The changes from @mvertens in #690 include a change to med_phases_prep_rof_mod that appears to zero the irrigation flux if med_coupling_active(comprof,complnd) is false. That condition will occur when using DLND to provide forcings to ROF.
As @mvertens explained, this logic was needed because med_phases_prep_rof_irrig relies on the rof-to-lnd route handle being set up, and currently that isn't done when running with DLND.
The ideal thing to do in this case would be to set up the rof-to-lnd route handle even with DLND (similarly to how we set up a lnd-to-glc route handle even with DGLC), so that we can still call med_phases_prep_rof_irrig in the DLND case.
(If that isn't feasible, then I think it could be possible to skip med_phases_prep_rof_irrig but not zero the irrigation field, so that we use the normal, conservatively-mapped irrigation field. The problem with this is that it would mean that we'd have different irrigation mappings in the DLND vs. active land case. This would be non-ideal both because of the differences themselves, and because in this case, the differences could mean that we'd be more likely to end up with negative river water volumes in the DLND case - which is part of what the specialized irrigation mapping tries to avoid.)
The changes from @mvertens in #690 include a change to med_phases_prep_rof_mod that appears to zero the irrigation flux if
med_coupling_active(comprof,complnd)is false. That condition will occur when using DLND to provide forcings to ROF.As @mvertens explained, this logic was needed because
med_phases_prep_rof_irrigrelies on the rof-to-lnd route handle being set up, and currently that isn't done when running with DLND.The ideal thing to do in this case would be to set up the rof-to-lnd route handle even with DLND (similarly to how we set up a lnd-to-glc route handle even with DGLC), so that we can still call
med_phases_prep_rof_irrigin the DLND case.(If that isn't feasible, then I think it could be possible to skip
med_phases_prep_rof_irrigbut not zero the irrigation field, so that we use the normal, conservatively-mapped irrigation field. The problem with this is that it would mean that we'd have different irrigation mappings in the DLND vs. active land case. This would be non-ideal both because of the differences themselves, and because in this case, the differences could mean that we'd be more likely to end up with negative river water volumes in the DLND case - which is part of what the specialized irrigation mapping tries to avoid.)