The validation of results of groundwater monitoring network optimization for Dehgolan Plain

In recent decades, optimization of piezometers and the reducing costs of data logging have been favored by scientific societies and become very conventional. Dehgolan plain is one of the great plains of Kurdistan province, which has been significantly exploited in recent years due to a lot of authorized and illegal wells. These wells have caused a sharp decrease in water table in this area. The purpose of this study is validating of optimization the groundwater monitoring network of Dehgolan aquifer using kriging. Hence, the active piezometers of this plain have been used in different water stress periods including 12 months of 2006, 2011 and 2016. Therefore, for each month, the best variogram was chosen based on the lowest nugget effect and ceiling and the maximum range. After calculating the estimation error during cross validation, one piezometer with least root mean square error was removed. The optimal network was determined by repeating this operation until the error of the estimate was less than 11% of the first interpolation. The results show that removable piezometers vary for different months and in a long time none of them can be ignored. Although removing of the piezometers during the first stress period did not pose a particular problem, in next water stress periods created problem and rejected the validity and reliability of optimization results due to decreased water table over time that cause to dry some piezometers. One should be very cautious in monitoring network optimization because using the seasonal data instead monthly data can be misleading and reduces the error.