Review of groundwater quality monitoring network by combining clustering and geostatistical methods in Tehran-Karaj study area

Protecting the quantity and quality of water resources has always been of great importance in all human societies, and in order to maintain the quality of these resources, numerous monitoring and remedial measures have been taken in most countries of the world. In this regard, water quality monitoring is considered as one of the essential tools and as an integrated activity to evaluate the physical, chemical and biological factors of water that are related to human health and living organisms. In the present study, a model of combining geostatistical methods (Kriging), clustering and entropy theory has been proposed to review and present the groundwater quality monitoring network. This model is presented under the first and second approaches in Tehran-Karaj study area. The first approach, without using the clustering method, reviews the existing monitoring network without using only entropy theory and Kriging geostatistical method as an estimator. The second approach uses the k-means clustering method, entropy theory and Kriging geostatistical method as an estimator to investigate the effect of combining these three methods on the review of the existing monitoring network and then the results of the first and second approaches are compared. The proposed final monitoring network with 44 wells has an average forecast error rate of 19 and a cost reduction of 34 percent compared to the cost of the current monitoring network. Also, using clustering, the average percentage of estimation error has been reduced by 20 percent compared to the case without clustering.