Changes between Version 238 and Version 239 of DevelopmentActivities/ORCHIDEE-DOFOCO

Timestamp:

2020-01-08T12:21:28+01:00 (4 years ago)

Author:

luyssaert

Comment:

--

DevelopmentActivities/ORCHIDEE-DOFOCO

@@ -239 +239 @@
 === Harvest ===
 All biomass harvest is recorded in a harvest variable Harvest_pool, this variable also stores the mass and dimensions of the harvest/mortality (absolute numbers!). Related variables were introduced: harvest_type,harvest_cut, and harvest_area. Wood product pools and fluxes from wood and biomass decomposition are calculated in a separate module dedicated to wood use. The dimension of the wood use pools were externalized and can be changed to better address regional differences when aiming for regional simulations. The default 1, 10 and 100 year pools should be replaced by 2, 17 and 50 years pools in Europe. 
+
+=== Leaf area index map ===
+Four flags have been identified that control the model behavior in terms of lai: ok_stomate, ok_pheno, impose_veg and read_lai. There is a 5th implicit flag which is whether restart files are used or not. If a restart file is used, the lai values will come from the sechiba restart file which is read at t=48. Given that each flag can take two values, we have 32 configurations in total. Out of these 32 configurations 10 are defined of which about 5 to 7 seem to be intended (for more details see Start and restart - Table 1). Many of the remaining 22 settings are inconsistent (i.e. running stomate to calculate a lai and reading an lai_map to prescribe lai), duplicate other settings, or would require further developments to work properly. Furthermore, the current code does not stop or warn when inconsistent settings are selected. The table (see Start and restart) proposes a scheme with 2 flags which can run with our without restart files, thus resulting in 8 different ways to control the lai in sechiba or the initial lai in stomate. The remaining 2 combinations are inconsistent and will stop the model.
+
+In the ORCHIDEE and ORCHIDEE-CN, canopy structure is prescribed by a single variable lai and the assumption of a turbid medium (Lambert-Beer). Consequently reading an lai value suffice to prescribe the entire canopy. In ORCHIDEE-CN-CAN, however, canopy structure has become a 3D property that can be calculated from the leaf biomass, stem biomass, the number of individuals and the assumptions that the trees follow a Poisson distribution in the horizontal plain, that the crowns are spherical and that the leaf biomass is uniformly distributed within the crowns. 
+
+The functionality to simply prescribe an lai value (either through impose or by reading a map) will need to be replaced by functionality that prescribes or reads the biomass, leaf age, and number of individuals. Given that the current lai map is based on a previous ORCHIDEE run, the same approach could be used to generate a spatially explicit canopy map that contains biomass, individuals and leaf age distribution. Nevertheless, reading an observed lai, i.e., from MODIS, and using it to force ORCHIDEE-CN-CAN would require a substantial number of assumptions to turn an aggregated 1D lai value into a disaggregated 3D canopy at the PFT level.
 
 === Land cover change (with age classes) ===