glacier surface types use ice in latent heat flux calculation?

[adamrhster@gmail_com]

I am curious to know if the surface energy balance parameterization in CLM4.5 for a bare glaciated land surface (with no snow on it) constrains the top glacier layer to be at or below the melting point. The CLM4.5 documentation states that glaciers are initialized below freezing in the model (250 K) but does not indicate that ice is present in the latent heat flux calculation. Is this so? From what I can tell, glaciers are treated as “soil” with ice like albedo and conductivity, but can theoretically have a ground temperature that exceeds the melting point. Is this so?

Also, I am specifically asking about glacier units that are fed in from a surface file and fixed on the land grid (do not evolve dynamically, e.g. beta version of CESM with 2-way coupling to CISM).

 

[adamrhster@gmail_com]

Well it turns out I stumbled over the answer. For CLM4.5, all glacier land types are treated as "soil", but are intialized with an ice lens filling the volume of each soil layer. If interactive ice sheets are turned on (in CESM1.2 - BG, FG, IG TG compsets), then glacier_mec surface types (mec = multiple elevation classes, currently only over greenland I think) will force the ice lens mass to be constant if any is melted during runtime. This does not appear to be the case for glacier surface types, where no such constraint is enforced. If a glacier surface type becomes depleted in its ice lens during runtime, the surface energy balance will allow for surface/subsurface temperatures to exceed the freezing point. This is not true of glacier_mec surface types.