Developing an eco-physiological process-based model of soybean growth and yield (MATCRO-Soy v.1): model calibration and evaluation

Abstract

MATCRO-Soy is an eco-physiological process-based crop model for soybean (Glycine max L. (Merr.)). It was developed by modifying the parameters of MATCRO-Rice, which integrates crop growth processes with a land surface model. The original model was modified using data from the literature and field experiments conducted in countries around the world. The reliability of the model was extensively validated by comparing the simulated yields with observed yields at global, national, and grid-cell levels. Moderate correlations were detected between the yields predicted by MATCRO-Soy and yield data in the Food and Agriculture Organization's FAOSTAT database, with correlation coefficients of 0.81 (p<0.001) for the global average yield and 0.45 (p<0.01) for the global average detrended yield over a 34 year period (1981-2014). Furthermore, validation at the grid-cell level revealed a statistically significant correlation between the MATCRO-Soy simulated yield and the observed yield in 66 % of the grid cells in the global yield map. These results highlight the model's ability to reproduce soybean yield under different environmental conditions, integrating soil water availability and nitrogen fertilizer levels. The MATCRO-Soy model may enhance our understanding of crop physiology, especially crop responses to climate change. Its application may support efforts to reduce uncertainty in projections of the effects of climate change on soybean crops.