[Met_help] MODE netCDF question

[John Halley Gotway]

Bill,

MET doesn't read the NetCDF output of WRF directly, and we don't have any plans to add that.  We require that some post-processing be done to the WRF output files before it can be verified.
Specifically, it needs to be interpolated horizontally from the staggered grid the model uses to a more regular grid (currently we support data on Lambert Conformal, Polar Stereographic, Mercator, and
Lat/Lon projections).  Also, the vertical levels that the models use need to be converted to pressure level prior to verifying upper-air forecasts.  We recommend that WRF users run their WRF output
through the WRF Post-Processor (WPP), which does both of these steps and writes out a GRIB file.

For the next release of MET, we're working on two things that relate to this:
(1) Enhance our support for NetCDF files, specifically those conforming to the CF-convention of NetCDF.
(2) Read the output of the "pinterp" tool, which is a post-processor for WRF-ARW data that the MMM folks have made available.

However, it sounds like you're students don't actually NetCDF WRF output files right now.  They have their data in some binary format.  In order to use it in METv2.0, they'll need to either convert it
to GRIB version 1 or to NetCDF, making it look like the NetCDF output of the PCP-Combine tool in MET - which is rather specific.  If they'd like to do the latter, just have them write
met_help at ucar.edu to ask for more details.

Hope that helps.

John

wgallus at iastate.edu wrote:
> John,
> 
> Hope things are going well there.  I have 2 grad students working on
> different projects who we'd like to have them use MODE.  We have MET2.0
> installed here (I haven't tested it yet).    I know the tutorial
> mentions MODE can handle both GRIB and NetCDF input.  But, I also seem
> to recall there was a time where it maybe didn't handle the raw WRF
> version of NetCDF.
> 
> Can you tell me -- can MODE handle the raw wrf_output form of NetCDF? 
> Or, is there some conversion needed?  I know when you helped me with my
> binary data, I had to run things through TWO conversions. I thought the
> first one converted into wrf-style NetCDF but then you had to write that
> code to re-convert to a different NetCDF.  I could be remembering
> incorrectly, though.   I think my students have their precip data in
> binary form, but I'd guess they could adjust and instead use the raw
> wrf_output files if those can be input directly into MODE.
> 
> Bill
>