Responsiveness of rice yield to climate change in China: historical analysis and future projections

Understanding the meteorological driving factors of changes in rice production is critical to global food security. In this study, we apply multivariate adaptive regression spline algorithm to examine the effect of climatic variables on rice yield in China. We find that the importance of temperature is much higher than that of precipitation. The hinge function shows significant nonlinear and interaction effects in yield response function. Based on the projected climate change scenarios of lower rate global warming (SSP1-2.6) and higher rate global warming (SSP5-8.5), we further predict the rice yield, and find that the yield will increase by 0.09% and 0.28% under both scenarios in the medium term. However, the continuous climate change in the long-term will lead to yield loss (0.39%) in SSP1-2.6, and a higher loss (7.39%) in SSP5-8.5. The difference in predicted rice yield in 15 provinces also indicates a heterogeneous impact of climate change.