Validation of the SUBSTOR Model for Simulating Soil Water Content

Abstract

Crop growth models are increasingly important for the evaluation of agricultural management practices. These models rely on accurate simulation of soil water content. This study was conducted to compare soil water contents measured, in 1992 and 1993, on 20 sites cropped to potatoes, and simulated by the SUBSTOR model. A sensitivity analysis was performed to identify the input parameters of SUBSTOR which most influence the water content variations. Sensitivity analysis results showed that water retention parameters (saturation and field capacity) have an important effect on most hydrological processes and soil water content. Water retention parameters were estimated by published relationships. Simulated soil water contents were in close agreement to measured values with a mean absolute error between measured and predicted soil water contents ranging from 2.7 to 4.6%. On the average, simulated water contents ranged from 3.3% below measured values to 4.6% above. Soil parameters estimated with the published relationships were in the same order of magnitude as optimized parameters. Generally, the model overestimated measured water contents, due principally to an underestimation of evapotranspiration, particularly in the middle of the growing season.