Annual mean of LPJmL's per-time-step decomposition response: a modified Arrhenius temperature function after Lloyd and Taylor (1994), multiplied by a cubic function of the soil's degree of saturation, and capped at 1.
Three properties matter when reading the result. The temperature term is normalised so that it is exactly 1 at 10 degrees Celsius, which is what the "at 10 degrees" in LPJmL's rate constants refers to. The moisture term is not normalised at all and peaks at 0.937. And the cap is imposed by LPJmL's source rather than by its published equations: it binds from about 10.7 degrees at optimal moisture, so every warm, well-watered cell returns exactly 1 and the fast pool cannot turn over faster than its nominal rate. The cap is applied per time step, before averaging, because capping the mean is a different function.
Source
Schaphoff, S. et al. (2018). LPJmL4 - a dynamic global vegetation
model with managed land - Part 1: Model description. Geoscientific Model
Development, 11, 1343-1375. doi:10.5194/gmd-11-1343-2018
, Eq. 45
(temperature) and Eq. 96 (moisture); temperature form after Lloyd, J. &
Taylor, J. A. (1994). doi:10.2307/2389824
. The cap at 1 is in LPJmL's
source (src/soil/littersom.c) and not in the published equations.
Arguments
- temp_soil_c
Numeric soil temperature series (degrees Celsius). LPJmL drives this with per-layer soil temperature, not air temperature; the two differ in both damping and lag, so supplying air temperature is a substitution the caller is making, not a detail.
- theta
Numeric series of the soil's degree of saturation (0-1), as LPJmL's fractional soil water content reports it. Use it raw; it is already a saturation fraction, not a volumetric content needing a porosity divisor.