Kernel Function Model for Planning of Agricultural Groundwater Development

Abstract

A linked kernel-optimization model for the planning of optimal groundwater development for irrigation is presented. The planning ensures optimization of zonal crop patterns subject to the constraints on the maximum water table depth and the stream-aquifer interflow at the dynamic equilibrium. The model is computationally inexpensive as compared to the traditional linked simulation-optimization models. Its use is demonstrated by applying it to a canal command area in India. Five kernel models are developed relating the maximum water table depth and four critical stream-aquifer interflow rates to the crop areas. The necessary data base is generated by using a physically based precalibrated simulation model of groundwater flow. The kernel models are linked to a genetic algorithm-based optimizer for arriving at the optimal cropping pattern and the associated pumping pattern. The near-optimal solution so obtained is further