Soil Hydraulic Parameter from surface file for eCLM

[jjokella]

During a meeting-discussion ( @Toeroeoe ), it came up that the functionality of setting soil hydraulic parameters from surface files, while used for perturbing the parameters in eCLM-PDAF, is not restricted to eCLM-PDAF.

It can make sense to set soil hydraulic parameters specifically for a standalone eCLM simulation, e.g. when these parameters are known from literature.

Therefore, the #ifdef USE_PDAF statements are removed in src/clm5/biogeophys/SoilStateInitTimeConstMod.F90.

• arrays for parameter reading are only allocated once the namelist settings are invoked
• documentation of soil hydraulic parmaters is separated from the PDAF-specific atmospheric forcing noise documentation

Namelist settings

• soil_hyd_inparm_from_file
• soil_hyd_inparm_from_file_adj

Documentation

• https://hpscterrsys.github.io/eCLM/users_guide/introduction/soil_hydraulic_parameters.html

[kvrigor]

is the additional locally allocated memory for the introduced arrays, which will now be allocated for all eCLM simulations, rather than only for eCLM-PDAF simulations.

This could be avoided if extra parameter checking on the surface file (e.g. parameters_in_file) is not used as a switch to use modified soil hydraulic parameters. Perhaps making this as an explicit namelist setting under clm_soilstate_inparm would be better?

eCLM/src/clm5/biogeophys/SoilStateInitTimeConstMod.F90

Lines 56 to 58 in 7877e5c

	character(len=*), parameter :: nl_name = 'clm_soilstate_inparm' ! Namelist name
	! MUST agree with name in namelist and read
	namelist / clm_soilstate_inparm / organic_frac_squared

[jjokella]

I think the namelist setting idea is the correct way to go! I added a implementation.

This way:

• memory is only allocated if really needed
• multiple error handling can be inserted
  • only one of the parameter input types can be true
  • all of the parameters have to be in the surface data file