set scm_iop_lhflxshflxTg = T for BOMEX & DYCOMSII-RF01 SCAM cases

[adamrher]

scm_iop_lhflxshflxTg = F for all scam cases, meaning lhflx,shflx & Tg are not read in from the IOP file, but computed by the surface components. Some IOP's have protocols where they need to be read in from the IOP file, typically a fixed surface flux with time. Two such cases are BOMEX & DYCOMSII-RF01, and so scm_iop_lhflxshflxTg needs to be set to .true. for these cases.

There may be other IOP cases in the usermods directory that also need to follow the protocol. I will add them to this issue as I become aware.

[adamrher]

@swrneale does SAS require reading in surface fluxes from the IOP file? It is not currently doing that.

[swrneale]

It should be. I think @jtruesdal committed that and it's a key part of the experimental design as those fluxes are semi-analytical. On a side note, don't we still have to be a little careful allowing the surface fluxes to be calculated since CLM is not initialized correctly is it? Maybe this has changed recently, I know John has been working on it.

[jtruesdal]

We've been working with a branch for SAS which still needs to be pushed to cam_dev. I think this will be a good time to update/correct that case in cam_dev along with all the rest. With regard to the land bias Rich is correct and I think most of the land cases should have had surface fluxes specified but were missing that default. I do have a CTSM issue #1362 to allow use of spun up initial land conditions and I would like to update the defaults here with that in mind. I think for most IOPs that means specifying a limited set of the surface forcing and allowing SCAM to calculate surface fluxes. We also allow SCAM to relax any prognostic to obs or a climatological average. All of the IOPs currently use relaxation to provide a realistic aerosol profile and a few of them relax T to obs over a 10 day time scale near the surface and 2 days e-folding time in the upper atmosphere. I'll send a link to a google doc which can be edited by you to pin down appropriate defaults for these cases. Including @andrewgettelman on this as well.