, I haven't string the MET tools together in this order before.
When you get some sample MPR data which contains persistence data in
the
CLIMO column, I'd love to take a look. Then I could test it out with
STAT-Analysis to make sure it's working as expected.
Thanks,
John
On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT <
met_help at ucar.edu> wrote:
>
> Wed May 25 12:45:15 2016: Request 76495 was acted upon.
> Transaction: Ticket created by robert.craig.2 at us.af.mil
> Queue: met_help
> Subject: Persistence Data
> Owner: Nobody
> Requestors: robert.craig.2 at us.af.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
>
>
> John or Julie, I need to use persistence data as my reference
forecast for
> the HiRA verification I am working on. It will be used to
calculated Brier
> Skill Scores. The documentation mentions the use of climo data for
this.
> Is it as simple as providing a persistence forecast for the climo
mean
> filename and field?
>
> Also I want to put a plug in for determining how to get MODIS regrid
to
> compile. We have a guy here who wants to use it in dust
verification.
>
> Thanks
> Bob
>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #76495] Persistence Data
From: robert.craig.2 at us.af.mil
Time: Thu May 26 14:03:26 2016
John, thanks for all your help on this. I was creating a second set
of MPR and PCT for the persistence values. But I like your idea
better if it works. One thing with the changes we have made. I no
longer have the forecast thresholds in the PCT line. They got
replaced by the contingency table threshold values. Also, in the MPR
lines I had placed 3 for my neighborhood size but it became 4 in the
PCT file - weird. I might have to keep the different thresholds
separate in separate files - not the preferred option.
Bob
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, May 26, 2016 2:54 PM
To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
<robert.craig.2 at us.af.mil>
Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL
<matthew.sittel.ctr at us.af.mil>
Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
Bob,
I emailed Matt this morning with info on compiling modis_regrid.
Regarding persistence...
I see that you'd like to compute BSS in the HiRA framework as the
brier score of the model relative to persistence. Correct me if I'm
wrong, but you've written code to generate the MPR lines to mimic the
output of Point-Stat. The second to last column of the MPR line type
is named "CLIMO". I'd suggest that you fill that column with that
persistence value.
*BUT* this has an implication in the logic of processing this data
through STAT-Analysis.
When we run MPR -> PCT, the PCT line type no longer contains
information about persistence. Instead, you'd need to go directly
from MPR -> PSTD...
since MPR has climo values, it should use that in the computation of
BSS for the output PSTD line type.
However, I haven't string the MET tools together in this order before.
When you get some sample MPR data which contains persistence data in
the CLIMO column, I'd love to take a look. Then I could test it out
with STAT-Analysis to make sure it's working as expected.
Thanks,
John
On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT <
met_help at ucar.edu> wrote:
>
> Wed May 25 12:45:15 2016: Request 76495 was acted upon.
> Transaction: Ticket created by robert.craig.2 at us.af.mil
> Queue: met_help
> Subject: Persistence Data
> Owner: Nobody
> Requestors: robert.craig.2 at us.af.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495
> >
>
>
> John or Julie, I need to use persistence data as my reference
forecast
> for the HiRA verification I am working on. It will be used to
> calculated Brier Skill Scores. The documentation mentions the use
of climo data for this.
> Is it as simple as providing a persistence forecast for the climo
mean
> filename and field?
>
> Also I want to put a plug in for determining how to get MODIS regrid
> to compile. We have a guy here who wants to use it in dust
verification.
>
> Thanks
> Bob
>
>
>
------------------------------------------------
Subject: Persistence Data
From: John Halley Gotway
Time: Thu May 26 14:12:42 2016
Bob,
If you're able to send me a sample file along with the stat-analysis
command that make the 3 change to 4, I could take a look and figure
out
what's going on.
John
On Thu, May 26, 2016 at 2:03 PM, robert.craig.2 at us.af.mil via RT <
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
>
> John, thanks for all your help on this. I was creating a second set
of
> MPR and PCT for the persistence values. But I like your idea better
if it
> works. One thing with the changes we have made. I no longer have
the
> forecast thresholds in the PCT line. They got replaced by the
contingency
> table threshold values. Also, in the MPR lines I had placed 3 for
my
> neighborhood size but it became 4 in the PCT file - weird. I might
have to
> keep the different thresholds separate in separate files - not the
> preferred option.
>
> Bob
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, May 26, 2016 2:54 PM
> To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
<robert.craig.2 at us.af.mil>
> Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL <
> matthew.sittel.ctr at us.af.mil>
> Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
>
> Bob,
>
> I emailed Matt this morning with info on compiling modis_regrid.
>
> Regarding persistence...
>
> I see that you'd like to compute BSS in the HiRA framework as the
brier
> score of the model relative to persistence. Correct me if I'm
wrong, but
> you've written code to generate the MPR lines to mimic the output of
> Point-Stat. The second to last column of the MPR line type is named
> "CLIMO". I'd suggest that you fill that column with that
persistence value.
>
> *BUT* this has an implication in the logic of processing this data
through
> STAT-Analysis.
>
> When we run MPR -> PCT, the PCT line type no longer contains
information
> about persistence. Instead, you'd need to go directly from MPR ->
PSTD...
> since MPR has climo values, it should use that in the computation of
BSS
> for the output PSTD line type.
>
> However, I haven't string the MET tools together in this order
before.
> When you get some sample MPR data which contains persistence data in
the
> CLIMO column, I'd love to take a look. Then I could test it out
with
> STAT-Analysis to make sure it's working as expected.
>
> Thanks,
> John
>
> On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Wed May 25 12:45:15 2016: Request 76495 was acted upon.
> > Transaction: Ticket created by robert.craig.2 at us.af.mil
> > Queue: met_help
> > Subject: Persistence Data
> > Owner: Nobody
> > Requestors: robert.craig.2 at us.af.mil
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495
> > >
> >
> >
> > John or Julie, I need to use persistence data as my reference
forecast
> > for the HiRA verification I am working on. It will be used to
> > calculated Brier Skill Scores. The documentation mentions the use
of
> climo data for this.
> > Is it as simple as providing a persistence forecast for the climo
mean
> > filename and field?
> >
> > Also I want to put a plug in for determining how to get MODIS
regrid
> > to compile. We have a guy here who wants to use it in dust
verification.
> >
> > Thanks
> > Bob
> >
> >
> >
>
>
>
>
------------------------------------------------
Subject: Persistence Data
From: John Halley Gotway
Time: Thu May 26 14:15:12 2016
Also, I should mention that we do have an idea for future development
in
STAT-Analysis to would apply to this case, but it doesn't currently
exist.
Basically, pass STAT-Analysis MET output from two different models,
have it
generate some statistics for each model, and then have it compute
skill
scores as the performance of one model relative to another.
Would that sort of logic be useful?
John
On Thu, May 26, 2016 at 2:12 PM, John Halley Gotway <johnhg at ucar.edu>
wrote:
> Bob,
>
> If you're able to send me a sample file along with the stat-analysis
> command that make the 3 change to 4, I could take a look and figure
out
> what's going on.
>
> John
>
> On Thu, May 26, 2016 at 2:03 PM, robert.craig.2 at us.af.mil via RT <
> met_help at ucar.edu> wrote:
>
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
>>
>> John, thanks for all your help on this. I was creating a second
set of
>> MPR and PCT for the persistence values. But I like your idea
better if it
>> works. One thing with the changes we have made. I no longer have
the
>> forecast thresholds in the PCT line. They got replaced by the
contingency
>> table threshold values. Also, in the MPR lines I had placed 3 for
my
>> neighborhood size but it became 4 in the PCT file - weird. I might
have to
>> keep the different thresholds separate in separate files - not the
>> preferred option.
>>
>> Bob
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Thursday, May 26, 2016 2:54 PM
>> To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
<robert.craig.2 at us.af.mil>
>> Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL <
>> matthew.sittel.ctr at us.af.mil>
>> Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
>>
>> Bob,
>>
>> I emailed Matt this morning with info on compiling modis_regrid.
>>
>> Regarding persistence...
>>
>> I see that you'd like to compute BSS in the HiRA framework as the
brier
>> score of the model relative to persistence. Correct me if I'm
wrong, but
>> you've written code to generate the MPR lines to mimic the output
of
>> Point-Stat. The second to last column of the MPR line type is
named
>> "CLIMO". I'd suggest that you fill that column with that
persistence value.
>>
>> *BUT* this has an implication in the logic of processing this data
>> through STAT-Analysis.
>>
>> When we run MPR -> PCT, the PCT line type no longer contains
information
>> about persistence. Instead, you'd need to go directly from MPR ->
PSTD...
>> since MPR has climo values, it should use that in the computation
of BSS
>> for the output PSTD line type.
>>
>> However, I haven't string the MET tools together in this order
before.
>> When you get some sample MPR data which contains persistence data
in the
>> CLIMO column, I'd love to take a look. Then I could test it out
with
>> STAT-Analysis to make sure it's working as expected.
>>
>> Thanks,
>> John
>>
>> On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT <
>> met_help at ucar.edu> wrote:
>>
>> >
>> > Wed May 25 12:45:15 2016: Request 76495 was acted upon.
>> > Transaction: Ticket created by robert.craig.2 at us.af.mil
>> > Queue: met_help
>> > Subject: Persistence Data
>> > Owner: Nobody
>> > Requestors: robert.craig.2 at us.af.mil
>> > Status: new
>> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495
>> > >
>> >
>> >
>> > John or Julie, I need to use persistence data as my reference
forecast
>> > for the HiRA verification I am working on. It will be used to
>> > calculated Brier Skill Scores. The documentation mentions the
use of
>> climo data for this.
>> > Is it as simple as providing a persistence forecast for the climo
mean
>> > filename and field?
>> >
>> > Also I want to put a plug in for determining how to get MODIS
regrid
>> > to compile. We have a guy here who wants to use it in dust
>> verification.
>> >
>> > Thanks
>> > Bob
>> >
>> >
>> >
>>
>>
>>
>>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #76495] Persistence Data
From: robert.craig.2 at us.af.mil
Time: Thu May 26 14:45:18 2016
John, I placed the MPR and PCT files on the server. The command line
for the MPR -> PCT is (python list):
['/h/WXQC/met-5.1/bin/stat_analysis', '-lookin',
'/h/data/global/WXQC/data/met/mdlob_pairs/regular', '-out_stat',
'/h/data/global/WXQC/data/met/ens_cont_tbl/GALWEM_APCP_12hr_1_3_PCT_0Z_pct.stat',
'-config', '/h/WXQC/met-
5.1/data/config/STATAnalysisConfig_hira_bs_updated', '-out_fcst_thresh
ge0,ge0.1,ge0.2,ge0.3,ge0.4,ge0.5,ge0.6,ge0.7,ge0.8,ge0.9,ge1.0', '-
out_obs_thresh eq1', '-by FCST_VAR', '-by FCST_THRESH', '-v', '6']
The command line for the PCT to PSTD is:
['/h/WXQC/met-5.1/bin/stat_analysis', '-lookin',
'/h/data/global/WXQC/data/met/ens_cont_tbl', '-out',
'/h/data/global/WXQC/data/met/summary/GALWEM_APCP_12hr_50_3_PSTD_0Z',
'-job aggregate_stat', '-line_type PCT', '-out_line_type PSTD', '-by
FCST_VAR', '-by FCST_THRESH', '-v', '6']
Hope the helps. As for the question - your idea for stat_anal changes
would be useful - it definitely would make my task easier.
Bob
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, May 26, 2016 3:15 PM
To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
<robert.craig.2 at us.af.mil>
Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL
<matthew.sittel.ctr at us.af.mil>
Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
Also, I should mention that we do have an idea for future development
in STAT-Analysis to would apply to this case, but it doesn't currently
exist.
Basically, pass STAT-Analysis MET output from two different models,
have it generate some statistics for each model, and then have it
compute skill scores as the performance of one model relative to
another.
Would that sort of logic be useful?
John
On Thu, May 26, 2016 at 2:12 PM, John Halley Gotway <johnhg at ucar.edu>
wrote:
> Bob,
>
> If you're able to send me a sample file along with the stat-analysis
> command that make the 3 change to 4, I could take a look and figure
> out what's going on.
>
> John
>
> On Thu, May 26, 2016 at 2:03 PM, robert.craig.2 at us.af.mil via RT <
> met_help at ucar.edu> wrote:
>
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
>>
>> John, thanks for all your help on this. I was creating a second
set
>> of MPR and PCT for the persistence values. But I like your idea
>> better if it works. One thing with the changes we have made. I no
>> longer have the forecast thresholds in the PCT line. They got
>> replaced by the contingency table threshold values. Also, in the
MPR
>> lines I had placed 3 for my neighborhood size but it became 4 in
the
>> PCT file - weird. I might have to keep the different thresholds
>> separate in separate files - not the preferred option.
>>
>> Bob
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Thursday, May 26, 2016 2:54 PM
>> To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
>> <robert.craig.2 at us.af.mil>
>> Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL <
>> matthew.sittel.ctr at us.af.mil>
>> Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
>>
>> Bob,
>>
>> I emailed Matt this morning with info on compiling modis_regrid.
>>
>> Regarding persistence...
>>
>> I see that you'd like to compute BSS in the HiRA framework as the
>> brier score of the model relative to persistence. Correct me if
I'm
>> wrong, but you've written code to generate the MPR lines to mimic
the
>> output of Point-Stat. The second to last column of the MPR line
type
>> is named "CLIMO". I'd suggest that you fill that column with that
persistence value.
>>
>> *BUT* this has an implication in the logic of processing this data
>> through STAT-Analysis.
>>
>> When we run MPR -> PCT, the PCT line type no longer contains
>> information about persistence. Instead, you'd need to go directly
from MPR -> PSTD...
>> since MPR has climo values, it should use that in the computation
of
>> BSS for the output PSTD line type.
>>
>> However, I haven't string the MET tools together in this order
before.
>> When you get some sample MPR data which contains persistence data
in
>> the CLIMO column, I'd love to take a look. Then I could test it
out
>> with STAT-Analysis to make sure it's working as expected.
>>
>> Thanks,
>> John
>>
>> On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT <
>> met_help at ucar.edu> wrote:
>>
>> >
>> > Wed May 25 12:45:15 2016: Request 76495 was acted upon.
>> > Transaction: Ticket created by robert.craig.2 at us.af.mil
>> > Queue: met_help
>> > Subject: Persistence Data
>> > Owner: Nobody
>> > Requestors: robert.craig.2 at us.af.mil
>> > Status: new
>> > Ticket <URL:
>> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495
>> > >
>> >
>> >
>> > John or Julie, I need to use persistence data as my reference
>> > forecast for the HiRA verification I am working on. It will be
>> > used to calculated Brier Skill Scores. The documentation
mentions
>> > the use of
>> climo data for this.
>> > Is it as simple as providing a persistence forecast for the climo
>> > mean filename and field?
>> >
>> > Also I want to put a plug in for determining how to get MODIS
>> > regrid to compile. We have a guy here who wants to use it in
dust
>> verification.
>> >
>> > Thanks
>> > Bob
>> >
>> >
>> >
>>
>>
>>
>>
>
------------------------------------------------
Subject: Persistence Data
From: John Halley Gotway
Time: Thu May 26 15:11:04 2016
Bob,
I see what's going on. When Point-Stat runs, the user defines the
interpolation width in the config file as N. For example, a value of
1
means the nearest point, 2 means a 2x2 box (so 4 points), 3 means a
3x3 box
(so 9 points). The INTERP_PNTS output column contains the value NxN.
When
STAT-Analysis parses that line it takes the square root of
INTERP_PNTS,
rounds that to the nearest integer, and stores that as the
interpolation
width, N. When it writes it's output, it fills the INTERP_PNTS column
with
NxN.
In your data, INTERP_PNTS is set to 3... the sqrt of 3 is 1.73... that
is
rounded to 2. Then 2x2 = 4 is written to the output INTERP_PNTS
columns.
When you write 3, do you really mean a 3x3 box... so 9 points? If so,
you
might consider writing 9 there instead for consistency with other
usage in
MET.
Thanks,
John
On Thu, May 26, 2016 at 2:45 PM, robert.craig.2 at us.af.mil via RT <
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
>
> John, I placed the MPR and PCT files on the server. The command
line for
> the MPR -> PCT is (python list):
>
> ['/h/WXQC/met-5.1/bin/stat_analysis', '-lookin',
> '/h/data/global/WXQC/data/met/mdlob_pairs/regular', '-out_stat',
>
'/h/data/global/WXQC/data/met/ens_cont_tbl/GALWEM_APCP_12hr_1_3_PCT_0Z_pct.stat',
> '-config',
> '/h/WXQC/met-5.1/data/config/STATAnalysisConfig_hira_bs_updated',
> '-out_fcst_thresh
> ge0,ge0.1,ge0.2,ge0.3,ge0.4,ge0.5,ge0.6,ge0.7,ge0.8,ge0.9,ge1.0',
> '-out_obs_thresh eq1', '-by FCST_VAR', '-by FCST_THRESH', '-v', '6']
>
> The command line for the PCT to PSTD is:
>
> ['/h/WXQC/met-5.1/bin/stat_analysis', '-lookin',
> '/h/data/global/WXQC/data/met/ens_cont_tbl', '-out',
>
'/h/data/global/WXQC/data/met/summary/GALWEM_APCP_12hr_50_3_PSTD_0Z',
'-job
> aggregate_stat', '-line_type PCT', '-out_line_type PSTD', '-by
FCST_VAR',
> '-by FCST_THRESH', '-v', '6']
>
>
> Hope the helps. As for the question - your idea for stat_anal
changes
> would be useful - it definitely would make my task easier.
>
> Bob
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, May 26, 2016 3:15 PM
> To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
<robert.craig.2 at us.af.mil>
> Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL <
> matthew.sittel.ctr at us.af.mil>
> Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
>
> Also, I should mention that we do have an idea for future
development in
> STAT-Analysis to would apply to this case, but it doesn't currently
exist.
>
> Basically, pass STAT-Analysis MET output from two different models,
have
> it generate some statistics for each model, and then have it compute
skill
> scores as the performance of one model relative to another.
>
> Would that sort of logic be useful?
>
> John
>
> On Thu, May 26, 2016 at 2:12 PM, John Halley Gotway
<johnhg at ucar.edu>
> wrote:
>
> > Bob,
> >
> > If you're able to send me a sample file along with the stat-
analysis
> > command that make the 3 change to 4, I could take a look and
figure
> > out what's going on.
> >
> > John
> >
> > On Thu, May 26, 2016 at 2:03 PM, robert.craig.2 at us.af.mil via RT <
> > met_help at ucar.edu> wrote:
> >
> >>
> >> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495 >
> >>
> >> John, thanks for all your help on this. I was creating a second
set
> >> of MPR and PCT for the persistence values. But I like your idea
> >> better if it works. One thing with the changes we have made. I
no
> >> longer have the forecast thresholds in the PCT line. They got
> >> replaced by the contingency table threshold values. Also, in the
MPR
> >> lines I had placed 3 for my neighborhood size but it became 4 in
the
> >> PCT file - weird. I might have to keep the different thresholds
> >> separate in separate files - not the preferred option.
> >>
> >> Bob
> >>
> >> -----Original Message-----
> >> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> >> Sent: Thursday, May 26, 2016 2:54 PM
> >> To: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
> >> <robert.craig.2 at us.af.mil>
> >> Cc: SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL <
> >> matthew.sittel.ctr at us.af.mil>
> >> Subject: Re: [rt.rap.ucar.edu #76495] Persistence Data
> >>
> >> Bob,
> >>
> >> I emailed Matt this morning with info on compiling modis_regrid.
> >>
> >> Regarding persistence...
> >>
> >> I see that you'd like to compute BSS in the HiRA framework as the
> >> brier score of the model relative to persistence. Correct me if
I'm
> >> wrong, but you've written code to generate the MPR lines to mimic
the
> >> output of Point-Stat. The second to last column of the MPR line
type
> >> is named "CLIMO". I'd suggest that you fill that column with
that
> persistence value.
> >>
> >> *BUT* this has an implication in the logic of processing this
data
> >> through STAT-Analysis.
> >>
> >> When we run MPR -> PCT, the PCT line type no longer contains
> >> information about persistence. Instead, you'd need to go
directly from
> MPR -> PSTD...
> >> since MPR has climo values, it should use that in the computation
of
> >> BSS for the output PSTD line type.
> >>
> >> However, I haven't string the MET tools together in this order
before.
> >> When you get some sample MPR data which contains persistence data
in
> >> the CLIMO column, I'd love to take a look. Then I could test it
out
> >> with STAT-Analysis to make sure it's working as expected.
> >>
> >> Thanks,
> >> John
> >>
> >> On Wed, May 25, 2016 at 12:45 PM, robert.craig.2 at us.af.mil via RT
<
> >> met_help at ucar.edu> wrote:
> >>
> >> >
> >> > Wed May 25 12:45:15 2016: Request 76495 was acted upon.
> >> > Transaction: Ticket created by robert.craig.2 at us.af.mil
> >> > Queue: met_help
> >> > Subject: Persistence Data
> >> > Owner: Nobody
> >> > Requestors: robert.craig.2 at us.af.mil
> >> > Status: new
> >> > Ticket <URL:
> >> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=76495
> >> > >
> >> >
> >> >
> >> > John or Julie, I need to use persistence data as my reference
> >> > forecast for the HiRA verification I am working on. It will be
> >> > used to calculated Brier Skill Scores. The documentation
mentions
> >> > the use of
> >> climo data for this.
> >> > Is it as simple as providing a persistence forecast for the
climo
> >> > mean filename and field?
> >> >
> >> > Also I want to put a plug in for determining how to get MODIS
> >> > regrid to compile. We have a guy here who wants to use it in
dust
> >> verification.
> >> >
> >> > Thanks
> >> > Bob
> >> >
> >> >
> >> >
> >>
> >>
> >>
> >>
> >
>
>
>
>
------------------------------------------------
More information about the Met_help
mailing list