Some users report questions about how to perform the model calibration, notably regarding phenology, SLA and biomass.
Phenology
For phenological stages, the model currently uses 5 phenology parameters : SDJLevee, SDJBVP, SDJRPR, SDJMatu1 and SDJMatu2 (defined in /data/params/variety*.yaml). Users ask how to convert phenological observation dates into phenological parameters. In particular, they only report dates for emergence (levée), flowering (floraison), sowing (semis), and harvest (récolte). From these dates, thermal times have to be calculated, using the temperature base (Tbase) adapted to the considered crop species. We can then use the following reference timeline for clearer understanding of how to compute thermal time:
For SLA, users ask how they can use SLA measurements they have done to calibrate the model. There seems to be a confusion regarding the parameters SlaMin, SlaMax, and SlaPente, which influence the SLA simulation and should not be replaced directly with measured values. Instead, the sla variable (which is an intermediate variable computed daily in the model), should be compared to measured values to adjust SlaMin, SlaMax, and SlaPente parameters.
The calculation process of the sla variable is documented in the calculate_canopy_specific_leaf_area() function in bilan_carbo.py. The function description includes the conditions under which SLA is calculated, adjusted, and bounded by SlaMinand SlaMax. The documentation provides a detailed calculation method assuming young leaves have a higher SLA than old leaves, with specific conditions and parameters influencing SLA.
Biomass
As for SLA, users struggle to see how they can use biomass measurements for model calibration. Overall, the WOFOST documentation (WOFOST calibration guide : A Gentle Introduction to WOFOST, Annex 2) provides a useful calibration guide that is also valid in its logic for SARRA-Py, suggesting the order of parameter adjustment: start with phenology, then SLA, and finally biomass parameters ultimately impacting yield.
Hi,
Some users report questions about how to perform the model calibration, notably regarding phenology, SLA and biomass.
Phenology
For phenological stages, the model currently uses 5 phenology parameters :
SDJLevee,SDJBVP,SDJRPR,SDJMatu1andSDJMatu2(defined in/data/params/variety*.yaml). Users ask how to convert phenological observation dates into phenological parameters. In particular, they only report dates for emergence (levée), flowering (floraison), sowing (semis), and harvest (récolte). From these dates, thermal times have to be calculated, using the temperature base (Tbase) adapted to the considered crop species. We can then use the following reference timeline for clearer understanding of how to compute thermal time:SLA
For SLA, users ask how they can use SLA measurements they have done to calibrate the model. There seems to be a confusion regarding the parameters
SlaMin,SlaMax, andSlaPente, which influence the SLA simulation and should not be replaced directly with measured values. Instead, theslavariable (which is an intermediate variable computed daily in the model), should be compared to measured values to adjustSlaMin,SlaMax, andSlaPenteparameters.The calculation process of the
slavariable is documented in thecalculate_canopy_specific_leaf_area()function inbilan_carbo.py. The function description includes the conditions under which SLA is calculated, adjusted, and bounded bySlaMinandSlaMax. The documentation provides a detailed calculation method assuming young leaves have a higher SLA than old leaves, with specific conditions and parameters influencing SLA.Biomass
As for SLA, users struggle to see how they can use biomass measurements for model calibration. Overall, the WOFOST documentation (WOFOST calibration guide : A Gentle Introduction to WOFOST, Annex 2) provides a useful calibration guide that is also valid in its logic for SARRA-Py, suggesting the order of parameter adjustment: start with phenology, then SLA, and finally biomass parameters ultimately impacting yield.