[Met_help] [rt.rap.ucar.edu #86059] History for FW: HiRA Climo Reference

Tue Jul 9 12:03:27 MDT 2019

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Just making sure you got this. 

Thanks
Bob Craig

-----Original Message-----
From: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN 
Sent: Wednesday, June 27, 2018 4:24 PM
To: 'met_help at ucar.edu' <met_help at ucar.edu>
Cc: BURRIS, KEVIN D Capt USAF ACC 16 WS/WXE <kevin.burris.3 at us.af.mil>;
SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL
<matthew.sittel.ctr at us.af.mil>
Subject: HiRA Climo Reference

John, now that I am reading a non-corrupt climo file thanks to your help, I
can proceed and generate my MPR files from the data.  Looking through the
MPR file, attached, there is data in the climo column for the non-HiRA case
but there is no data in that column for the HiRA neighborhood (121).    I
attached my configuration file.   I would expect to get a neighborhood
generated values like I have in the forecast column.   

What are your thoughts?

Thanks
Bob

----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: FW: HiRA Climo Reference
From: John Halley Gotway
Time: Fri Jul 06 09:58:50 2018

Bob,

Sorry for the delay in getting back to you on this.  Took me a while
to
re-orient myself to this data.  I ran plot_data_plane to display the
forecast and climo fields:
   met-6.1/bin/plot_data_plane 53a8fcf9-57db-4b93-9c55-843bd8cb53f5
fcst.ps
'name="APCP_06"; level="A6";'
   met-6.1/bin/plot_data_plane cfsr_freq_p6_world_mo_6.nc climo.ps
'name="CFSR_APCP_T01_00"; level="A6";'

I've attached png versions of the output.

The forecast data contains 6-hourly accumulated precip values from 0
up to
about 70mm over a pretty small domain.
The climo data is global and contains percentages representing the
frequency of 6-hour precip exceeding some thresholds.  Presumably, the
"T01" means the frequency of exceeding 0.01 inches (or .254 mm).

The presumption in MET is that the climatology data you use is defined
the *same
way* as the forecast data.  And that is not the case with your data.
The
forecast is 6-hour accumulations and the climo is the frequency of 6-
hour
accumulations exceeding some threshold.  That's why this case is
difficult
to understand.

To illustrate what's going on, I ran Point-Stat for a single Station
ID:
14048534
At that point, the forecast value = 0.81, the obs vale = 0, and the
climo
value = 25.5.  I looked at neighborhood sizes 3, 5, 7, 9, and 11 and
computed the MIN, MAX, and MEAN in those neighborhoods.  I also
applied
HiRA using the same neighborhood sizes.  I've attached the resulting
MPR
file.

In HiRA, the logic is this.  At each observation lat/lon location,
look at
the surround forecast values and compute the ratio of them exceeding
the
specified threshold.  In my example, I switched from >=.254 to >=1.0.
In
the 7x7 box, 16.327% of the 49 forecast values are >=1.0.  But we also
apply identical logic to the climo data... compute the ratio of climo
values >=1.0.  In the 7x7 box, 100% of them do.  In the 9x9, that 100%
drops to 88.8% percent because we've included some 0's in the
neighborhood.  You know that because the MIN climo value of the 9x9
box is
0.

That threshold of >=1.0 (>=.254) is meaningful for the forecast
data... but
it is not meaningful for the climo data.

What I suspect you want to do is treat that climo field as the already
compute fractional coverage field... and compare those values directly
to
the HiRA coverage values.  But that's not how the logic is set up.

Hope that helps clarify what's going on.

Thanks,
John

On Thu, Jul 5, 2018 at 2:07 PM robert.craig.2 at us.af.mil via RT <
met_help at ucar.edu> wrote:

>
> Thu Jul 05 14:06:31 2018: Request 86059 was acted upon.
> Transaction: Ticket created by robert.craig.2 at us.af.mil
>        Queue: met_help
>      Subject: FW: HiRA Climo Reference
>        Owner: Nobody
>   Requestors: robert.craig.2 at us.af.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86059 >
>
>
> Just making sure you got this.
>
> Thanks
> Bob Craig
>
> -----Original Message-----
> From: CRAIG, ROBERT J GS-12 USAF ACC 16 WS/WXN
> Sent: Wednesday, June 27, 2018 4:24 PM
> To: 'met_help at ucar.edu' <met_help at ucar.edu>
> Cc: BURRIS, KEVIN D Capt USAF ACC 16 WS/WXE
<kevin.burris.3 at us.af.mil>;
> SITTEL, MATTHEW C CTR USAF AFMC AFLCMC/HBAW-OL
> <matthew.sittel.ctr at us.af.mil>
> Subject: HiRA Climo Reference
>
> John, now that I am reading a non-corrupt climo file thanks to your
help, I
> can proceed and generate my MPR files from the data.  Looking
through the
> MPR file, attached, there is data in the climo column for the non-
HiRA case
> but there is no data in that column for the HiRA neighborhood (121).
I
> attached my configuration file.   I would expect to get a
neighborhood
> generated values like I have in the forecast column.
>
> What are your thoughts?
>
> Thanks
> Bob
>
>

------------------------------------------------