fcst and das (run) should use the same land setup

[rtodling]

somehow this does not show as a problem in most cases! For example, x0046a (and prePP) and x0045a - and other runs, the forecast reproduces the DAS during the period over IAU (as it should). The explanation for what is going on is provided by Rolf - quote here from email exchange:

The fvpsas.log indicates that several Surface rc variables (see list below) were set explicitly in GEOS_SurfaceGridComp.rc, presumably because they have been uncommented by the setup script. But all of these variables match the default setting that is encoded in the MAPL_GetResource() calls. For example, in GEOS_SurfaceGridComp.F90, LAND_PARAMS defaults to "NRv7.2" when the LAND_PARAMS cannot be found in the resource file (as is the case in the run associated with prePP.gcm.log.*.txt):

if (LSM_CHOICE.eq.1) then
call MAPL_GetResource (SCF, LAND_PARAMS, label='LAND_PARAMS:', DEFAULT="NRv7.2", RC )

And LAND_PARAMS is also set to "NRv7.2" if the corresponding line in the Surface rc file is simply uncommented without making other changes.

In a nutshell, the model defines defaults in two places, the rc file and the MAPL_GetResource() calls. In recent tags, the respective defaults are consistent (which isn't true if you go back a year or so). Specifying the default settings in the rc file only makes the defaults show up conveniently in the log file, but nothing else changes unless you specify non-default values in the rc file.

Also: this is being labeled here as zero-diff since it is (incidentally) so for the choices land in the x-exps and FP/FPP/prePP runs. It is not the case for forecasts using the Icarus land -but I don't believe anyone is truly using this option anymore.