Assessing the Impact of Climatic Variability and Change on Maize Production in the Midwestern USA with Crop Simulation Models.

Weather and climate remain among the most important uncontrollable factors in agricultural production systems. In this study, two process-based dynamic crop simulation models were used to identify the impacts of climate on the production of maize in the Midwestern U.S.A. during the historical (1981-2012) and future (2041-2070) time periods. The 12-state region is a key global corn production area, responsible for more than 80% of U.S. domestic and 25% of total global production. The study is a part of the Useful to Useable (U2U) Project, a USDA NIFA-sponsored project seeking to improve the resilience and profitability of farming operations in the region amid climate variability and change.

The crop simulation models used in the study were CERES-Maize embedded in DSSATv4.5 (Hoogenboom et al., 2012) and the Hybrid-Maize model (Yang et al., 2004). Model validation was carried out with individual plot and county observations. The models were run with gridded weather data for representative soils and cultivars for historical and projected future time frames to examine spatial and temporal yield variability within the region. For the historical period, these comparisons included observed 18 sites, LIS 4km, and NARR 32 km gridded data, and gridded output from CRCM, HRM3, MM5, and WRFG regional climate models for future.  We also ran simulations for model-derived current contemporary (1971-2000) and future (2041-2070) periods at single sites across the domain with the delta method applied to the historical gridded input data. In all simulations, all input variables except climate were held at constant levels representative of current levels of technology to isolate the influence of weather and climate. Finally, the DSSAT model is being used to simulate the impacts of different management strategies (variation of planting date, fertilizer applications, tillage, and irrigation) during historical and projected future time frames.