Predicting land subsidence rate by estimating aquifer hydraulic parameters, obtained from aquifer-test data and optimization algorithms in the district 19 of Tehran

Groundwater modeling, land subsidence hazards and proper management of groundwater resources of the alluvial aquifer in the district 19 of Tehran, south of Tehran Plain need precise estimation of aquifer hydraulic parameters. Besides, traditional techniques and usual graphical methods have been approximate, expensive and time-consuming. In this paper, ten sets of aquifer test data were selected; moreover, these data were recorded in a well located in district 19 of Tehran, southwest of Tehran Plain during five years (2008 to 2012). For solving the previous methods’ problems, three computer codes have been developed to optimize aquifer parameters using three optimization approaches on the other hand, two kinds of genetic algorithms and a multi-elitist particle swarm optimization (MEPSO) which avoid getting stuck in local optima and save time. The efficacy and efficiency of the developed codes have been examined using ten sets of aquifer test data of a confined aquifer, and then their results have been compared with the results obtained by the graphical approach using AquiferTest software. Based on the fitness function, i.e. sum of square errors, the MEPSO and the GAs in descending order are more reliable for estimating the parameters contrast with the graphical method. Furthermore, the sensitivity analysis of the parameters during the performance of the optimization approaches has authenticated that the results obtained are enough precise and reliable. Then an equation has been presented according to the amounts of hydraulic conductivity which have been obtained using MEPSO during the years and the amounts of land subsidence rates which have been obtained using geodetic measurement methods to predict the amounts of land subsidence rates through the time when the amount of hydraulic conductivity will reach to Ultimately, based on the equation, after 30 years the amount of hydraulic conductivity will reach to and the total amount of land subsidence will be 0.5213 m from 2008 to 2038 . Moreover, land subsidence rates’ data obtained from interferometry synthetic aperture radar (InSAR) have confirmed the accuracy of the equations.