Indian Journal of Agronomy

CropSyst simulation model was tested during the rainy seasons of 2007 and 2008 and the winter seasons of 2007–08 and 2008–09 at New Delhi, to study the long and wide-spread adoptability of pearlmillet [Pennisetum glaucum (L.) R. Br. emend. Stuntz.]-chickpea (Cicer arietinum L.) cropping system by comparison of simulated and observed variables. The observed variables collected from the experimental data were used to study the effect of nitrogen (N) levels in pearlmillet and water levels in chickpea for pearlmillet-chickpea crop rotation. The calibration, validation and sensitivity analysis of CropSyst model was utilized to quantify and verify the interactive effect of different water and N treatments on the productivity of pearlmillet-chickpea crop rotation using measurements from field experiments. Results showed that CropSyst model performed well at lower levels of N treatments (60 kg/ha) whereas at higher levels of N treatments (90 kg/ha) the predicted values were lower than the observed values. The model also responded well to limited level of irrigation in pearlmillet, but in chickpea irrigation did not have a significant influence on biomass yield as predicted by the model. The model was tested for accuracy in determination of the crop parameters by conducting sensitivity analysis of the model, which depicted that crop parameters, viz. light to above-ground biomass conversion, specific leaf area, phenological degree-days, base temperature, stem/leaf partition coefficient needs more accuracy during its calibration and validation. Also, the root mean square error (RMSE) for biomass and grain yield was found to be 1.97 and 0.36 t/ha, being 5 and 7% of the observed mean, respectively, in pearlmillet whereas for chickpea it was 0.40 and 0.17 t/ha which, in turn, was 8 and 6% respectively. These low values of RMSE indicates that the CropSyst model is highly accurate in predicting grain yield and above-ground biomass of pearlmillet-chickpea crop rotation.