Snow thermal conductivity changes: Sturm1997 and Jordan1991

[kvrigor]

This PR implements @AdrienDams snow thermal conductivity fixes in CLM5.

Damseaux, A., Matthes, H., Dutch, V. R., Wake, L., and Rutter, N.: Impact of snow thermal conductivity schemes on pan-Arctic permafrost dynamics in the Community Land Model version 5.0, The Cryosphere, 19, 1539–1558, https://doi.org/10.5194/tc-19-1539-2025, 2025.

New namelist parameters in lnd_in

These fixes allow changing the snow thermal conductivity per land unit type. New namelist switches must be added in the lnd_in with defaults set to CLM6 defaults.

&clm_inparm
 snow_thermal_cond_method = 'Sturm1997'
 snow_thermal_cond_glc_method = 'Sturm1997'
 snow_thermal_cond_lake_method = 'Jordan1991'

For reference the namelist parameters reproducing the old defaults

&clm_inparm
 snow_thermal_cond_method = 'Jordan1991'
 snow_thermal_cond_glc_method = 'Jordan1991'
 snow_thermal_cond_lake_method = 'Jordan1991'

Code changes

Changes were based from these upstream PRs:

• Introduce namelist option for thermal conductivity of snow: Jordan vs. Sturm ESCOMP/CTSM#2148
• User control over snow thermal conductivity scheme over glaciers ESCOMP/CTSM#3072

[jjokella]

Testing this PR

• @Buliao1
• ICOS-sites as point-scale testcases:
  • first test: FI-Hyy (best observations)
  • possibly later added due to large amount of snow: FI-Sod / CH-Dav
• check that Jordan1991 reproduces previous snow variables
• check effect of Sturm1997-snow-thermal-conductivity on fluxes

TSMP2 branches that now include this PR:

• dev-pdaf-snow: https://github.com/HPSCTerrSys/TSMP2/tree/dev-pdaf-snow
• dev-pdaf-snow-params: https://github.com/HPSCTerrSys/TSMP2/tree/dev-pdaf-snow-params

Literature using Sturm1997:

Dutch, V. R., et al. (2024). Simulating net ecosystem exchange under seasonal snow cover at an arctic tundra site. Biogeosciences, 21(3), 825–841. http://dx.doi.org/10.5194/bg-21-825-2024

[jjokella]

Additional test

Came up in Z04-meeting: It would be interesting to check the influence of Sturm1997 on #90

[kvrigor]

Note on CI failure: #104 has to be merged first before this PR.

#104 actually depends on this PR since it already assumes the existence of the new snow_thermal_cond_* namelist parameters.

[Buliao1]

Hi all, here are the results:

1. CLM5.0 vs Jordan1991

Jordan1991 is able to reproduce the previous results very well.

2. CLM5.0 vs Sturm1997

Compared with the default CLM5.0, the Sturm1997 formulation yields improved snow depth simulations, especially in 2024.