tslin2 commented 1 year ago

56

Namelist Vtable, z changed to TERRAIN call latlon_to_ij() don’t need to +360.0 for ERA5 data, if longitude is positive.

CharlesZheZhang commented 1 year ago

Thanks Cenlin and Tzu-Shun. As we discussed this afternoon, I agree to merge this pull request to develop branch.

tslin2 commented 1 year ago

if geo longitude is negative, the longitude of precipitation needs to add 360 degree in the subroutine interp_rainfall_nearest_neighbor

https://github.com/NCAR/hrldas/blob/9df967daec7fcc827a30982469fa8d88824aac24/hrldas/HRLDAS_forcing/create_forcing.F#L2734

revise to

    if (forcing_type=="ERA5") then
       if (geo_em%lon(ii,jj) < 0) then
          call latlon_to_ij(datastruct%proj, geo_em%lat(ii,jj), geo_em%lon(ii,jj)+360.0, x, y)
       else
          call latlon_to_ij(datastruct%proj, geo_em%lat(ii,jj), geo_em%lon(ii,jj), x, y)
       endif
    else
       call latlon_to_ij(datastruct%proj, geo_em%lat(ii,jj), geo_em%lon(ii,jj), x, y)
    endif

tslin2 commented 1 year ago

Another issue with the soil thickness vertical interpolation The problem is due to the var(:,:,k) in the following line https://github.com/NCAR/hrldas/blob/9df967daec7fcc827a30982469fa8d88824aac24/hrldas/IO_code/module_hrldas_netcdf_io.F#L1909 changing to var(:,k,:) works fine.

NCAR / hrldas

fixing issues with ERA5 climate #78

56