The average productivity of cotton, sorghum and soybean in Maharashtra is considerably lower than their potential. There is also a large temporal and spatial variability in their productivity due to the spatial distribution of soils and its interaction with the rainfall pattern. WOFOST model-version 7.1 was validated and used for quantification of yield gaps under different rainfall patterns for cotton, sorghum and soybean on five soil series of Maharashtra. The validation results indicate that the model performed well with RMSE less than 20% and simulated the yields with Model efficiency (ME) values 0.73, 0.88 and 0.89 for cotton, sorghum and soybean, respectively. Between soil series, the variability in the mean simulated yield among years experiencing normal rainfall was higher for cotton (CV=53.9%) than for sorghum (CV=27.6%) or soybean (CY=20.1%) as soil parameters significantly affected the cotton yields. There was significant correlation between simulated yield with soil depth (r=0.91) and extractable soil moisture (r=0.96) for cotton; but not for sorghum and soybean. It is concluded that the WOFOST model could capture the effects of spatial distribution of soil and rainfall pattern on the yields of cotton, sorghum and soybean and can compliment other techniques in suggesting alternative crop options for aberrant rainfall situations.
