[Met_help] [rt.rap.ucar.edu #61776] History for RE: [rt.rap.ucar.edu #61428/61504] MODE setup question (UNCLASSIFIED)
John Halley Gotway via RT
met_help at ucar.edu
Mon Jun 10 13:54:54 MDT 2013
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Classification: UNCLASSIFIED
Caveats: NONE
Hi John,
I apologize for not going through channels. John Raby isn't in today, and I
wasn't sure how he normally contacts you with questions re: MODE. ...so, my
apologies upfront.
Had a "quick MODE logic question"...
I've built a script to automate our multiple hours of data processing with
MODE. The 3 options we'll be using are:
1. Define objects as all values between XXX.X and YYY.y
2. Define objects as all values greater than XXX.x
3. Define objects as all values less than XXX.x
Here are the parameters set in the script for each of the 3 cases:
1. Objects defined as "between XXX.X and YYY.y"
raw_thresh <= user entry (highest value)
conv_radius >= user entry (lowest value)
conv_thresh 0 (skips this step)
2. Objects defined as ">= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius >= user entry (lowest value accepted for object)
conv_thresh 1 (1 grid unit is default)
3. Objects defined as "<= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius <= user entry (highest value accepted for object)
conv_thresh 1 (1 grid unit is default)
Aside: The rest of the config file variables are default values.
Are the 3 cases (object definitions) and configuration parameters correctly
setup for MODE?
(Test results on the above, look promising.)
Please advise.
Thanks,
Gail Vaucher
Army Research Lab
WSMR, NM 88002
(575)678-3237
Classification: UNCLASSIFIED
Caveats: NONE
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61776] RE: [rt.rap.ucar.edu #61428/61504] MODE setup question (UNCLASSIFIED)
From: John Halley Gotway
Time: Mon Jun 10 13:27:11 2013
Gail,
Looks like you've got the radius and threshold settings switched
around. They should look like this:
1. Objects defined as "between XXX.X and YYY.y"
raw_thresh <= user entry (highest value)
conv_radius 0 (skips this step)
conv_thresh >= user entry (lowest value)
2. Objects defined as ">= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius 1 (1 grid unit is default)
conv_thresh >= user entry (lowest value accepted for object)
3. Objects defined as "<= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius 1 (1 grid unit is default)
conv_thresh <= user entry (highest value accepted for
object)
The convolution radius (conv_radius) controls how much smoothing is
done. The larger the radius, the smoother the resulting objects. I
don't know why you list a convolution radius of 1 as the
"default value". Using a convolution radius of 0 skips the smoothing
step, and will result in a large number of relatively small objects.
You should select the amount of smoothing in such a way to
define objects that capture the feature you're interested in studying.
For example, if you're interested in the performance of large MCS's,
you'd do a lot of smoothing. But if you're looking for
small popcorn convection, you'd set it lower. For precip, it's
important to match the amount of smoothing to the features you're
studying. But other fields, such as temperature and relative humidity
will already be pretty smooth. So you may choose to do no convolution
as all (set conv_radius = 0). Make sense?
Thanks,
John Halley Gotway
met_help at ucar.edu
On 06/07/2013 02:48 PM, gail.t.vaucher.civ at mail.mil via RT wrote:
>
> Fri Jun 07 14:48:28 2013: Request 61776 was acted upon.
> Transaction: Ticket created by gail.t.vaucher.civ at mail.mil
> Queue: met_help
> Subject: RE: [rt.rap.ucar.edu #61428/61504] MODE setup
question (UNCLASSIFIED)
> Owner: Nobody
> Requestors: gail.t.vaucher.civ at mail.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61776 >
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> Hi John,
>
> I apologize for not going through channels. John Raby isn't in
today, and I
> wasn't sure how he normally contacts you with questions re: MODE.
...so, my
> apologies upfront.
>
> Had a "quick MODE logic question"...
>
> I've built a script to automate our multiple hours of data
processing with
> MODE. The 3 options we'll be using are:
>
> 1. Define objects as all values between XXX.X and YYY.y
> 2. Define objects as all values greater than XXX.x
> 3. Define objects as all values less than XXX.x
>
> Here are the parameters set in the script for each of the 3 cases:
>
> 1. Objects defined as "between XXX.X and YYY.y"
> raw_thresh <= user entry (highest value)
> conv_radius >= user entry (lowest value)
> conv_thresh 0 (skips this step)
>
> 2. Objects defined as ">= XXX.X"
> raw_thresh >=-9999 (to avoid boundary false data)
> conv_radius >= user entry (lowest value accepted for object)
> conv_thresh 1 (1 grid unit is default)
>
> 3. Objects defined as "<= XXX.X"
> raw_thresh >=-9999 (to avoid boundary false data)
> conv_radius <= user entry (highest value accepted for
object)
> conv_thresh 1 (1 grid unit is default)
>
> Aside: The rest of the config file variables are default values.
>
> Are the 3 cases (object definitions) and configuration parameters
correctly
> setup for MODE?
>
> (Test results on the above, look promising.)
>
> Please advise.
>
> Thanks,
> Gail Vaucher
> Army Research Lab
> WSMR, NM 88002
> (575)678-3237
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #61428/61504] MODE setup question (UNCLASSIFIED)
From: gail.t.vaucher.civ at mail.mil
Time: Mon Jun 10 13:44:03 2013
Classification: UNCLASSIFIED
Caveats: NONE
Hi John,
Thanks for flagging the correction. (Something didn't look right,
which
prompted the email; Looks like you solved the mystery. Thanks!)
I see what you mean re: having a reduced need to smooth parameters
like
temperature & RH. Will change the default to conv_radius = 0 for all
3
options. (The config file I initially received as my "template" had
the
single grid unit. But, that was for another parameter.)
Thanks, again.
Gail
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Monday, June 10, 2013 1:27 PM
To: Vaucher, Gail T CIV (US)
Subject: Re: [rt.rap.ucar.edu #61776] RE: [rt.rap.ucar.edu
#61428/61504] MODE
setup question (UNCLASSIFIED)
Gail,
Looks like you've got the radius and threshold settings switched
around. They
should look like this:
1. Objects defined as "between XXX.X and YYY.y"
raw_thresh <= user entry (highest value)
conv_radius 0 (skips this step)
conv_thresh >= user entry (lowest value)
2. Objects defined as ">= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius 1 (1 grid unit is default)
conv_thresh >= user entry (lowest value accepted for object)
3. Objects defined as "<= XXX.X"
raw_thresh >=-9999 (to avoid boundary false data)
conv_radius 1 (1 grid unit is default)
conv_thresh <= user entry (highest value accepted for
object)
The convolution radius (conv_radius) controls how much smoothing is
done. The
larger the radius, the smoother the resulting objects. I don't know
why you
list a convolution radius of 1 as the "default value". Using a
convolution
radius of 0 skips the smoothing step, and will result in a large
number of
relatively small objects. You should select the amount of smoothing
in such a
way to define objects that capture the feature you're interested in
studying.
For example, if you're interested in the performance of large MCS's,
you'd do
a lot of smoothing. But if you're looking for small popcorn
convection, you'd
set it lower. For precip, it's important to match the amount of
smoothing to
the features you're studying. But other fields, such as temperature
and
relative humidity will already be pretty smooth. So you may choose to
do no
convolution as all (set conv_radius = 0). Make sense?
Thanks,
John Halley Gotway
met_help at ucar.edu
On 06/07/2013 02:48 PM, gail.t.vaucher.civ at mail.mil via RT wrote:
>
> Fri Jun 07 14:48:28 2013: Request 61776 was acted upon.
> Transaction: Ticket created by gail.t.vaucher.civ at mail.mil
> Queue: met_help
> Subject: RE: [rt.rap.ucar.edu #61428/61504] MODE setup
question
> (UNCLASSIFIED)
> Owner: Nobody
> Requestors: gail.t.vaucher.civ at mail.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61776
> >
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> Hi John,
>
> I apologize for not going through channels. John Raby isn't in
today,
> and I wasn't sure how he normally contacts you with questions re:
> MODE. ...so, my apologies upfront.
>
> Had a "quick MODE logic question"...
>
> I've built a script to automate our multiple hours of data
processing
> with MODE. The 3 options we'll be using are:
>
> 1. Define objects as all values between XXX.X and YYY.y
> 2. Define objects as all values greater than XXX.x
> 3. Define objects as all values less than XXX.x
>
> Here are the parameters set in the script for each of the 3 cases:
>
> 1. Objects defined as "between XXX.X and YYY.y"
> raw_thresh <= user entry (highest value)
> conv_radius >= user entry (lowest value)
> conv_thresh 0 (skips this step)
>
> 2. Objects defined as ">= XXX.X"
> raw_thresh >=-9999 (to avoid boundary false data)
> conv_radius >= user entry (lowest value accepted for object)
> conv_thresh 1 (1 grid unit is default)
>
> 3. Objects defined as "<= XXX.X"
> raw_thresh >=-9999 (to avoid boundary false data)
> conv_radius <= user entry (highest value accepted for
object)
> conv_thresh 1 (1 grid unit is default)
>
> Aside: The rest of the config file variables are default values.
>
> Are the 3 cases (object definitions) and configuration parameters
> correctly setup for MODE?
>
> (Test results on the above, look promising.)
>
> Please advise.
>
> Thanks,
> Gail Vaucher
> Army Research Lab
> WSMR, NM 88002
> (575)678-3237
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
>
>
Classification: UNCLASSIFIED
Caveats: NONE
------------------------------------------------
More information about the Met_help
mailing list