SIMPLECrop Modeling for rice: Calibration, Validation, Evaluation and Sensitivity analysis in An Giang Provine, Vietnam

Abstract

Rice, a staple crop of global significance, is instrumental in securing food resources and fostering socio-economic well-being. Maintaining the stability and resilience of rice production is paramount in the context of a changing climate and increasing population. Advanced crop modeling techniques have emerged as indispensable tools for understanding and predicting crop responses to varying environmental conditions. This study, conducted in Vietnam's An Giang Province in the Mekong Delta, addresses the critical need for robust rice production models. The region's susceptibility to climate change-induced temperature and rainfall shifts underscores the urgency of developing precise predictive models for rice cultivation. Employing the versatile SIMPLECrop model, we endeavor to address the tests of calibrating, validating, evaluating, and conducting sensitivity analysis for rice cultivation in this specific locale. Calibration, the process of aligning model parameters with observed data, involved iterative adjustments using optimization techniques. The validation process compared model predictions with independent observed data. Our evaluation demonstrated relative root-mean-square error (RRMSE), Nash-Sutcliffe efficiency (NSE) values ranging from 4.2% to 5.5% and 0.87 to 0.90 for open field conditions. During sensitivity analysis, two parameters, RUE and T_base, consistently exhibited the highest influence within the model. In the field condition model assessment, the SIMPLECrop model excelled in predicting crop yields in open-field settings. The RRMSE values fell within the accepted range for crop modeling, and the NSE coefficient indicated a strong fit performance, ensuring reliable predictions of rice yields across seasons.