Use of Multiple Environment Variety Trials Data to Simulate Maize Yields in the Ogallala Aquifer Region: A Two Model Approach

With a long-term goal to optimize use of groundwater in the Ogallala Aquifer Region (OAR) to sustain food production systems, this study was conducted to calibrate Decision Support System for Agrotechnology Transfer (DSSAT) and AquaCrop crop modeling platforms to simulate maize production at a regional scale using historic datasets. Calibration of the models with local crop growth data and crop management practices is important, but usually this in-season crop growth information is not available. This study determined the possibility of using maize variety trial data for the evaluation of the CSM-Crop Estimation through Resources and Environmental Synthesis-Maize and AquaCrop models in the OAR. The models were calibrated and tested in three counties in Nebraska. Both the models were then used to simulate irrigated maize yield during 1988 to 2015 for all three counties. The criteria for evaluating the performance of these crop models included statistical parameters and graphical analysis. The performance of both models were then compared with the observed yield from field variety test results and historic National Agricultural Statistical Service yields. The results indicated that difference between yield of calibrated DSSAT model and observed yield was less than 10% and AquaCrop root mean square error ranged from 740 to 1,820 kg/ha. Long-term comparison between observed and simulated Nebraska county yields also indicated confidence in calibrating crop models with typical end of season yield data and using these models for studying crop production at regional scales when detailed in-season crop growth observed data are not available.