In the fractional grid part code (GFS_surface_composites_post.F90), the surface exchange coefficients are re-computed. The bug is that z0max comes from noah-mp (when running with noah-mp), but ztmax comes from sfc_diff.f.
In addition, z0max and ztmax are always the same value in NOAH-MP, which I address in a different issue.
I believe the solution is to add bounded surface roughness [m] as an output in noahmpdrv.F90
Description
In the fractional grid part code (GFS_surface_composites_post.F90), the surface exchange coefficients are re-computed. The bug is that z0max comes from noah-mp (when running with noah-mp), but ztmax comes from sfc_diff.f. In addition, z0max and ztmax are always the same value in NOAH-MP, which I address in a different issue. I believe the solution is to add bounded surface roughness [m] as an output in noahmpdrv.F90