Evaluation and improvement of CERES-Maize evapotranspiration simulations under full and limited irrigation treatments in northern Colorado
Date
2011
Authors
DeJonge, Kendall, author
Garcia, Luis, advisor
Andales, Allan, advisor
Arabi, Mazdak, committee member
Hansen, Neil C., committee member
Ascough, Jim, II, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Population growth in urbanizing areas such as the Front Range of Colorado has led to increased pressure to transfer water from agriculture to municipalities. In some cases, farmers may remain agriculturally productive while practicing "limited or deficit irrigation," where substantial yields may be obtained with reduced water applications during the non-water sensitive growth stages. Savings in crop evapotranspiration (ET) could then be leased to municipalities or other entities as desired. This dissertation examined the benefit of limited irrigation in comparison with full irrigation in the northern Front Range of Colorado, in both an on-field context and in a crop modeling context. Because of Colorado water law the quantification of ET is especially important, as ET is considered a consumptive use of the crop and therefore can be transferred between entities. The overall goal was to improve understanding of both field and model maize yield response to limited water supplies, accurately simulate this management scenario with a crop growth model, and evaluate and improve model simulation of ET. This goal was achieved in the context of three studies that are included in this dissertation. First, the CERES-Maize corn growth model was calibrated and evaluated for full and limited irrigation of corn; the model generally simulated many aspects of crop growth, including yield, leaf area index, leaf growth, and phenology. The model performed better overall for the full irrigation treatment than the limited irrigation treatment. The model underestimated treatment differences in cumulative ET between treatments, simulating too little ET under full irrigation (-7.2% relative error over three years) and too much ET under limited irrigation (12.7% relative error over the same three years). Second, a global sensitivity analysis was performed on genotype and soil hydraulic parameters in addition to radiation use efficiency, using full and limited irrigation treatments and two global sensitivity analysis methods (Morris and Sobol'). Outputs evaluated included phenological stages, leaf growth, leaf area index, yield, and ET. The model showed similar sensitivities between treatments in regard to phenology and leaf expansion. However, leaf area index, yield, and ET were primarily sensitive to genotype parameters under full irrigation, but under limited irrigation showed increased sensitivity to soil hydraulic parameters. Results from both sensitivity analysis methods were highly correlated. Finally, the model processes that govern major aspects of the water balance, particularly the calculation of potential ET and the partitioning of this value into potential soil evaporation and potential plant transpiration, were evaluated and improved. A new equation for the crop coefficient as a function of LAI was added to better represent the ET demand based on plant canopy. This new formula, as well as a new coefficient determining ET partitioning, were evaluated using five years of management and weather data, and both irrigation treatments. Under full irrigation simulated yield was unchanged compared to previous simulations, while ET and water use efficiency (WUE) were simulated closer to observations. Under limited irrigation, ET and WUE simulations were improved, with RMSD reduced from 80.9 mm to 49.9 mm for ET and from 5.97 kg/ha-mm to 2.86 kg/ha-mm for WUE. While this model change is a significant improvement in regard to the estimation of ET under water stress, it is further recommended that future model changes attempt to incorporate physiological response to stress, such as stomatal conductance or canopy temperature, to better represent plant response to water stress.
Description
Rights Access
Subject
evapotranspiration
CERES-Maize
global sensitivity analysis
irrigation
modeling
water stress