[Met_help] [rt.rap.ucar.edu #38895] History for General MET Question

[RAL HelpDesk {for Tressa Fowler}]

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

We are proposing to the NUOPC group to use met to verify ensemble forecasts in a common way.  Even though MET was written with WRF in mind, this there any reason it shouldn't work with other models such as GFS, NOGAPS, - any gridded model?

Thanks

Bob


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

Subject: General MET Question
From: Tressa Fowler
Time: Tue Jun 29 13:23:10 2010

Hi Bob,

We would really like to see you guys use MET for NUOPC. I wonder if
maybe we should have a teleconference or something to go over some of
these details and find out how we can best help you. What do you
think?

We could discuss the various types of files you will be dealing with,
the output you would like to see, and possibly even the potential use
of our database and display system. We also could talk about providing
some training to the other NUOPC folks or arranging to get them to a
tutorial.

As for this specific question, John has provided the following
technical answer:

>From a software engineering perspective, he's correct that we
basically just need gridded model output that's on a regular grid and
pressure levels in the vertical.  A larger issue would be data
formats. They'd certainly need GRIB2 support... and perhaps other
formats as well?  Another issue would be how MET would be used within
the context of NUOPC.  If it becomes "operational" in some way, they'd
need a higher level of 24x7 support.

Thanks,

Tressa

On Mon Jun 28 13:38:06 2010, craigr at offutt.af.mil wrote:
> We are proposing to the NUOPC group to use met to verify ensemble
>    forecasts in a common way.  Even though MET was written with WRF
in
>    mind, this there any reason it shouldn't work with other models
>    such as GFS, NOGAPS, - any gridded model?
>
> Thanks
>
> Bob



------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #38895] General MET Question 
From: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Time: Thu Jul 01 14:38:31 2010

Tressa, Thanks for your input on this.  Even though I have suggested
MET, there has been some push back from at least one NUOPC member -
much of it because they aren't familiar with MET.  I have yet to hear
from other members.  As I get the opportunity, I will promote MET to
this committee. However, I was reminded that my only task is to come
up with a common set of metrics that all NUOPC players can use and not
dictate how they do it - so it will be an uphill battle.

The
following is a list of the parameters I am proposing we use.  Is there
anything there that MET can't do?

500mb Hgt Anomaly correlation
(using NCEP gridded climo data)

10 Meter Winds (probs > 20, 35
50kts) - Brier Score, RMSE/Bias of ensemble mean

Precipitation
(probs > .25, .50, 1", 2") - Brier Score, CRPS

850 mb Temperature -
RMSE/Bias of ensemble mean
700 mb Dewpoint - RMSE/Bias of ensemble
mean

This list will change at our next get-to-gether, but hopefully
not much.

Bob

-----Original Message-----
From: RAL HelpDesk
{for Tressa Fowler} [mailto:met_help at ucar.edu] 
Sent: Tuesday, June
29, 2010 2:23 PM
To: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: [rt.rap.ucar.edu #38895] General MET Question 

Hi Bob,
We would really like to see you guys use MET for NUOPC. I wonder if
maybe we should have a teleconference or something to go over some of
these details and find out how we can best help you. What do you
think? 

We could discuss the various types of files you will be
dealing with, the output you would like to see, and possibly even the
potential use of our database and display system. We also could talk
about providing some training to the other NUOPC folks or arranging to
get them to a tutorial. 

As for this specific question, John has
provided the following technical answer: 

>From a software
engineering perspective, he's correct that we basically just need
gridded model output that's on a regular grid and pressure levels in
the vertical.  A larger issue would be data formats. They'd certainly
need GRIB2 support... and perhaps other formats as well?  Another
issue would be how MET would be used within the context of NUOPC.  If
it becomes "operational" in some way, they'd need a higher level of
24x7 support.

Thanks,

Tressa

On Mon Jun 28 13:38:06 2010,
craigr at offutt.af.mil wrote:
> We are proposing to the NUOPC group to
use met to verify ensemble
>    forecasts in a common way.  Even
though MET was written with WRF in
>    mind, this there any reason
it shouldn't work with other models
>    such as GFS, NOGAPS, - any
gridded model?
> 
> Thanks
> 
> Bob

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