[Met_help] [rt.rap.ucar.edu #74883] History for rapid intensification filter not working for met-5.1
John Halley Gotway via RT
met_help at ucar.edu
Fri Jan 29 14:51:25 MST 2016
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hi there,
I've been trying to use the "rapid_inten" filter option in MET-TC, but it
doesn't appear to work with version 5.1. It appears that this filter is not
being applied. However, my script works fine with version 5.0.
Thanks,
Andy
--
Andrew Penny
Hurricane Model Diagnostician
SRG / National Hurricane Center
11691 SW 17th Street
Miami, FL 33165-2149
phone: 305.229.4457
email: andrew.penny at noaa.gov
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: John Halley Gotway
Time: Wed Jan 27 14:04:09 2016
Andy,
I have a guess as to what's going on here. In met-5.1, the name of
the
"rapid_inten" option was changed to "rirw" for a variety of reasons.
This is included in the met-5.1 release notes, here:
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
Here's the met-5.0 "rapid_inten" config file option:
rapid_inten = {
track = NONE;
time = "24";
exact = TRUE;
thresh = >=30.0;
}
And here's the met-5.1 "rirw" config file option:
rirw = {
track = NONE;
adeck = {
time = "24";
exact = TRUE;
thresh = >=30.0;
}
bdeck = adeck;
}
Notice here the big change is that we have separate setting for the
ADECK
and BDECK. So you could define different logic for the different
tracks
which should give finer control of the filtering.
Please look on pages 260-261 for more info on this:
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
Hopefully that helps.
Thanks,
John
On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA Affiliate via RT
<
met_help at ucar.edu> wrote:
>
> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> Transaction: Ticket created by andrew.penny at noaa.gov
> Queue: met_help
> Subject: rapid intensification filter not working for met-5.1
> Owner: Nobody
> Requestors: andrew.penny at noaa.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>
>
> Hi there,
>
> I've been trying to use the "rapid_inten" filter option in MET-TC,
but it
> doesn't appear to work with version 5.1. It appears that this filter
is not
> being applied. However, my script works fine with version 5.0.
>
> Thanks,
> Andy
>
> --
> Andrew Penny
> Hurricane Model Diagnostician
> SRG / National Hurricane Center
> 11691 SW 17th Street
> Miami, FL 33165-2149
> phone: 305.229.4457
> email: andrew.penny at noaa.gov
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #74883] rapid intensification filter not working for met-5.1
From: Andrew Penny - NOAA Affiliate
Time: Wed Jan 27 14:59:31 2016
Thanks John! My apologies for missing that.
One other question: is it possible to only verify cases that occur
within a range of intensity change rather than using just one
threshold (e.g., include a case if the intensity changes is between
-10 kt and +10 kt per 24 hours)? I'd like to verify only cases where
intensification/weakening did not occur.
Thanks,
Andy
On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT wrote:
> Andy,
>
> I have a guess as to what's going on here. In met-5.1, the name of
the
> "rapid_inten" option was changed to "rirw" for a variety of reasons.
>
> This is included in the met-5.1 release notes, here:
>
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
>
> Here's the met-5.0 "rapid_inten" config file option:
>
> rapid_inten = {
>
> track = NONE;
>
> time = "24";
>
> exact = TRUE;
>
> thresh = >=30.0;
>
> }
>
> And here's the met-5.1 "rirw" config file option:
>
> rirw = {
>
> track = NONE;
>
> adeck = {
>
> time = "24";
>
> exact = TRUE;
>
> thresh = >=30.0;
>
> }
>
> bdeck = adeck;
>
> }
> Notice here the big change is that we have separate setting for the
ADECK
> and BDECK. So you could define different logic for the different
tracks
> which should give finer control of the filtering.
>
> Please look on pages 260-261 for more info on this:
>
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
>
> Hopefully that helps.
>
> Thanks,
> John
>
> On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA Affiliate via
RT <
> met_help at ucar.edu> wrote:
>
>>
>> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
>> Transaction: Ticket created by andrew.penny at noaa.gov
>> Queue: met_help
>> Subject: rapid intensification filter not working for met-5.1
>> Owner: Nobody
>> Requestors: andrew.penny at noaa.gov
>> Status: new
>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>>
>>
>> Hi there,
>>
>> I've been trying to use the "rapid_inten" filter option in MET-TC,
but it
>> doesn't appear to work with version 5.1. It appears that this
filter is not
>> being applied. However, my script works fine with version 5.0.
>>
>> Thanks,
>> Andy
>>
>> --
>> Andrew Penny
>> Hurricane Model Diagnostician
>> SRG / National Hurricane Center
>> 11691 SW 17th Street
>> Miami, FL 33165-2149
>> phone: 305.229.4457
>> email: andrew.penny at noaa.gov
>>
>>
>
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: John Halley Gotway
Time: Wed Jan 27 15:18:08 2016
Andy,
Yes, you should be able to do that in met-5.1. Read the info about
the
generalized thresholding available in this release:
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
try using this:
thresh = >-10&&<10;
Also look on page 43 and 44 of the MET User's Guide for information
about
defining complex thresholds like this:
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
Please let me know if/how well that works. This is new functionality,
and
I'm curious how smoothly it works.
Thanks,
John
On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA Affiliate via RT
<
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>
> Thanks John! My apologies for missing that.
>
> One other question: is it possible to only verify cases that occur
within
> a range of intensity change rather than using just one threshold
(e.g.,
> include a case if the intensity changes is between -10 kt and +10 kt
per 24
> hours)? I'd like to verify only cases where
intensification/weakening did
> not occur.
>
> Thanks,
> Andy
>
>
> On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT wrote:
>
> > Andy,
> >
> > I have a guess as to what's going on here. In met-5.1, the name
of the
> > "rapid_inten" option was changed to "rirw" for a variety of
reasons.
> >
> > This is included in the met-5.1 release notes, here:
> >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> >
> > Here's the met-5.0 "rapid_inten" config file option:
> >
> > rapid_inten = {
> >
> > track = NONE;
> >
> > time = "24";
> >
> > exact = TRUE;
> >
> > thresh = >=30.0;
> >
> > }
> >
> > And here's the met-5.1 "rirw" config file option:
> >
> > rirw = {
> >
> > track = NONE;
> >
> > adeck = {
> >
> > time = "24";
> >
> > exact = TRUE;
> >
> > thresh = >=30.0;
> >
> > }
> >
> > bdeck = adeck;
> >
> > }
> > Notice here the big change is that we have separate setting for
the ADECK
> > and BDECK. So you could define different logic for the different
tracks
> > which should give finer control of the filtering.
> >
> > Please look on pages 260-261 for more info on this:
> >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> >
> > Hopefully that helps.
> >
> > Thanks,
> > John
> >
> > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA Affiliate
via RT <
> > met_help at ucar.edu> wrote:
> >
> >>
> >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> >> Transaction: Ticket created by andrew.penny at noaa.gov
> >> Queue: met_help
> >> Subject: rapid intensification filter not working for met-5.1
> >> Owner: Nobody
> >> Requestors: andrew.penny at noaa.gov
> >> Status: new
> >> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> >>
> >>
> >> Hi there,
> >>
> >> I've been trying to use the "rapid_inten" filter option in MET-
TC, but
> it
> >> doesn't appear to work with version 5.1. It appears that this
filter is
> not
> >> being applied. However, my script works fine with version 5.0.
> >>
> >> Thanks,
> >> Andy
> >>
> >> --
> >> Andrew Penny
> >> Hurricane Model Diagnostician
> >> SRG / National Hurricane Center
> >> 11691 SW 17th Street
> >> Miami, FL 33165-2149
> >> phone: 305.229.4457
> >> email: andrew.penny at noaa.gov
> >>
> >>
> >
>
>
>
>
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: Andrew Penny - NOAA Affiliate
Time: Thu Jan 28 06:37:28 2016
Hi John,
Thanks for the info. I plan to test combining thresholds out shortly.
I've been experimenting with the rapid intensity threshold filter, and
I'm
getting different results between v5.0 and v5.1. I've done a simple
test
such that I only want to verify cases where the 24 h intensity change
is
more than 15 kt. One caveat is that I'm using the same .tcst file to
test
both versions. Has the format of the .tcst file changed such that this
might be an issue? I've attached the files here for you to see the
differences in the number of cases. Any ideas?
Thanks for your help,
Andy
On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Andy,
>
> Yes, you should be able to do that in met-5.1. Read the info about
the
> generalized thresholding available in this release:
>
>
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
>
> try using this:
> thresh = >-10&&<10;
>
> Also look on page 43 and 44 of the MET User's Guide for information
about
> defining complex thresholds like this:
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
>
> Please let me know if/how well that works. This is new
functionality, and
> I'm curious how smoothly it works.
>
> Thanks,
> John
>
> On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA Affiliate via
RT <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> >
> > Thanks John! My apologies for missing that.
> >
> > One other question: is it possible to only verify cases that occur
within
> > a range of intensity change rather than using just one threshold
(e.g.,
> > include a case if the intensity changes is between -10 kt and +10
kt per
> 24
> > hours)? I'd like to verify only cases where
intensification/weakening did
> > not occur.
> >
> > Thanks,
> > Andy
> >
> >
> > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT wrote:
> >
> > > Andy,
> > >
> > > I have a guess as to what's going on here. In met-5.1, the name
of the
> > > "rapid_inten" option was changed to "rirw" for a variety of
reasons.
> > >
> > > This is included in the met-5.1 release notes, here:
> > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > >
> > > Here's the met-5.0 "rapid_inten" config file option:
> > >
> > > rapid_inten = {
> > >
> > > track = NONE;
> > >
> > > time = "24";
> > >
> > > exact = TRUE;
> > >
> > > thresh = >=30.0;
> > >
> > > }
> > >
> > > And here's the met-5.1 "rirw" config file option:
> > >
> > > rirw = {
> > >
> > > track = NONE;
> > >
> > > adeck = {
> > >
> > > time = "24";
> > >
> > > exact = TRUE;
> > >
> > > thresh = >=30.0;
> > >
> > > }
> > >
> > > bdeck = adeck;
> > >
> > > }
> > > Notice here the big change is that we have separate setting for
the
> ADECK
> > > and BDECK. So you could define different logic for the
different
> tracks
> > > which should give finer control of the filtering.
> > >
> > > Please look on pages 260-261 for more info on this:
> > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > >
> > > Hopefully that helps.
> > >
> > > Thanks,
> > > John
> > >
> > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA Affiliate
via RT
> <
> > > met_help at ucar.edu> wrote:
> > >
> > >>
> > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > >> Queue: met_help
> > >> Subject: rapid intensification filter not working for met-
5.1
> > >> Owner: Nobody
> > >> Requestors: andrew.penny at noaa.gov
> > >> Status: new
> > >> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> >
> > >>
> > >>
> > >> Hi there,
> > >>
> > >> I've been trying to use the "rapid_inten" filter option in MET-
TC, but
> > it
> > >> doesn't appear to work with version 5.1. It appears that this
filter
> is
> > not
> > >> being applied. However, my script works fine with version 5.0.
> > >>
> > >> Thanks,
> > >> Andy
> > >>
> > >> --
> > >> Andrew Penny
> > >> Hurricane Model Diagnostician
> > >> SRG / National Hurricane Center
> > >> 11691 SW 17th Street
> > >> Miami, FL 33165-2149
> > >> phone: 305.229.4457
> > >> email: andrew.penny at noaa.gov
> > >>
> > >>
> > >
> >
> >
> >
> >
>
>
--
Andrew Penny
Hurricane Model Diagnostician
SRG / National Hurricane Center
11691 SW 17th Street
Miami, FL 33165-2149
phone: 305.229.4457
email: andrew.penny at noaa.gov
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: John Halley Gotway
Time: Thu Jan 28 09:57:07 2016
Andrew,
Looking closely at the output you sent, I see that you used different
rapid
intensification thresholds for the different versions of MET:
In met-5.0: -rapid_inten_thresh >=15.000
In met-5.1: -rirw_thresh >=30.0
The higher threshold in 5.1 explains the fewer matching lines. When I
run
the same job using the >=15 through 5.0 and 5.1 on the data file you
sent,
both match the same 135 out of 3101 track lines. That isn't the same
as
141 out of 3545 track lines reported in your met-5.0 output, but
perhaps
you didn't send me the full set of track data files you were using.
Hopefully that explains it. But please let us know if find any other
discrepancies between 5.0 and 5.1.
Thanks,
John
On Thu, Jan 28, 2016 at 6:37 AM, Andrew Penny - NOAA Affiliate via RT
<
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>
> Hi John,
>
> Thanks for the info. I plan to test combining thresholds out
shortly.
>
> I've been experimenting with the rapid intensity threshold filter,
and I'm
> getting different results between v5.0 and v5.1. I've done a simple
test
> such that I only want to verify cases where the 24 h intensity
change is
> more than 15 kt. One caveat is that I'm using the same .tcst file to
test
> both versions. Has the format of the .tcst file changed such that
this
> might be an issue? I've attached the files here for you to see the
> differences in the number of cases. Any ideas?
>
> Thanks for your help,
>
> Andy
>
> On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Andy,
> >
> > Yes, you should be able to do that in met-5.1. Read the info
about the
> > generalized thresholding available in this release:
> >
> >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> >
> > try using this:
> > thresh = >-10&&<10;
> >
> > Also look on page 43 and 44 of the MET User's Guide for
information about
> > defining complex thresholds like this:
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> >
> > Please let me know if/how well that works. This is new
functionality,
> and
> > I'm curious how smoothly it works.
> >
> > Thanks,
> > John
> >
> > On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA Affiliate via
RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > >
> > > Thanks John! My apologies for missing that.
> > >
> > > One other question: is it possible to only verify cases that
occur
> within
> > > a range of intensity change rather than using just one threshold
(e.g.,
> > > include a case if the intensity changes is between -10 kt and
+10 kt
> per
> > 24
> > > hours)? I'd like to verify only cases where
intensification/weakening
> did
> > > not occur.
> > >
> > > Thanks,
> > > Andy
> > >
> > >
> > > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT wrote:
> > >
> > > > Andy,
> > > >
> > > > I have a guess as to what's going on here. In met-5.1, the
name of
> the
> > > > "rapid_inten" option was changed to "rirw" for a variety of
reasons.
> > > >
> > > > This is included in the met-5.1 release notes, here:
> > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > >
> > > > Here's the met-5.0 "rapid_inten" config file option:
> > > >
> > > > rapid_inten = {
> > > >
> > > > track = NONE;
> > > >
> > > > time = "24";
> > > >
> > > > exact = TRUE;
> > > >
> > > > thresh = >=30.0;
> > > >
> > > > }
> > > >
> > > > And here's the met-5.1 "rirw" config file option:
> > > >
> > > > rirw = {
> > > >
> > > > track = NONE;
> > > >
> > > > adeck = {
> > > >
> > > > time = "24";
> > > >
> > > > exact = TRUE;
> > > >
> > > > thresh = >=30.0;
> > > >
> > > > }
> > > >
> > > > bdeck = adeck;
> > > >
> > > > }
> > > > Notice here the big change is that we have separate setting
for the
> > ADECK
> > > > and BDECK. So you could define different logic for the
different
> > tracks
> > > > which should give finer control of the filtering.
> > > >
> > > > Please look on pages 260-261 for more info on this:
> > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > >
> > > > Hopefully that helps.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA
Affiliate via
> RT
> > <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > >>
> > > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> > > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > > >> Queue: met_help
> > > >> Subject: rapid intensification filter not working for
met-5.1
> > > >> Owner: Nobody
> > > >> Requestors: andrew.penny at noaa.gov
> > > >> Status: new
> > > >> Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> > >
> > > >>
> > > >>
> > > >> Hi there,
> > > >>
> > > >> I've been trying to use the "rapid_inten" filter option in
MET-TC,
> but
> > > it
> > > >> doesn't appear to work with version 5.1. It appears that this
filter
> > is
> > > not
> > > >> being applied. However, my script works fine with version
5.0.
> > > >>
> > > >> Thanks,
> > > >> Andy
> > > >>
> > > >> --
> > > >> Andrew Penny
> > > >> Hurricane Model Diagnostician
> > > >> SRG / National Hurricane Center
> > > >> 11691 SW 17th Street
> > > >> Miami, FL 33165-2149
> > > >> phone: 305.229.4457
> > > >> email: andrew.penny at noaa.gov
> > > >>
> > > >>
> > > >
> > >
> > >
> > >
> > >
> >
> >
>
>
> --
> Andrew Penny
> Hurricane Model Diagnostician
> SRG / National Hurricane Center
> 11691 SW 17th Street
> Miami, FL 33165-2149
> phone: 305.229.4457
> email: andrew.penny at noaa.gov
>
>
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: Andrew Penny - NOAA Affiliate
Time: Thu Jan 28 10:02:45 2016
Hi John,
If you look at the configure file I sent "TCStatConfig_INT_ERR_POS15",
both
thresholds should be set to >=15 kt. I'm not sure why v5.1 is using
>=30
kt. Is something wrong with the syntax that I'm using (see below)?
rapid_inten = {
track = BDECK;
time = "24";
exact = TRUE;
thresh = >=15.0;
}
rirw = {
track = BDECK;
time = "24";
exact = TRUE;
thresh = >=15.0;
}
Thanks,
Andy
On Thu, Jan 28, 2016 at 4:57 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Andrew,
>
> Looking closely at the output you sent, I see that you used
different rapid
> intensification thresholds for the different versions of MET:
>
> In met-5.0: -rapid_inten_thresh >=15.000
> In met-5.1: -rirw_thresh >=30.0
>
> The higher threshold in 5.1 explains the fewer matching lines. When
I run
> the same job using the >=15 through 5.0 and 5.1 on the data file you
sent,
> both match the same 135 out of 3101 track lines. That isn't the
same as
> 141 out of 3545 track lines reported in your met-5.0 output, but
perhaps
> you didn't send me the full set of track data files you were using.
>
> Hopefully that explains it. But please let us know if find any
other
> discrepancies between 5.0 and 5.1.
>
> Thanks,
> John
>
> On Thu, Jan 28, 2016 at 6:37 AM, Andrew Penny - NOAA Affiliate via
RT <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> >
> > Hi John,
> >
> > Thanks for the info. I plan to test combining thresholds out
shortly.
> >
> > I've been experimenting with the rapid intensity threshold filter,
and
> I'm
> > getting different results between v5.0 and v5.1. I've done a
simple test
> > such that I only want to verify cases where the 24 h intensity
change is
> > more than 15 kt. One caveat is that I'm using the same .tcst file
to test
> > both versions. Has the format of the .tcst file changed such that
this
> > might be an issue? I've attached the files here for you to see the
> > differences in the number of cases. Any ideas?
> >
> > Thanks for your help,
> >
> > Andy
> >
> > On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Andy,
> > >
> > > Yes, you should be able to do that in met-5.1. Read the info
about the
> > > generalized thresholding available in this release:
> > >
> > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > >
> > > try using this:
> > > thresh = >-10&&<10;
> > >
> > > Also look on page 43 and 44 of the MET User's Guide for
information
> about
> > > defining complex thresholds like this:
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > >
> > > Please let me know if/how well that works. This is new
functionality,
> > and
> > > I'm curious how smoothly it works.
> > >
> > > Thanks,
> > > John
> > >
> > > On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA Affiliate
via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
>
> > > >
> > > > Thanks John! My apologies for missing that.
> > > >
> > > > One other question: is it possible to only verify cases that
occur
> > within
> > > > a range of intensity change rather than using just one
threshold
> (e.g.,
> > > > include a case if the intensity changes is between -10 kt and
+10 kt
> > per
> > > 24
> > > > hours)? I'd like to verify only cases where
intensification/weakening
> > did
> > > > not occur.
> > > >
> > > > Thanks,
> > > > Andy
> > > >
> > > >
> > > > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT wrote:
> > > >
> > > > > Andy,
> > > > >
> > > > > I have a guess as to what's going on here. In met-5.1, the
name of
> > the
> > > > > "rapid_inten" option was changed to "rirw" for a variety of
> reasons.
> > > > >
> > > > > This is included in the met-5.1 release notes, here:
> > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > >
> > > > > Here's the met-5.0 "rapid_inten" config file option:
> > > > >
> > > > > rapid_inten = {
> > > > >
> > > > > track = NONE;
> > > > >
> > > > > time = "24";
> > > > >
> > > > > exact = TRUE;
> > > > >
> > > > > thresh = >=30.0;
> > > > >
> > > > > }
> > > > >
> > > > > And here's the met-5.1 "rirw" config file option:
> > > > >
> > > > > rirw = {
> > > > >
> > > > > track = NONE;
> > > > >
> > > > > adeck = {
> > > > >
> > > > > time = "24";
> > > > >
> > > > > exact = TRUE;
> > > > >
> > > > > thresh = >=30.0;
> > > > >
> > > > > }
> > > > >
> > > > > bdeck = adeck;
> > > > >
> > > > > }
> > > > > Notice here the big change is that we have separate setting
for the
> > > ADECK
> > > > > and BDECK. So you could define different logic for the
different
> > > tracks
> > > > > which should give finer control of the filtering.
> > > > >
> > > > > Please look on pages 260-261 for more info on this:
> > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > >
> > > > > Hopefully that helps.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA
Affiliate via
> > RT
> > > <
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > >>
> > > > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> > > > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > > > >> Queue: met_help
> > > > >> Subject: rapid intensification filter not working for
met-5.1
> > > > >> Owner: Nobody
> > > > >> Requestors: andrew.penny at noaa.gov
> > > > >> Status: new
> > > > >> Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> > > >
> > > > >>
> > > > >>
> > > > >> Hi there,
> > > > >>
> > > > >> I've been trying to use the "rapid_inten" filter option in
MET-TC,
> > but
> > > > it
> > > > >> doesn't appear to work with version 5.1. It appears that
this
> filter
> > > is
> > > > not
> > > > >> being applied. However, my script works fine with version
5.0.
> > > > >>
> > > > >> Thanks,
> > > > >> Andy
> > > > >>
> > > > >> --
> > > > >> Andrew Penny
> > > > >> Hurricane Model Diagnostician
> > > > >> SRG / National Hurricane Center
> > > > >> 11691 SW 17th Street
> > > > >> Miami, FL 33165-2149
> > > > >> phone: 305.229.4457
> > > > >> email: andrew.penny at noaa.gov
> > > > >>
> > > > >>
> > > > >
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Andrew Penny
> > Hurricane Model Diagnostician
> > SRG / National Hurricane Center
> > 11691 SW 17th Street
> > Miami, FL 33165-2149
> > phone: 305.229.4457
> > email: andrew.penny at noaa.gov
> >
> >
>
>
--
Andrew Penny
Hurricane Model Diagnostician
SRG / National Hurricane Center
11691 SW 17th Street
Miami, FL 33165-2149
phone: 305.229.4457
email: andrew.penny at noaa.gov
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: John Halley Gotway
Time: Thu Jan 28 10:35:50 2016
Andrew,
Look closely at that section of the config file for met-5.1. It's
structured slightly differently than 5.0. Try using this instead:
rirw = {
track = BDECK;
bdeck = {
time = "24";
exact = TRUE;
thresh = >=15.0;
}
adeck = bdeck;
}
Your met-5.1 didn't follow this format, so it reverted to the default
values.
Within "rirw" the "adeck" and "bdeck" dictionaries define the RI/RW
logic
separately. For example, you may be to filter on RI/RW of >=30.0 in
the
BDECK and >=15.0 in the ADECK if your model has a low intensity bias.
In
met-5.0, you had to use the same definition for both. In met-5.1, you
can
define them separately.
In the example above, I defined the bdeck criteria and then just
applied
the same to the adeck by saying "adeck = bdeck;" However it doesn't
really
matter because we're only applying the filtering logic to the BDECK
since
"track = BDECK;".
These changes came as a result of adding the new "rirw" job type in
met-5.1. This job uses the "rirw" criteria to define *events* in both
the
ADECK and BDECK. And it uses those events to populate a 2x2
contingency
table, from which statistics can be derived.
The really nice thing about it is that you can define a matching time
window. So, if your forecast did have RI but was, let's say, 12 hours
late
you could still count that as a "hit" in the contingency table. Using
the
sample data you sent me, here's how you could run this type of job:
met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC
By default, it requires an exact match in time. Try allowing it to be
+/-
12 hours:
met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC -rirw_window 12
That causes 2 of the misses (FN_OY) to turn into hits (FY_OY).
Lastly, try turning on all possible output types (CTC, CTS, and MPR)
and
redirect the output to a file:
met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC,CTS,MPR
-rirw_window
12 -out rirw_example.out
The resulting output file is attached.
Thanks,
John
On Thu, Jan 28, 2016 at 10:02 AM, Andrew Penny - NOAA Affiliate via RT
<
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>
> Hi John,
>
> If you look at the configure file I sent
"TCStatConfig_INT_ERR_POS15", both
> thresholds should be set to >=15 kt. I'm not sure why v5.1 is using
>=30
> kt. Is something wrong with the syntax that I'm using (see below)?
>
> rapid_inten = {
> track = BDECK;
> time = "24";
> exact = TRUE;
> thresh = >=15.0;
> }
>
> rirw = {
> track = BDECK;
> time = "24";
> exact = TRUE;
> thresh = >=15.0;
> }
>
> Thanks,
> Andy
>
>
> On Thu, Jan 28, 2016 at 4:57 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Andrew,
> >
> > Looking closely at the output you sent, I see that you used
different
> rapid
> > intensification thresholds for the different versions of MET:
> >
> > In met-5.0: -rapid_inten_thresh >=15.000
> > In met-5.1: -rirw_thresh >=30.0
> >
> > The higher threshold in 5.1 explains the fewer matching lines.
When I
> run
> > the same job using the >=15 through 5.0 and 5.1 on the data file
you
> sent,
> > both match the same 135 out of 3101 track lines. That isn't the
same as
> > 141 out of 3545 track lines reported in your met-5.0 output, but
perhaps
> > you didn't send me the full set of track data files you were
using.
> >
> > Hopefully that explains it. But please let us know if find any
other
> > discrepancies between 5.0 and 5.1.
> >
> > Thanks,
> > John
> >
> > On Thu, Jan 28, 2016 at 6:37 AM, Andrew Penny - NOAA Affiliate via
RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > >
> > > Hi John,
> > >
> > > Thanks for the info. I plan to test combining thresholds out
shortly.
> > >
> > > I've been experimenting with the rapid intensity threshold
filter, and
> > I'm
> > > getting different results between v5.0 and v5.1. I've done a
simple
> test
> > > such that I only want to verify cases where the 24 h intensity
change
> is
> > > more than 15 kt. One caveat is that I'm using the same .tcst
file to
> test
> > > both versions. Has the format of the .tcst file changed such
that this
> > > might be an issue? I've attached the files here for you to see
the
> > > differences in the number of cases. Any ideas?
> > >
> > > Thanks for your help,
> > >
> > > Andy
> > >
> > > On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Andy,
> > > >
> > > > Yes, you should be able to do that in met-5.1. Read the info
about
> the
> > > > generalized thresholding available in this release:
> > > >
> > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > >
> > > > try using this:
> > > > thresh = >-10&&<10;
> > > >
> > > > Also look on page 43 and 44 of the MET User's Guide for
information
> > about
> > > > defining complex thresholds like this:
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > >
> > > > Please let me know if/how well that works. This is new
> functionality,
> > > and
> > > > I'm curious how smoothly it works.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA Affiliate
via
> RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > > > >
> > > > > Thanks John! My apologies for missing that.
> > > > >
> > > > > One other question: is it possible to only verify cases that
occur
> > > within
> > > > > a range of intensity change rather than using just one
threshold
> > (e.g.,
> > > > > include a case if the intensity changes is between -10 kt
and +10
> kt
> > > per
> > > > 24
> > > > > hours)? I'd like to verify only cases where
> intensification/weakening
> > > did
> > > > > not occur.
> > > > >
> > > > > Thanks,
> > > > > Andy
> > > > >
> > > > >
> > > > > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT
wrote:
> > > > >
> > > > > > Andy,
> > > > > >
> > > > > > I have a guess as to what's going on here. In met-5.1,
the name
> of
> > > the
> > > > > > "rapid_inten" option was changed to "rirw" for a variety
of
> > reasons.
> > > > > >
> > > > > > This is included in the met-5.1 release notes, here:
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > > >
> > > > > > Here's the met-5.0 "rapid_inten" config file option:
> > > > > >
> > > > > > rapid_inten = {
> > > > > >
> > > > > > track = NONE;
> > > > > >
> > > > > > time = "24";
> > > > > >
> > > > > > exact = TRUE;
> > > > > >
> > > > > > thresh = >=30.0;
> > > > > >
> > > > > > }
> > > > > >
> > > > > > And here's the met-5.1 "rirw" config file option:
> > > > > >
> > > > > > rirw = {
> > > > > >
> > > > > > track = NONE;
> > > > > >
> > > > > > adeck = {
> > > > > >
> > > > > > time = "24";
> > > > > >
> > > > > > exact = TRUE;
> > > > > >
> > > > > > thresh = >=30.0;
> > > > > >
> > > > > > }
> > > > > >
> > > > > > bdeck = adeck;
> > > > > >
> > > > > > }
> > > > > > Notice here the big change is that we have separate
setting for
> the
> > > > ADECK
> > > > > > and BDECK. So you could define different logic for the
different
> > > > tracks
> > > > > > which should give finer control of the filtering.
> > > > > >
> > > > > > Please look on pages 260-261 for more info on this:
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > > >
> > > > > > Hopefully that helps.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA
Affiliate
> via
> > > RT
> > > > <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > >>
> > > > > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> > > > > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > > > > >> Queue: met_help
> > > > > >> Subject: rapid intensification filter not working for
> met-5.1
> > > > > >> Owner: Nobody
> > > > > >> Requestors: andrew.penny at noaa.gov
> > > > > >> Status: new
> > > > > >> Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> > > > >
> > > > > >>
> > > > > >>
> > > > > >> Hi there,
> > > > > >>
> > > > > >> I've been trying to use the "rapid_inten" filter option
in
> MET-TC,
> > > but
> > > > > it
> > > > > >> doesn't appear to work with version 5.1. It appears that
this
> > filter
> > > > is
> > > > > not
> > > > > >> being applied. However, my script works fine with version
5.0.
> > > > > >>
> > > > > >> Thanks,
> > > > > >> Andy
> > > > > >>
> > > > > >> --
> > > > > >> Andrew Penny
> > > > > >> Hurricane Model Diagnostician
> > > > > >> SRG / National Hurricane Center
> > > > > >> 11691 SW 17th Street
> > > > > >> Miami, FL 33165-2149
> > > > > >> phone: 305.229.4457
> > > > > >> email: andrew.penny at noaa.gov
> > > > > >>
> > > > > >>
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Andrew Penny
> > > Hurricane Model Diagnostician
> > > SRG / National Hurricane Center
> > > 11691 SW 17th Street
> > > Miami, FL 33165-2149
> > > phone: 305.229.4457
> > > email: andrew.penny at noaa.gov
> > >
> > >
> >
> >
>
>
> --
> Andrew Penny
> Hurricane Model Diagnostician
> SRG / National Hurricane Center
> 11691 SW 17th Street
> Miami, FL 33165-2149
> phone: 305.229.4457
> email: andrew.penny at noaa.gov
>
>
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: Andrew Penny - NOAA Affiliate
Time: Thu Jan 28 11:24:32 2016
Hi John,
Thanks so much for the explanation. I can definitely see the
usefulness of
having the capability to do something like that. I'll have to
experiment
with some of these features when I have more time.
I've corrected the syntax to match what is expected for v5.1, and all
is
working well now. I'm able to reproduce the results I obtained with
v5.0.
Thanks for your help,
Andy
On Thu, Jan 28, 2016 at 5:35 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Andrew,
>
> Look closely at that section of the config file for met-5.1. It's
> structured slightly differently than 5.0. Try using this instead:
>
> rirw = {
> track = BDECK;
> bdeck = {
> time = "24";
> exact = TRUE;
> thresh = >=15.0;
> }
> adeck = bdeck;
> }
>
> Your met-5.1 didn't follow this format, so it reverted to the
default
> values.
>
> Within "rirw" the "adeck" and "bdeck" dictionaries define the RI/RW
logic
> separately. For example, you may be to filter on RI/RW of >=30.0 in
the
> BDECK and >=15.0 in the ADECK if your model has a low intensity
bias. In
> met-5.0, you had to use the same definition for both. In met-5.1,
you can
> define them separately.
>
> In the example above, I defined the bdeck criteria and then just
applied
> the same to the adeck by saying "adeck = bdeck;" However it doesn't
really
> matter because we're only applying the filtering logic to the BDECK
since
> "track = BDECK;".
>
> These changes came as a result of adding the new "rirw" job type in
> met-5.1. This job uses the "rirw" criteria to define *events* in
both the
> ADECK and BDECK. And it uses those events to populate a 2x2
contingency
> table, from which statistics can be derived.
>
> The really nice thing about it is that you can define a matching
time
> window. So, if your forecast did have RI but was, let's say, 12
hours late
> you could still count that as a "hit" in the contingency table.
Using the
> sample data you sent me, here's how you could run this type of job:
>
> met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
> rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC
>
> By default, it requires an exact match in time. Try allowing it to
be +/-
> 12 hours:
>
> met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
> rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC -rirw_window
12
>
> That causes 2 of the misses (FN_OY) to turn into hits (FY_OY).
>
> Lastly, try turning on all possible output types (CTC, CTS, and MPR)
and
> redirect the output to a file:
>
> met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel HB48,HCTC,HIVB
-job
> rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC,CTS,MPR
-rirw_window
> 12 -out rirw_example.out
>
> The resulting output file is attached.
>
> Thanks,
> John
>
>
> On Thu, Jan 28, 2016 at 10:02 AM, Andrew Penny - NOAA Affiliate via
RT <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> >
> > Hi John,
> >
> > If you look at the configure file I sent
"TCStatConfig_INT_ERR_POS15",
> both
> > thresholds should be set to >=15 kt. I'm not sure why v5.1 is
using >=30
> > kt. Is something wrong with the syntax that I'm using (see below)?
> >
> > rapid_inten = {
> > track = BDECK;
> > time = "24";
> > exact = TRUE;
> > thresh = >=15.0;
> > }
> >
> > rirw = {
> > track = BDECK;
> > time = "24";
> > exact = TRUE;
> > thresh = >=15.0;
> > }
> >
> > Thanks,
> > Andy
> >
> >
> > On Thu, Jan 28, 2016 at 4:57 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Andrew,
> > >
> > > Looking closely at the output you sent, I see that you used
different
> > rapid
> > > intensification thresholds for the different versions of MET:
> > >
> > > In met-5.0: -rapid_inten_thresh >=15.000
> > > In met-5.1: -rirw_thresh >=30.0
> > >
> > > The higher threshold in 5.1 explains the fewer matching lines.
When I
> > run
> > > the same job using the >=15 through 5.0 and 5.1 on the data file
you
> > sent,
> > > both match the same 135 out of 3101 track lines. That isn't the
same
> as
> > > 141 out of 3545 track lines reported in your met-5.0 output, but
> perhaps
> > > you didn't send me the full set of track data files you were
using.
> > >
> > > Hopefully that explains it. But please let us know if find any
other
> > > discrepancies between 5.0 and 5.1.
> > >
> > > Thanks,
> > > John
> > >
> > > On Thu, Jan 28, 2016 at 6:37 AM, Andrew Penny - NOAA Affiliate
via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
>
> > > >
> > > > Hi John,
> > > >
> > > > Thanks for the info. I plan to test combining thresholds out
shortly.
> > > >
> > > > I've been experimenting with the rapid intensity threshold
filter,
> and
> > > I'm
> > > > getting different results between v5.0 and v5.1. I've done a
simple
> > test
> > > > such that I only want to verify cases where the 24 h intensity
change
> > is
> > > > more than 15 kt. One caveat is that I'm using the same .tcst
file to
> > test
> > > > both versions. Has the format of the .tcst file changed such
that
> this
> > > > might be an issue? I've attached the files here for you to see
the
> > > > differences in the number of cases. Any ideas?
> > > >
> > > > Thanks for your help,
> > > >
> > > > Andy
> > > >
> > > > On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > > Andy,
> > > > >
> > > > > Yes, you should be able to do that in met-5.1. Read the
info about
> > the
> > > > > generalized thresholding available in this release:
> > > > >
> > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > >
> > > > > try using this:
> > > > > thresh = >-10&&<10;
> > > > >
> > > > > Also look on page 43 and 44 of the MET User's Guide for
information
> > > about
> > > > > defining complex thresholds like this:
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > >
> > > > > Please let me know if/how well that works. This is new
> > functionality,
> > > > and
> > > > > I'm curious how smoothly it works.
> > > > >
> > > > > Thanks,
> > > > > John
> > > > >
> > > > > On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA
Affiliate via
> > RT <
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > > > > >
> > > > > > Thanks John! My apologies for missing that.
> > > > > >
> > > > > > One other question: is it possible to only verify cases
that
> occur
> > > > within
> > > > > > a range of intensity change rather than using just one
threshold
> > > (e.g.,
> > > > > > include a case if the intensity changes is between -10 kt
and +10
> > kt
> > > > per
> > > > > 24
> > > > > > hours)? I'd like to verify only cases where
> > intensification/weakening
> > > > did
> > > > > > not occur.
> > > > > >
> > > > > > Thanks,
> > > > > > Andy
> > > > > >
> > > > > >
> > > > > > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT
wrote:
> > > > > >
> > > > > > > Andy,
> > > > > > >
> > > > > > > I have a guess as to what's going on here. In met-5.1,
the
> name
> > of
> > > > the
> > > > > > > "rapid_inten" option was changed to "rirw" for a variety
of
> > > reasons.
> > > > > > >
> > > > > > > This is included in the met-5.1 release notes, here:
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > > > >
> > > > > > > Here's the met-5.0 "rapid_inten" config file option:
> > > > > > >
> > > > > > > rapid_inten = {
> > > > > > >
> > > > > > > track = NONE;
> > > > > > >
> > > > > > > time = "24";
> > > > > > >
> > > > > > > exact = TRUE;
> > > > > > >
> > > > > > > thresh = >=30.0;
> > > > > > >
> > > > > > > }
> > > > > > >
> > > > > > > And here's the met-5.1 "rirw" config file option:
> > > > > > >
> > > > > > > rirw = {
> > > > > > >
> > > > > > > track = NONE;
> > > > > > >
> > > > > > > adeck = {
> > > > > > >
> > > > > > > time = "24";
> > > > > > >
> > > > > > > exact = TRUE;
> > > > > > >
> > > > > > > thresh = >=30.0;
> > > > > > >
> > > > > > > }
> > > > > > >
> > > > > > > bdeck = adeck;
> > > > > > >
> > > > > > > }
> > > > > > > Notice here the big change is that we have separate
setting for
> > the
> > > > > ADECK
> > > > > > > and BDECK. So you could define different logic for the
> different
> > > > > tracks
> > > > > > > which should give finer control of the filtering.
> > > > > > >
> > > > > > > Please look on pages 260-261 for more info on this:
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > > > >
> > > > > > > Hopefully that helps.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > John
> > > > > > >
> > > > > > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA
Affiliate
> > via
> > > > RT
> > > > > <
> > > > > > > met_help at ucar.edu> wrote:
> > > > > > >
> > > > > > >>
> > > > > > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted upon.
> > > > > > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > > > > > >> Queue: met_help
> > > > > > >> Subject: rapid intensification filter not working
for
> > met-5.1
> > > > > > >> Owner: Nobody
> > > > > > >> Requestors: andrew.penny at noaa.gov
> > > > > > >> Status: new
> > > > > > >> Ticket <URL:
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> > > > > >
> > > > > > >>
> > > > > > >>
> > > > > > >> Hi there,
> > > > > > >>
> > > > > > >> I've been trying to use the "rapid_inten" filter option
in
> > MET-TC,
> > > > but
> > > > > > it
> > > > > > >> doesn't appear to work with version 5.1. It appears
that this
> > > filter
> > > > > is
> > > > > > not
> > > > > > >> being applied. However, my script works fine with
version 5.0.
> > > > > > >>
> > > > > > >> Thanks,
> > > > > > >> Andy
> > > > > > >>
> > > > > > >> --
> > > > > > >> Andrew Penny
> > > > > > >> Hurricane Model Diagnostician
> > > > > > >> SRG / National Hurricane Center
> > > > > > >> 11691 SW 17th Street
> > > > > > >> Miami, FL 33165-2149
> > > > > > >> phone: 305.229.4457
> > > > > > >> email: andrew.penny at noaa.gov
> > > > > > >>
> > > > > > >>
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Andrew Penny
> > > > Hurricane Model Diagnostician
> > > > SRG / National Hurricane Center
> > > > 11691 SW 17th Street
> > > > Miami, FL 33165-2149
> > > > phone: 305.229.4457
> > > > email: andrew.penny at noaa.gov
> > > >
> > > >
> > >
> > >
> >
> >
> > --
> > Andrew Penny
> > Hurricane Model Diagnostician
> > SRG / National Hurricane Center
> > 11691 SW 17th Street
> > Miami, FL 33165-2149
> > phone: 305.229.4457
> > email: andrew.penny at noaa.gov
> >
> >
>
>
--
Andrew Penny
Hurricane Model Diagnostician
SRG / National Hurricane Center
11691 SW 17th Street
Miami, FL 33165-2149
phone: 305.229.4457
email: andrew.penny at noaa.gov
------------------------------------------------
Subject: rapid intensification filter not working for met-5.1
From: John Halley Gotway
Time: Thu Jan 28 11:38:53 2016
Andrew,
Great, I'll go ahead and resolve this ticket.
Thanks,
John
On Thu, Jan 28, 2016 at 11:24 AM, Andrew Penny - NOAA Affiliate via RT
<
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
>
> Hi John,
>
> Thanks so much for the explanation. I can definitely see the
usefulness of
> having the capability to do something like that. I'll have to
experiment
> with some of these features when I have more time.
>
> I've corrected the syntax to match what is expected for v5.1, and
all is
> working well now. I'm able to reproduce the results I obtained with
v5.0.
>
> Thanks for your help,
> Andy
>
>
> On Thu, Jan 28, 2016 at 5:35 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Andrew,
> >
> > Look closely at that section of the config file for met-5.1. It's
> > structured slightly differently than 5.0. Try using this instead:
> >
> > rirw = {
> > track = BDECK;
> > bdeck = {
> > time = "24";
> > exact = TRUE;
> > thresh = >=15.0;
> > }
> > adeck = bdeck;
> > }
> >
> > Your met-5.1 didn't follow this format, so it reverted to the
default
> > values.
> >
> > Within "rirw" the "adeck" and "bdeck" dictionaries define the
RI/RW logic
> > separately. For example, you may be to filter on RI/RW of >=30.0
in the
> > BDECK and >=15.0 in the ADECK if your model has a low intensity
bias. In
> > met-5.0, you had to use the same definition for both. In met-5.1,
you
> can
> > define them separately.
> >
> > In the example above, I defined the bdeck criteria and then just
applied
> > the same to the adeck by saying "adeck = bdeck;" However it
doesn't
> really
> > matter because we're only applying the filtering logic to the
BDECK since
> > "track = BDECK;".
> >
> > These changes came as a result of adding the new "rirw" job type
in
> > met-5.1. This job uses the "rirw" criteria to define *events* in
both
> the
> > ADECK and BDECK. And it uses those events to populate a 2x2
contingency
> > table, from which statistics can be derived.
> >
> > The really nice thing about it is that you can define a matching
time
> > window. So, if your forecast did have RI but was, let's say, 12
hours
> late
> > you could still count that as a "hit" in the contingency table.
Using
> the
> > sample data you sent me, here's how you could run this type of
job:
> >
> > met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel
HB48,HCTC,HIVB -job
> > rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC
> >
> > By default, it requires an exact match in time. Try allowing it
to be
> +/-
> > 12 hours:
> >
> > met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel
HB48,HCTC,HIVB -job
> > rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC -rirw_window
12
> >
> > That causes 2 of the misses (FN_OY) to turn into hits (FY_OY).
> >
> > Lastly, try turning on all possible output types (CTC, CTS, and
MPR) and
> > redirect the output to a file:
> >
> > met-5.1/bin/tc_stat -lookin ep222015.tcst -amodel
HB48,HCTC,HIVB -job
> > rirw -by AMODEL -rirw_thresh \>=15 -out_line_type CTC,CTS,MPR
> -rirw_window
> > 12 -out rirw_example.out
> >
> > The resulting output file is attached.
> >
> > Thanks,
> > John
> >
> >
> > On Thu, Jan 28, 2016 at 10:02 AM, Andrew Penny - NOAA Affiliate
via RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > >
> > > Hi John,
> > >
> > > If you look at the configure file I sent
"TCStatConfig_INT_ERR_POS15",
> > both
> > > thresholds should be set to >=15 kt. I'm not sure why v5.1 is
using
> >=30
> > > kt. Is something wrong with the syntax that I'm using (see
below)?
> > >
> > > rapid_inten = {
> > > track = BDECK;
> > > time = "24";
> > > exact = TRUE;
> > > thresh = >=15.0;
> > > }
> > >
> > > rirw = {
> > > track = BDECK;
> > > time = "24";
> > > exact = TRUE;
> > > thresh = >=15.0;
> > > }
> > >
> > > Thanks,
> > > Andy
> > >
> > >
> > > On Thu, Jan 28, 2016 at 4:57 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Andrew,
> > > >
> > > > Looking closely at the output you sent, I see that you used
different
> > > rapid
> > > > intensification thresholds for the different versions of MET:
> > > >
> > > > In met-5.0: -rapid_inten_thresh >=15.000
> > > > In met-5.1: -rirw_thresh >=30.0
> > > >
> > > > The higher threshold in 5.1 explains the fewer matching lines.
When
> I
> > > run
> > > > the same job using the >=15 through 5.0 and 5.1 on the data
file you
> > > sent,
> > > > both match the same 135 out of 3101 track lines. That isn't
the same
> > as
> > > > 141 out of 3545 track lines reported in your met-5.0 output,
but
> > perhaps
> > > > you didn't send me the full set of track data files you were
using.
> > > >
> > > > Hopefully that explains it. But please let us know if find
any other
> > > > discrepancies between 5.0 and 5.1.
> > > >
> > > > Thanks,
> > > > John
> > > >
> > > > On Thu, Jan 28, 2016 at 6:37 AM, Andrew Penny - NOAA Affiliate
via
> RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883 >
> > > > >
> > > > > Hi John,
> > > > >
> > > > > Thanks for the info. I plan to test combining thresholds out
> shortly.
> > > > >
> > > > > I've been experimenting with the rapid intensity threshold
filter,
> > and
> > > > I'm
> > > > > getting different results between v5.0 and v5.1. I've done a
simple
> > > test
> > > > > such that I only want to verify cases where the 24 h
intensity
> change
> > > is
> > > > > more than 15 kt. One caveat is that I'm using the same .tcst
file
> to
> > > test
> > > > > both versions. Has the format of the .tcst file changed such
that
> > this
> > > > > might be an issue? I've attached the files here for you to
see the
> > > > > differences in the number of cases. Any ideas?
> > > > >
> > > > > Thanks for your help,
> > > > >
> > > > > Andy
> > > > >
> > > > > On Wed, Jan 27, 2016 at 10:18 PM, John Halley Gotway via RT
<
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > > Andy,
> > > > > >
> > > > > > Yes, you should be able to do that in met-5.1. Read the
info
> about
> > > the
> > > > > > generalized thresholding available in this release:
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > > >
> > > > > > try using this:
> > > > > > thresh = >-10&&<10;
> > > > > >
> > > > > > Also look on page 43 and 44 of the MET User's Guide for
> information
> > > > about
> > > > > > defining complex thresholds like this:
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > > >
> > > > > > Please let me know if/how well that works. This is new
> > > functionality,
> > > > > and
> > > > > > I'm curious how smoothly it works.
> > > > > >
> > > > > > Thanks,
> > > > > > John
> > > > > >
> > > > > > On Wed, Jan 27, 2016 at 2:59 PM, Andrew Penny - NOAA
Affiliate
> via
> > > RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> >
> > > > > > >
> > > > > > > Thanks John! My apologies for missing that.
> > > > > > >
> > > > > > > One other question: is it possible to only verify cases
that
> > occur
> > > > > within
> > > > > > > a range of intensity change rather than using just one
> threshold
> > > > (e.g.,
> > > > > > > include a case if the intensity changes is between -10
kt and
> +10
> > > kt
> > > > > per
> > > > > > 24
> > > > > > > hours)? I'd like to verify only cases where
> > > intensification/weakening
> > > > > did
> > > > > > > not occur.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > Andy
> > > > > > >
> > > > > > >
> > > > > > > On Jan 27, 2016, at 4:04 PM, John Halley Gotway via RT
wrote:
> > > > > > >
> > > > > > > > Andy,
> > > > > > > >
> > > > > > > > I have a guess as to what's going on here. In met-
5.1, the
> > name
> > > of
> > > > > the
> > > > > > > > "rapid_inten" option was changed to "rirw" for a
variety of
> > > > reasons.
> > > > > > > >
> > > > > > > > This is included in the met-5.1 release notes, here:
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/support/release_notes/METv5.1_release_notes.php
> > > > > > > >
> > > > > > > > Here's the met-5.0 "rapid_inten" config file option:
> > > > > > > >
> > > > > > > > rapid_inten = {
> > > > > > > >
> > > > > > > > track = NONE;
> > > > > > > >
> > > > > > > > time = "24";
> > > > > > > >
> > > > > > > > exact = TRUE;
> > > > > > > >
> > > > > > > > thresh = >=30.0;
> > > > > > > >
> > > > > > > > }
> > > > > > > >
> > > > > > > > And here's the met-5.1 "rirw" config file option:
> > > > > > > >
> > > > > > > > rirw = {
> > > > > > > >
> > > > > > > > track = NONE;
> > > > > > > >
> > > > > > > > adeck = {
> > > > > > > >
> > > > > > > > time = "24";
> > > > > > > >
> > > > > > > > exact = TRUE;
> > > > > > > >
> > > > > > > > thresh = >=30.0;
> > > > > > > >
> > > > > > > > }
> > > > > > > >
> > > > > > > > bdeck = adeck;
> > > > > > > >
> > > > > > > > }
> > > > > > > > Notice here the big change is that we have separate
setting
> for
> > > the
> > > > > > ADECK
> > > > > > > > and BDECK. So you could define different logic for
the
> > different
> > > > > > tracks
> > > > > > > > which should give finer control of the filtering.
> > > > > > > >
> > > > > > > > Please look on pages 260-261 for more info on this:
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v5.1.pdf
> > > > > > > >
> > > > > > > > Hopefully that helps.
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > > John
> > > > > > > >
> > > > > > > > On Wed, Jan 27, 2016 at 10:46 AM, Andrew Penny - NOAA
> Affiliate
> > > via
> > > > > RT
> > > > > > <
> > > > > > > > met_help at ucar.edu> wrote:
> > > > > > > >
> > > > > > > >>
> > > > > > > >> Wed Jan 27 10:46:01 2016: Request 74883 was acted
upon.
> > > > > > > >> Transaction: Ticket created by andrew.penny at noaa.gov
> > > > > > > >> Queue: met_help
> > > > > > > >> Subject: rapid intensification filter not working
for
> > > met-5.1
> > > > > > > >> Owner: Nobody
> > > > > > > >> Requestors: andrew.penny at noaa.gov
> > > > > > > >> Status: new
> > > > > > > >> Ticket <URL:
> > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=74883
> > > > > > >
> > > > > > > >>
> > > > > > > >>
> > > > > > > >> Hi there,
> > > > > > > >>
> > > > > > > >> I've been trying to use the "rapid_inten" filter
option in
> > > MET-TC,
> > > > > but
> > > > > > > it
> > > > > > > >> doesn't appear to work with version 5.1. It appears
that
> this
> > > > filter
> > > > > > is
> > > > > > > not
> > > > > > > >> being applied. However, my script works fine with
version
> 5.0.
> > > > > > > >>
> > > > > > > >> Thanks,
> > > > > > > >> Andy
> > > > > > > >>
> > > > > > > >> --
> > > > > > > >> Andrew Penny
> > > > > > > >> Hurricane Model Diagnostician
> > > > > > > >> SRG / National Hurricane Center
> > > > > > > >> 11691 SW 17th Street
> > > > > > > >> Miami, FL 33165-2149
> > > > > > > >> phone: 305.229.4457
> > > > > > > >> email: andrew.penny at noaa.gov
> > > > > > > >>
> > > > > > > >>
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > > --
> > > > > Andrew Penny
> > > > > Hurricane Model Diagnostician
> > > > > SRG / National Hurricane Center
> > > > > 11691 SW 17th Street
> > > > > Miami, FL 33165-2149
> > > > > phone: 305.229.4457
> > > > > email: andrew.penny at noaa.gov
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Andrew Penny
> > > Hurricane Model Diagnostician
> > > SRG / National Hurricane Center
> > > 11691 SW 17th Street
> > > Miami, FL 33165-2149
> > > phone: 305.229.4457
> > > email: andrew.penny at noaa.gov
> > >
> > >
> >
> >
>
>
> --
> Andrew Penny
> Hurricane Model Diagnostician
> SRG / National Hurricane Center
> 11691 SW 17th Street
> Miami, FL 33165-2149
> phone: 305.229.4457
> email: andrew.penny at noaa.gov
>
>
------------------------------------------------
More information about the Met_help
mailing list