Modelling and Assessment of Climate Change Impact on Rainfed Rice Cultivation in a Sub-humid Subtropical Region

Changing climate is one of the main factors affecting to decline food production because crop growth is sensitive to climatic changes. In this study, DSSAT and APSIM models were calibrated and validated at 15 different locations of the Kangsabati river basin and compared based on three output parameters for three famous variety, viz. Swarna, Lalat, and MTU 1010 to identify which one was the most accurate in reproducing rainfed rice yield in the research location. The results suggest that DSSAT model can simulate rice crop growth and yield better than APSIM model in this region. In terms of rice yield, Swarna, a long-duration rice variety, had a greater RMSE (144.63 kg ha⁻¹) than the short-duration rice variety Lalat and MTU 1010. However, the DSSAT model performed better than the APSIM model in modeling rainfed rice crop growth and yield (r² > 0.86 and greater ME). Bias-corrected RCM (HadGEM3-RA) have been linked to the DSSAT model to analyze yield changes of three popular rice varieties (Swarna, Lalat, and MTU 1010) for future climate scenario (2030s & 2040s). The average decrease in rice yield of Swarna is projected to be around 7% in 2030s and 9% in 2040s, and under RCP 8.5, average decrease in rice yield of Swarna is projected to be 12.8% (2030s) and 15.4% (2040s), which is much higher than production of Lalat and MTU 1010 in both the cases. In conclusion, DSSAT might be a helpful tool for determining the effects of management techniques and climate change on various rice varieties.