Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Parameter Estimation, Transferability Evaluation and Predictive Uncertainty Analysis of a Sugarbeet Model Using PEST

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2018 ASABE Annual International Meeting  1800231.(doi:10.13031/aim.201800231)
Authors:   Zhulu Lin, Mohammad J Anar
Keywords:   Biofuel, Crop and Environment REsource Synthesis (CERES), Decision Support System for Agrotechnology Transfer (DSSAT), Parameter Estimation (PEST), Predictive uncertainty analysis, Sugarbeet.

Abstract. Sugarbeet (Beta vulgaris) is considered as one of the most viable feedstock alternatives to corn for biofuel production after herbicide resistant sugarbeet was deregulated by United States Department of Agriculture in 2012. Although many sugarbeet models have been found in the literature, most models are restricted to the regions and conditions for which they were developed and require different file and data structures. The Decision Support System for Agrotechnology Transfer (DSSAT) developed by an international network of scientists provides a common framework for a cropping system study. It currently has plant growth modules for more than 40 crops but does not include sugarbeet. In a recent study, Crop and Environment REsource Synthesis (CERES) Beet model was modified and incorporated into the DSSAT to model sugarbeet growth, titled “CSM-CERES-Beet”. In this study, PEST was used for parameter estimation, transferability evaluation, and predictive uncertainty analysis of CSM-CERES-Beet. The model was evaluated against two sets of plant growth data collected for different sugarbeet varieties grown in two different regions and under different conditions – one in Romania (Europe) during 1997-1998 and the other in North Dakota (USA) during 2014-2016. After model calibration for specific cultivars, the CSM-CERES-Beet model performed well for the simulation of leaf area index, leaf number, leaf or top weight, and root weight for both datasets (d-statistic = 0.783-0.993, rRMSE = 0.127-1.014). However, the uncertainty analysis revealed that the calibrated CSM-CERES-Beet consistently over-predicted leaf numbers with false confidence. CSM-CERES-Beet could be applied for predicting sugarbeet yield for different soil and climatic conditions and various management scenarios for the Red River Valley in the US and other regions with environmental conditions favorable for sugarbeet production.

(Download PDF)    (Export to EndNotes)