[Met_help] using NCEP stage IV data

[John Halley Gotway]

Todd,

Yes, Grid-Stat is doing what you guessed.  It tries to compute a bunch of
matched pairs, and if one of those value (fcst or obs) is missing, it
doesn't use that matched pair.

There are 2 parameters in the Grid-Stat configuration file related to how
to handle missing data: interp_thresh and nbr_thresh.  Please take a look
at the comments in the config file for a description of what they do:
http://www.dtcenter.org/met/users/support/online_tutorial/METv2.0/grid_stat/GridStatConfig_default

If you were to run the MODE tool on this data, you could select how you'd
like the missing data to be handled.  In the MODE configuration file, the
mask_missing_flag parameter tells MODE how to handle missing data in one
or both of the input files:
http://www.dtcenter.org/met/users/support/online_tutorial/METv2.0/mode/WrfModeConfig_default

Hope that helps.  Just let us know if any more questions arise in your use
of MET.

Thanks,
John Halley Gotway
johnhg at ucar.edu

> We are currently verifying WRF precip forecasts using NCEP stage IV
> data.  The NCEP stage IV data is a composite of regional data, and
> occasionally regions are missing from the composite.  The missing areas
> are masked using a bitmap in the grib1 encoding.  I've attached an
> example image of stage IV data.  The blue region is where data exists
> and the black region is the region that is masked out.
>
> I'm wondering how the grid_stat tool handles the missing regions,
> especially when calculating GSS.  That is, what does grid_stat do with
> precip points where WRF forecast data is available, but stage IV data is
> not available (it's masked out)?
>
> We're verifying both the 'FULL' domain and the domain using the
> CONUS.poly region.
>
> Thank you.
>
> Todd Hutchinson
>
> _______________________________________________
> Met_help mailing list
> Met_help at mailman.ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/met_help
>