IRRIGATED CROP YIELDS REVEAL DISTINCT INFLUENCES OF HEAT AND MOISTURE

Abstract

Climate change presents a growing challenge to global food systems in addition to challenges to human and natural ecosystem health. While statistical crop models offer insights into how rising temperatures affect crop yield, a complete understanding of projected impacts has been limited by often-simplified assumptions about the independence of temperature and water stress. Here, we comprehensively analyze the covariance between commonly used predictors in statistical crop models. Using paired rainfed and irrigated yield observations across the US, we are able to control for the effects of water availability and independently evaluate the influence of temperature on yields. We find that temperature-based predictors often absorb the effects of water stress, inflating their apparent importance in statistical yield models. These findings have implications for the magnitude of projected yield damages and advance our understanding of the agricultural consequences of climate change.