logical variable "LOCAL" in WRF physics subroutine SFLX

[Dan Tyler]

Hi,

We're in the process of developing MPAS-A towards an end of using it to simulate the Martian atmosphere.  I did a similar thing with MM5 during my doctoral work in atmospheric dynamics at Oregon State University (I'm still here).

Using MPAS-A, we've already done a lot of Simple GCM (SGCM) cases where the only physics is: 1) relaxing air temps to an equilibrium temperature field computed offline, and 2) Rayleigh Friction drag near the surface.  We've already got surface pressures and all constants so as to be correct for Mars, which may be why we're having problems with activating the land use model.

What we're doing now is incorporating more realistic physics by turning on some of the available physics packages, with the near term goal of plugging in a grey IR scheme to the radiation call.  What we want to understand at that point is what problems the rigid lid might cause in relation to the very large atmospheric thermal tides, and what sensitivities there are.  All this, I suppose, is mostly beside the point...

The point:

Now, with the default YSU PBL scheme active and the default surface layer M-O scheme active (both set in namelist.atmosphere), things still work great.  When I try to activate the land surface model (LSM) scheme, that causes the WRF soil model routine SFLX to be called, the code makes NaNs in the very first time step.  I'm trying to figure out why, and looking for the easiest way to do so.  Of course, there is a great deal of land information required with the only choice available, NOAH.  I've probably missed on initializing something, or the liquid phase of water issue is being problematic, not sure...

What I do find in the code, in the call to subroutine LSM (in module_sf_noahdrv.F), is a logical flag "LOCAL" that passed to SFLX.  Since our initial simulations and tests with a diurnal cycle will be on a "cue ball" Mars, the use of this flag might be exactly what we need, but I'm not sure how to control it and whether it can be used to set globally uniform surface values.

First off, is this possible?

Second, where do I do that, meaning set globally uniform values?  Finally, does this seem like a reasonable course of action?

thanks,

  Dan Tyler

  Oregon State University