Terrain data after making initial condition is different from static.nc

[JasonLO]

Hi everyone:

I encounter an issue that is when I compared the ter variable in init.nc (the netcdf after making initial condition) and static.nc (the netcdf after making static data process), the difference is not only around boundary but whole domain! And the difference can up to several hundred meters, which is quite large value.
The MPAS version is v8.2.2 but not sure previous version also has the same issue or not. Seeking some comments about this issue.
Thank you!

Jason

 

[JasonLO]

Oh, I think I found the reason. The config_smooth_surfaces will apply smoothing to the terrain, so ter wil differ from static.nc

 

[gdicker]

You're correct the differences are expected behavior, due to how the init_atmosphere model works.

This terrain smoothing is controlled in the config_nsmterrain setting. The config_smooth_surfaces setting is for the other surfaces in the vertical direction - the "zeta" surfaces.

 

[JasonLO]

Hi gdicker,

Thanks your response! It's good to hear that my guess is right

Jason

 

[gdicker]

I should have added: If you want this turned off, you can do so with config_nsmterrain = 0. We sometimes do this when the land file should be taken "as-is."

Good luck with your next steps!