Estimating the Spatially Varying Responses of Corn Yields toWeather Variations using GeographicallyWeighted Panel Regression

Researchers have extensively studied crop yield response to weather variations, while only a limited number of studies have attempted to identify spatial heterogeneity in this relationship. We explore spatial heterogeneity in corn yield response to weather by combining geographically weighted regression and panel regression. We find that temperature tends to have negative effects on U.S. corn yields in warmer regions and positive effects in cooler regions, with spatial heterogeneity at a fine scale. The spatial pattern of precipitation effects is more complicated. A further analysis shows that precipitation effects are sensitive to the existence of irrigation systems.