Author(s): Mustafa Tamer Ayvaz, Alper Elci
Linked Author(s): Tamer Ayvaz
Keywords: Water resources, groundwater quality, monitoring network design, genetic algorithm, interpolation
Abstract: Planning and management of groundwater systems require appropriate monitoring data for groundwater quality and aquifer hydraulic head measurements. These data are usually collected from monitoring wells which are spatially distributed in the basin. Since monitoring of groundwater systems can be a time consuming and costly task, a minimum number of monitoring wells with an optimum spatial distribution is desired. Therefore, the conception of monitoring networks becomes an important engineering optimization problem. For this purpose, a genetic algorithm (GA) based optimization approach is proposed in this study for seeking the optimum groundwater quality monitoring network, starting with an already available set of monitoring wells. The goal of the proposed approach is to determine the optimum numbers and locations of the monitoring wells which provides equivalent amount of groundwater quality information with those obtained by using all available monitoring wells. This task was accomplished by representing each monitoring location with a binary bit in a GA chromosome to determine whether the associated location will be selected for the network. Then, the corresponding configuration fitness value was calculated by interpolating the associated water quality data over the field. The configuration fitness consists of the two objectives which are the maximization of the Nash-Sutcliffe model efficiency and the minimization of the number of monitoring wells in the newly generated configuration. Integration of these two objectives in an optimization framework results in best solutions with a minimum number of monitoring wells over the entire basin. Applicability of the proposed solution approach was evaluated by using groundwater quality data for the Gediz River Basin, a major basin located in western Turkey. The model results indicate that the proposed approach significantly reduced the number of monitoring wells with a relatively small deviation of the spatial distribution of the studied water quality parameter
Year: 2017