UPP vs. ndown for nesting

[tomgowan]

Hi,

I have netCDF wrfout files for a coarse, 10-km WRF simulation covering a large domain. I would like to use these wrfout files as initial and boundary conditions for a small, high-resolution, 2-km WRF simulation contained within the larger, 10-km domain.

It sounds like there are two ways to do this:
1) convert the wrfout files to grib using UPP and then link them into WPS to create met_em files
2) use ndown to directly read in the netCDF wrfout files and create the ICs and BCs for the high-resolution domain.

Is one of these methods preferred? Is one more accurate, reliable, and/or computationally efficient than the other? Any advice/insights would be greatly appreciated.

Thanks,
Tom

 

[kwerner]

Hi Tom,
In order to use the ndown method, you will also need to recreate the nested domain with WPS to create met_em* files for domain 2, so that you can then run real.exe for both domains, and then use the wrfout* files from the coarse domain to nest down to a finer-resolution domain. You can read the specifics here.

I'm not sure that one method is more accurate than the other. As my group doesn't manage UPP, I don't have any experience running it and can't comment on which method is preferred. There is, however, a UPP section in this forum, and you may be able to gain some insight there.