Complete design revamp of surface flux calculators to use simple fixed point iterations, and spatially varying roughness and land surface resistance parameters.

[akshaysridhar]

No description provided.

[glwagner]

We have a design coming together in ClimaOcean if it is of relevance. The major change is to recognize that there are three states in play for surface fluxes (not just two): the two "bulk" states on either side of the interface, and the surface state. Currently the design of surface fluxes assumes that there is just one bulk state (the atmosphere) and one surface state. Additionally, incoming and outgoing radiation comes into play.

For example, a typical model for surface soil moisture effectively computes the surface humidity at the air-soil interface as a function of the "interior" soil temperature (from which a saturation specific humidity is derived). Similarly, skin temperature formulations model the surface temperature as a function of the interior state, atmos state, and additional imposed fluxes like radiation.

[akshaysridhar]

We have a design coming together in ClimaOcean if it is of relevance. The major change is to recognize that there are three states in play for surface fluxes (not just two): the two "bulk" states on either side of the interface, and the surface state. Currently the design of surface fluxes assumes that there is just one bulk state (the atmosphere) and one surface state. Additionally, incoming and outgoing radiation comes into play.

For example, a typical model for surface soil moisture effectively computes the surface humidity at the air-soil interface as a function of the "interior" soil temperature (from which a saturation specific humidity is derived). Similarly, skin temperature formulations model the surface temperature as a function of the interior state, atmos state, and additional imposed fluxes like radiation.

@glwagner Can you please link the ClimaOcean design issue here? Thanks!

[glwagner]

There's no design issue, but the implementation is being done here: CliMA/ClimaOcean.jl#294

I can walk you through the code if you like. We can use that excuse to develop some documentation too.