search

CABLE-LSM / CABLE

Home to the CABLE land surface model and its documentation
https://cable.readthedocs.io/en/latest/
Other
8 stars 3 forks source link

xkleaf no longer used in casa_coeffplant #275

Open rml599gh opened 2 months ago

rml599gh commented 2 months ago

Line 767 of casa_cnp.F90 refers to trac ticket 242. It should be trac ticket 243. That ticket has removed the use of xkleaf, xkleafcold and xkleafdry in the calculation of casaflux%kplant(:,leaf). These terms seem to account for cold and drought stress and are calculated in casa_xrateplant. This is a change since CABLE2.4 so, I assume, is a real difference between ESM1.5 and ESM1.5/CABLE3. Was the impact of this change tested? If we don't want to retain the cold/dry stress capability then code could be cleaned up to remove casa_xrateplant and reading of various parameters from pftlookup file. Or the original code should be corrected (if it was a bug) and reinstated.

rml599gh commented 2 months ago

The change to the code was made at revision 8690 on 8/2/22. While the ticket describes the addition of xkleafcold and xkleafdry as 'clearly a bug' the addition does appear to be consistent with Equation 15 in Arora and Boer (2005) from which the cold and drought stress parameterisation was taken. Each term is a rate of conversion between leaf carbon and litter. The sum comprises 'normal' conversion and adds any addition conversion when temperature is below a threshold or soil moisture is below a threshold. 'Normal' conversion (from input parameters) is moderated by xkleaf. When using prescribed phenology, this is 1 for most of the year but 0 for the 14 days when leaves are growing so there is no conversion to litter in this period (unless it is too cold or dry). When using climate-based phenology, xkleaf is set to 100 for phenology phases 0 and 3 (leaf loss and no leaves). I assume this just means that 'normal' conversion is increased by two orders of magnitude and would swamp any additional cold or drought induced conversion. Recommend testing the impact of this change and then reinstating the code. May also be worth checking the sensitivity to parameter choice as the maximum drought and cold leaf loss rates appear to be quite different to those used in Arora and Boer while the 'shape parameters' are the same as Arora and Boer.