Identification of the optimum groundwater quality monitoring network using a genetic algorithm based optimization approach

Management of groundwater requires a sufficient coverage of accurate groundwater quality data. These data are usually collected from monitoring wells which are spatially distributed in the river basin or the groundwater body that is studied. A minimum number of monitoring wells with an optimum spatial distribution is desired to ensure a cost-effective observation of the groundwater body. Therefore, the configuration of groundwater monitoring networks and the number of required wells becomes an important engineering optimization problem. The goal of this study is to find an optimum monitoring network with the fewest wells that provides sufficient spatial coverage on groundwater quality. With the presented method redundant wells in an already existing network are identified. Here, a genetic algorithm (GA) based optimization approach is used in which each monitoring well in the watershed is represented with a binary GA bit to evaluate if the corresponding monitoring well will be selected for the network. The proposed approach can solve the problem by simultaneously optimizing two conflicting objectives. The first objective is the maximization of the match between the interpolated groundwater quality concentration distributions obtained using data from all available monitoring wells and the wells from the newly generated network. The match is primarily evaluated using the Nash-Sutcliffe (NS) model efficiency. Groundwater quality is represented by the water quality index WQI that aggregates several quality parameters. The second objective deals with the minimization of the number of monitoring wells in the newly generated network by considering cost-related constraints. These two objectives are integrated in a single objective function where different combinations of both objectives are investigated by considering two cases. The applicability of the proposed approach is evaluated for the groundwater monitoring network of the Gediz River Basin (GRB) which is one of the most important river basins in Turkey. Findings indicate that the proposed approach significantly reduces the number of monitoring wells with a relatively small deviation of the spatial distribution of the WQI values. Also, the monitoring network is optimized such that sampling points are removed from less polluted areas and selected in areas with higher pollutant concentrations.