Estimation of hydraulic parameters by using VES sounding and neural network techniques in the semi-arid Khanasser valley region, Syria

Abstract

This paper discusses the in near–real time processing of Global Navigation Satellite An alternative approach based on using Vertical Electrical Sounding (VES) measurements and Artificial Neural Network (ANN) technique is newly proposed for computing the hydraulic conductivity K and the transmissivity T of an aquifer. VES measurements in the locations, where available water samples exist are required in such an approach, in order to train a neural network with fitting capability to evaluate both the hydraulic conductivity and transmissivity. The hydraulic conductivity and transmissivity are thereafter extrapolated by the use of trained neural network, even in the VES points where no water samples exist. This approach is practiced and tested in the Khanasser valley, Northern Syria, where the hydraulic conductivity and the transmissivity of the Quaternary aquifer is computed. We find an acceptable agreement between the hydraulic conductivity values obtained by the new approach and those obtained by the pumping test, which range between 0.864 and 8.64 m/day.