Estimation of non-linear root water uptake parameter using genetic algorithms

Abstract

The present study is concerned with the estimation of non-linear root water uptake parameter using Genetic Algorithm (GA) technique. Various models have been proposed to predict root water uptake. However, from the studies and observations it has been found that the nature of root water uptake is non-linear. The non-linear root water uptake model termed as O–R model coupled with soil moisture flow equation is developed to predict the root water uptake pattern. The O–R model incorporates a parameter ‘β’ to account for the non-linearity in root water uptake. In the present study, parameter β is estimated through inverse modelling using GA optimization technique. The parameter is optimized by minimizing the difference between model predicted and experimentally observed percentage soil moisture depletion in the root zone. To check the efficacy of the developed model, the optimization procedure is validated from hypothetically generated percentage soil moisture depletion corresponding to an assumed β value. The model is applied to wheat crop (Triticum) for estimation of non-linear root water uptake parameter. The results show that linked simulation optimization model based on GA method accurately determines the non-linear root water uptake parameter for a given crop.