(SOLVED) Reducing topography in WRF (a different method)

[Liliputian_123]

Dear WRF experts,

I have a post here about reducing the topography of the geo_em files (https://forum.mmm.ucar.edu/phpBB3/viewtopic.php?f=30&t=8794&e=1&view=unread#unread). I used the read_wrf_nc to multiply a constant on the topography. However, it is hard specify the "specific" region where this should be done.

I would like to ask some opinon.
What if I reduce the topography of the input files before creating the geo_em files? For example, If my input is fron NCEP 6-hourly data, then I'll multiply the topography variable here with a constan.

If i do the former method, it will be hard to estimate the region for nested runs as I have to estimate the same location in 2 files with diffrent resolution.

Is the 2nd method also correct?

I'll appreciate any advice on this matter.

 

[mgduda]

In principle, you could either modify the topography input files or modify the HGT_M field in the geo_em.d01.nc file -- either would ultimately have the same effect (ignoring small perturbations introduced by the horizontal interpolation methods of the geogrid program in the former case). However, the WRF model does not use the topography field from, e.g., the NCEP datasets, but rather, the WRF model uses the topography field provided by the geogrid program. So, if you'd like to modify the original topography data, you would need to modify the binary tiles of either GMTED2010 or GTOPO30 data that are processed by the geogrid program.

 

[Liliputian_123]

Thank you. I understand it now.