U.S. Department of Agriculture: Agricultural Research Service, Lincoln, Nebraska
Document Type
Article
Date of this Version
2017
Citation
A.A. Pinto and D. Zilberman (eds.), Modeling, Dynamics, Optimization and Bioeconomics II, Springer Proceedings in Mathematics & Statistics 195, DOI 10.1007/978-3-319-55236-1_16
Abstract
U.S. agriculture has made impressive strides over the past 50 years in crop yield and input productivity growth, especially since the advent of genetically modified crops in 1996. However, future growth rates could decline if U.S. agriculture does not sufficiently adapt to climate change. We examine the magnitudes of weather impacts on U.S. corn yields during 1960–2011—with a focus on intense precipitation and nitrogen use efficiency—and use the empirical results to forecast yields for the subsequent 20 years (2012-2031). We improve upon past methodologies by employing dynamic Bayesian regressions. These dynamic models permit rapid updating of new information, consistent with both pronounced yield growth in recent years and agricultural adaptation to changing growing conditions. We find that corn yields will increase by 27–41% over 2011 yields in top-growing states, though yields will gradually decline in less-productive states where climate change impacts could be among the most harmful. Our forecasts are generally robust to the empirical specification and assumptions about the econometric disturbance term, and have similar out-of-sample performance. To the extent that increasingly intense rainfall could contribute to nitrogen and other nutrient leaching, farmers may need to adjust nutrient applications in response to changing production environments.
Comments
U.S. government work.