Multi-objective groundwater management using genetic algorithms in Kerbala desert area, Iraq

Abstract

Kerbala desert is one of the most important areas in Iraq, due to the large number of groundwater aquifers in the region. Based on this, many agricultural investment projects have been established in this area recently, with multiple new wells being drilled in the desert. All such wells penetrate the Dammam confined aquifer to a depth of 260 to 300 m, and uncontrolled use of these wells, and the additional wells expected as the area develops, is likely to lead to problems with groundwater sources, including depletion of water levels, well interference, and groundwater pollution. This study introduces a multi objectives groundwater management model based on an integrated simulation-optimisation (S/O) model. The simulation of groundwater flow was developed using a finite-difference numerical model, MODFLOW, under GMS software for various conditions, and evaluation of calibration of the transient model was determined by observing the measured and calculated aquifer heads. A multi-objective optimisation model was applied to maximise the pumping rate and minimise pumping costs, and this was solved using a genetic algorithm method supported by two further mechanisms, Pareto optimality ranking and fitness sharing. A set of Pareto optimal solutions as determined in the final generation was thus created for the multi-objective function (maximum pumping rate and minimum pumping cost) to help decision-makers (DM). Based on the Pareto solutions set, a DM or the designer may thus choose a preferred solution, or a compromise solution may be derived by considering the full set of Pareto optimal solutions.