search

CABLE-LSM / CABLE

Home to the CABLE land surface model and its documentation
https://cable.readthedocs.io/en/latest/
Other
12 stars 6 forks source link

wbliq bug in cbl_soilsnow_main.F90 #484

Open rml599gh opened 21 hours ago

rml599gh commented 21 hours ago

wbliq is treated as a local variable in cbl_soilsnow_main.F90 but needs to be written into ssnow%wbliq. Otherwise ssnow%wbliq does not change over time with implications for its use when fwsoil_switch='Haverd2013'.

Originally noticed when trying to enable Haverd2013 with ESM1.5/ESM1.6 (#483) but will also impact offline runs using soilsnow and Haverd2013 but not groundwater.

Testing at selected single sites (with ACCESS met forcing) showed impacts for GPP.

10 day running mean GPP shown for a 1 year simulation. Black uses fwsoil_switch='standard'. Red is the result of just changing fwsoil_switch to 'Haverd2013'. Green is the result with code changes in cbl_soilsnow_main.F90 (two instances of wbliq = ssnow%wb - ssnow%wbice changed to ssnow%wbliq = ssnow%wb - ssnow%wbice)

Image

har917 commented 20 hours ago

There is other development work occurring in the AM3 codebase which intersects with this (seeking to improve energy/water balance and allowing for liquid/frozen water to have different densities). This bug could be a non-issue once CABLE3-MAIN is sync'd with that AM3 work (and ESM1.6 picks up the latest CABLE3)

rml599gh commented 20 hours ago

Thanks @har917 for the reminder that the simple fix I'd tested is assuming that rho_water = rho_ice. Sounds like we are best to wait for the more correct, AM3, version.