Decision support for agricultural water management

Abstract

Irrigation water use is the major pressure limiting the availability of fresh water resources in the
 Mediterranean. Efficient irrigation scheduling programs (IRSPs) are able to reduce water
 consumption; however, their selection and placement in large agricultural landscapes depend on
 location specific characteristics and economic indicators. Towards this end, a novel and efficient
 Decision Support Tool (DST) is developed in MATLAB-programming, able to assess the
 effectiveness of different IRSPs in reducing total agricultural water use at the catchment scale along
 with their impact on crop yields. The DST integrates a look-up table with data on irrigation water
 amounts and crop yields at different locations within a catchment, populated by a hydrological and
 crop growth estimator: the process-based SWAT model, into a multi-objective Genetic Algorithm,
 which serves as the optimization engine for the allocation of measures across the agricultural land.
 The optimization scheme leads rapidly to the optimal trade-off frontier between the conflicting
 objectives providing spatial allocations of IRSPs. The tool was implemented in the Ali Efenti
 catchment demonstrating optimal solutions that could save more than 10% of water by reducing
 cotton yields less than 5% from the baseline. The study highlights the potential of the tool to assist in
 the development of cost-effective water saving plans at the catchment level in order to reduce the
 risk of desertification in intensively cultivated areas.