A PSO-SVD Based Hybrid Approach for Inversion of Vertical Electrical Sounding Data – A Case Study from Western Maharashtra

Direct Current Vertical electrical sounding (VES) surveys always play a very important role in the successful identification of groundwater reservoirs. A wrong interpretation of the VES data may lead to unwanted consequences of delineating no water table at all. In the present work, we have shown the application of Particle Swarm Optimization (PSO) constrained on the results of Singular Value Decomposition (SVD) based inversion for the estimation of layer thickness and layer resistivity. The results obtained from SVD-based inversion are used to define the search space for model population generation for the PSO to operate. This hybrid approach utilizes the advantages of both the methods viz. good exploration and exploitation abilities of PSO to find a global solution and ability of SVD to reach in the vicinity of the probable solution. The presented approach is applied for the inversion of VES data from a sounding location in crystalline hard rock area of Sindhudurg district of Western Maharashtra region in India. The interpretation of the inverted results shows the water table to be at around 11 meters, which matches considerably with the known depth of water table from previous studies in the area northern to the present study area.