Comparison of two soya bean simulation models under climate change : I Model calibration and sensitivity analyses

To analyse the effects of climate change on soya bean growth and production, both a simple growth model, SOYBEANW, and a comprehensive model, CROPGRO, have been applied. Both models were calibrated and tested against results from soya bean trials at Toulouse, France. The sensitivity of model results to changed values of weather variables was determined. The comparison of the results from both models indicated the sort of conditions in which model results differed and may become less reliable. The start date of seed filling at Toulouse was predicted well by both models, but the simulated duration of seed filling from both models was too short in most years. Irrigated seed production was calculated reasonably well by both models, although in some years simulated yields were rather high. Seed yields from SOYBEANW were strongly dependent on the soil nitrogen supply, whereas in CROPGRO this dependence was almost nil. Without irrigation, the yield reduction due to water shortage in dry years as simulated with CROPGRO was stronger than the observed yield reduction. For SOYBEANW the opposite applied (i.e. too small reduction). Irrigated seed yields from both models increased with both increasing solar radiation and atmospheric CO2. The optimum for irrigated production from SOYBEANW was at present temperatures in Toulouse and from CROPGRO at 2 to 4°C higher temperatures. Water-limited seed yields from both models were almost insensitive to temperature, increased with an increase in atmospheric CO2 and precipitation, and decreased with increasing solar radiation for CROPGRO but did not change for SOYBEANW