Skip to main content
ARS Home » Plains Area » Fort Collins, Colorado » Center for Agricultural Resources Research » Rangeland Resources & Systems Research » Research » Publications at this Location » Publication #356391
Title: Evaluating RZWQM2-CERES-Maize and water production functions for predicting irrigated maize yield and biomass in eastern Colorado

Author
item SIMA, NATHAN - Colorado State University
item ANDALES, ALLAN - Colorado State University
item Harmel, Daren
item Ma, Liwang
item Trout, Thomas

Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/15/2018
Publication Date: 1/2/2019
Citation: Sima, N.Q., Andales, A.A., Harmel, R.D., Ma, L., Trout, T.J. 2019. Evaluating RZWQM2-CERES-Maize and water production functions for predicting irrigated maize yield and biomass in eastern Colorado. Transactions of the ASABE. 62(1):1-11. https://doi.org/10.13031/trans.13045.
DOI: https://doi.org/10.13031/trans.13045

Interpretive Summary: Complex models have been developed to understand the interactions among cropping system components. Although these models are widely published in literature, their use to guide irrigation management is sparse. In this study, the CERES-Maize crop model in Root Zone Water Quality Model was compared to a crop water production function for its ability to predict biomass and grain yield of irrigated maize in eastern Colorado. Results showed that the water production function in general performed better than the CERES-Maize crop model after considering both experimental and simulation uncertainties. Thus, for practical purposes, water production functions may be more favorable than complex crop models because of their simplicity and no need for calibration. However, complex models are still useful to simulate the behavior of system variables that are not measured experimentally and to predict the interactions among system components. This study also revealed that a complex crop model could have compensated its deficiency in simulating management effects by fitting variety-specific parameters. As a result, the parameters are site- or data-specific and may vary from location to location, which is exactly what crop modelers try to avoid by developing complex mechanistic models instead of simplified relations.

Technical Abstract: Complex crop models have been developed to understand the interactions among system components. Although these models are widely published in literature, their use to guide irrigation management is sparse. In this study, the CERES-Maize crop model in Root Zone Water Quality Model was compared to an empirical regression-type crop water production function for its ability to predict biomass and grain yield of irrigated maize in eastern Colorado. Results showed that the water production function in general performed better than the CERES-Maize crop model based on a modified F-test and d-index after considering both experimental and simulation uncertainties. Thus, for practical purposes, water production functions may be more favorable than mechanistic crop models because of their simplicity and no need for calibration. However, mechanistic models are still useful to simulate the behavior of system variables that are not measured experimentally and to predict the interactions among system components. This study also revealed that a mechanistic crop model could have compensated its deficiency in simulating management effects by fitting cultivar parameters. As a result, the fitted cultivar parameters are site- or data-specific and may vary from location to location, which is exactly what crop modelers try to avoid by developing mechanistic models instead of regression equations.