Corn Water Use Modeling

Regional Corn Modeling at a High Resolution Scale: A Yield Based Approach and Blue vs Green Water Assessment

Agricultural water use currently accounts for as much as 70 percent of total water use.  Couple this with an expected 50 percent increase in global water resource demand in the next 40 years and a potential shift in rainfall patterns associated with climate change; one can begin to see the major challenges ahead for the current generations.  Predicting the future agricultural water demands patterns is essential to ensure a sustainable world for a growing human population.  We developed a crop water demand simulation process incorporating the CERES-Maize model in the Decisions Support System for Agrotechnology Transfer (DSSAT) Cropping System Model (CSM) program, version 4.0, the MATrix LABoratory program (MATLAB), and regional comprehensive datasets including the NASA Agroclimatology Archive, the International Soil Reference and Information Centre (ISRIC) World Inventory of Soil Emission Potentials (WISE) soils database, the Center for Sustainability and the Global Environment (SAGE) Harvest Area and Yields of 175 crops and Crop Calendar database, and the United States Department of Agriculture (USDA) National Agricultural Statistics Service (NASS).

Phase 1 of the project involved a preliminary a calibration of the CERES-Maize model to the largest corn production regions in the U.S., USDA Economic Research Service (ERS) Region 1, also known as the Heartland Region.  The calibration procedure exhibited relative success at modeling at the regional scale, obtaining a R2 value of 0.8.  However, the results were not replicated in the validation step, producing a R2 value of 0.07.  In addition, global evapotranspiration was estimated for the entire globe.  The process began by obtain water use efficiencies (WUE) from the current scientific literature.  The WUE values were then applied to global corn yields to calculate the evapotranspiration of corn.  Finally, based on the equation Q = P – ET – RO, blue and green water was calculated for the globe.  The areas where ET was greater than the difference between precipitation and runoff were designated blue water areas, while the areas where ET was less than the difference were designated as green water areas.  Given the poor predictive ability of the model produced during Phase 1, the calibration strategy was revised for the regional crop modeling procedure.  Maize production ERS region 1 was re-evaluated during Phase 2 and a temporal dimension was added to the process.

Phase 2 focused on derived a grid specific definition of maize genetic coefficients to act as predictor of maize yields.  A preliminary sensitivity analysis was conducted to determine the optimal starting ranges for calibration process.  Coding for the calibration procedure is currently underway and new results are expected soon.

The Sustainability Consortium helped fund this project.

Contact:
Marty Matlock, 479-575-2849, mmatlock@uark.edu