Calibration and Validation of CSM‐CROPGRO‐Cotton Model Using Lysimeter Data in the Texas High Plains

Abstract

AbstractTexas High Plains (THP), one of the most important food and fiber producing regions in the Ogallala Aquifer Region, currently faces rapid decline of groundwater levels. Predicted climate extremes and high temporal variability in growing season precipitation may require growers to pump more groundwater from the Ogallala Aquifer to meet higher crop water demand. The Decision Support System for Agrotechnology Transfer (DSSAT) Cropping System Model (CSM) is a widely used crop simulation tool for evaluating impacts of different water and crop management practices, including irrigation on crop yield and water use efficiency. In this study, CROPGRO‐Cotton module of the DSSAT was calibrated and validated using 2000, 2001, 2002, and 2010 irrigated lysimeter field data managed by the USDA‐ARS (United States Department of Agriculture ‐ Agricultural Research Service) Conservation and Production Research Laboratory at Bushland, TX. The lysimeter field consisted of four equal plots designated as NE, SE, NW, and SW. Crop growth characteristics including leaf area index (LAI), above ground biomass (AGB), evapotranspiration (ET), soil moisture, and lint yield of 2000‐NE, 2000‐SE, and 2001‐NE, were used for calibration and 2002‐NE, 2010‐NE, and 2010‐SE were used for validation. The calibrated and validated model was used to simulate the long term (1924–2012) crop yield and seasonal crop ET. During the calibration process, some of the cultivar and ecotype parameters that influence LAI, AGB, and lint yield were adjusted for better statistical results. Measured and simulated LAI, AGB, ET, soil moisture, and lint yield showed good agreement during calibration and validation as indicated by performance statistics such as r2 from 0.70 to 0.82, and percent error (PE) = −0.85 to 17.3% for LAI; r2 = 0.89 to 0.95, and PE = −7.36 to −13.66% for AGB; and r2= 0.90 to 0.94, and PE = 3.20 to 3.44% for ET during calibration and validation, respectively. The model underestimated ET during peak vegetative growth and development stage except in some circumstances. The calibrated and validated model was able to simulate lint yield and seasonal ET during a long term (1924–2012) historic period for Bushland, TX, under irrigated conditions. The calibrated model could be used to schedule ET based irrigation management practices in the THP and to estimate future ET for other modeling experiments.