[Met_help] [rt.rap.ucar.edu #73740] History for wrf model verfication

John Halley Gotway via RT met_help at ucar.edu
Thu Oct 29 15:44:24 MDT 2015


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

Hi Met help,
I have wrfoutput for 4 months.  I need to valid the data using Met package
in linux machine.  Kindly suggest the tools required. for installing the
Met package and suggest the version as I am new to MET tools.  I have point
data like

latitude longitude and value of rainfall
example
lat      lon     rainfall
12.0  76.0  12
12.1  76.0  13

Kindly suggest the scores that can be computed here.
Thanks and regards


*K. Naga Ratna*


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

Subject: wrf model verfication
From: John Halley Gotway
Time: Thu Oct 15 14:24:02 2015

K. Naga Ratna,

It sounds like you are in the following situation:
(1) You have 4 months of gridded wrfout files.
(2) You have ASCII point observations of accumulated precipitation.

Here's what I would suggest:
(1) Postprocess your wrfout files using either the Unified
PostProcessor or
the WRF p-interp utility.  I'd recommend UPP since it writes GRIB1
files
which MET handles very well.

(2) Since you're verifying precipitation, you'll need your model
accumulations to match your observed accumulations.  Figure out the
accumulation interval of your observations (maybe it's 6-hourly, 12-
hourly,
or 24-hourly).  Run the MET pcp_combine tool on your model data to get
your
forecast precip into the same accumulation interval as your obs,
ending at
the same valid time.

For example, perhaps your observations are 24-hour accumulations,
ending at
12Z every day.  You'd need to define your forecast accumulations in
exactly
the same way.

If you're using WRF-ARW output with a runtime accumulation, you'll
likely
need to use the pcp_combine "-subtract" option.

(3) Format you ascii point observations into the format prescribed for
the
MET ascii2nc tool.  See the user's guide or ascii2nc usage statement
for
details.  Run your reformatted observations through the MET ascii2nc
tool.

(4) Lastly, run the MET point_stat tool for each valid time to compare
the
forecast precip to the observed point precipitation.

The currently available version of MET is 5.0, but we will likely be
releasing version 5.1 on Monday.

Hope that helps.

Thanks,
John Halley Gotway
met_help at ucar.edu

On Thu, Oct 15, 2015 at 2:14 PM, The RT System itself via RT <
met_help at ucar.edu> wrote:

>
> Thu Oct 15 14:14:09 2015: Request 73740 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by RT_System
>        Queue: met_help
>      Subject: wrf model verfication
>        Owner: johnhg
>   Requestors: knratna at gmail.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=73740 >
>
>
> This transaction appears to have no content
>

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


More information about the Met_help mailing list