Quantifying the Impacts of Early and Late Growing Season Precipitation on Midwestern US Corn Production: A Downscaling and Modeling Approach

Water availability and accessibility is one of the primary drivers of crop growth and production. While corn (Zea maize L.) is particularly sensitive to lengthy dry spells, especially during the reproductive stages, on-farm decision making and flexible management strategies can help reduce risk and mitigate some of the negative effects during dry times. Using 2012 as a recent example of significant drought in the Midwestern United States, in this study, we simulate corn production at seven locations across the Corn Belt under multiple management strategies. In order to do so, we first implement a modified version of a non-parametric resampling method, FResampler, to downscale tercile-based probabilistic seasonal climate forecasts for the early (April, May, June) and late (July, August, September) growing season through a resampling of historical weather series. The seasonal climate forecasts are based on three distinct scenarios: a dry, near-normal, and wet forecast. Outside of the downscaled months, the observed weather for 2012 was used. The downscaled weather series, based on the forecasts, were then used as inputs for a process-based crop simulation, Decision Support System for Agrotechnology Transfer v4.7. This approach allows us to evaluate the effects of different management strategies under two contexts. First, in a decision-making capacity, the early season downscaling allows for identification of best management strategies based on information that would be available to a producer at the beginning of a growing season, before the crop has been planted and fertilizers applied, in the case of a late season drought. Second, the late season downscaling, where the early growing season is known, allows us to evaluate which management strategies provide the best outcomes given an uncertain late growing season in a hindcast context. Past research and preliminary results indicate that FResampler, in combination with crop models, displays substantial utility in evaluating the effects of variable management on corn production considering weather uncertainty and variability.