OPTIMAL MONITORING NETWORK DESIGN FOR EFFICIENT IDENTIFICATION OF UNKNOWN GROUNDWATER POLLUTION SOURCES

Application of linked simulation-optimization approach for solving groundwater identification problems is well established. Pollutant concentration measurements from different sets of monitoring locations, when used in a linked simulation-optimization approach, results in different degrees of accuracy of source identification. Moreover, the accuracy of source identification results depends on the number and spatiotemporal locations of pollutant concentrations measurements. This study aims at improving the accuracy of source identification results, by using concentration measurements from an optimally designed monitoring network. A linked simulation optimization based methodology is used for optimal source identification. Genetic programming based impact factor is used for designing the optimal monitoring network. Concentration measurement data from the designed network is then used, in the Simulated Annealing based linked simulation-optimization model for efficient source identification. The potential application of the developed methodology is demonstrated by evaluating its performance for an illustrative study area. These performance evaluation results show improvement in the efficiency in source identification when such designed monitoring networks are utilized.