Multivariate logarithmic modeling of grain production in the Yangtze River Basin: incorporating extreme weather factors.

Abstract

Abstract Whilst essential to the nutrition of societies, grain crops are demonstrated to be largely susceptible to the influence of anthropological climate change and extreme weather. However, few previous attempts at modeling grain yield took thorough consideration about the potential impact of extreme temperature events (ETEs) on average (or per-hectare) grain yield. From historical data in a Chinese agriculture hub, namely the Middle-Lower Yangtze Plains (MLYP) region, through a 2-step, nested OLS-FGLS multivariate log-log regression model, this study underscored the strong, sustained and significant negative influence ETEs had on grain production in the last 32 years in MLYP provinces of Jiangsu, Zhejiang, Anhui, Jiangxi, Hubei, and Hunan; supported the literature with further evidence of global warming reducing crop productivity; and corroborated previous studies highlighting a reduction in crop productivity sourced from inefficient distribution and management of labor in the context of technological advancements. Climate-model-based provincial predictions through Shared Socioeconomic Pathways (SSPs) indicate a strong need for agricultural workers and scientists to address the increasing threat of future heat and cold stress through both micro-level (such as genomics-assisted breeding) and macro-level (such as AI-mediated farm management tools), in order for them to be prepared for a wide range of climate change scenarios.