Optimal Design of Groundwater Monitoring Network Using the Combined Election-Kriging Method

Abstract

Groundwater monitoring requires a great deal of cost and time that optimizing the quantitative groundwater monitoring network with selecting the optimal number of sampling wells and determining their optimal location for reducing the cost and time of quantitative groundwater assessment are necessary. The data from 110 observation wells of Neyshabur plain in Iran range from the year 1986 to 2016 were studied. The combined Election-Kriging method was used to analyze these data, and the Pareto chart was plotted to determine the optimal number and location in two scenarios. The first scenario was to determine the optimal wells location among the existing wells and the second scenario was to determine the optimal wells location for monitoring groundwater levels throughout the plain. To limit the search space, the maximum and minimum number of monitoring network wells were selected 30 and 85 respectively. Based on the results, the selected method accurately provided the appropriate location of wells, so that in the first scenario, the RMSE values for the number of wells were in the range of 0.71 to 2.34 m. which are acceptable. In the second scenario, the RMSE values of between 1.04 and 2.89 m were obtained, which are appropriate values according to the objective of the problem. Also, the distribution of the wells selected in the area is also uniform in most numbers according to the second scenario, which indicates the good accuracy of the method.