In CABLE lakes are currently (Nov 2018) represented as permanently saturated bare soil. To be consistent with this the soil colour albedo should always take the saturated limit. However, within ACCESS (at least), the albedo code is called both prior and after to the lake tile is adjusted to saturation. Different parts of the code can use different albedos and energy conservation is lost.
The simplest method to avoid this is to adjust the soil colour parameterisation over lakes: In soilcol_albedo change
keyword_keepgit
type_defect
| by inh599In CABLE lakes are currently (Nov 2018) represented as permanently saturated bare soil. To be consistent with this the soil colour albedo should always take the saturated limit. However, within ACCESS (at least), the albedo code is called both prior and after to the lake tile is adjusted to saturation. Different parts of the code can use different albedos and energy conservation is lost.
The simplest method to avoid this is to adjust the soil colour parameterisation over lakes: In soilcol_albedo change
DO ib = 1,2 ! Number of wavebands (vis, nir)
END DO
to
DO ib = 1,2 ! Number of wavebands (vis, nir) inc = MAX(0.11-0.40*ssnow%wb(:,1), 0._r_2)
inc = MAX(0.11-0.40*ssnow%sfc, 0._r_2) ENDWHERE albsod(:,ib) = MIN(albsat(INT(soil%soilcol),ib)+inc, albdry(INT(soil%soilcol),ib)) albsoi(:,ib) = albsod(:,ib) END DO
Minor adjustments maybe needed for the GW model
Issue migrated from trac:207 at 2023-11-27 11:27:27 +1100