Inversion of Surface Nuclear Magnetic Resonance for Groundwater Exploration

Abstract

AbstractThis paper presents briefly the forward theory of surface nuclear magnetic resonance for groundwater exploration. This method computes the numerical integral of nuclear magnetic resonance kernel function which contains the double Bessel function using the Gauss quadrature and continued fraction expansion. Then we discuss the influence of the conductivity on the signal of nuclear magnetic resonance. Based on singular value decomposition algorithm, we have performed an inversion study on the theoretical model of surface nuclear magnetic resonance for detecting groundwater and gained the ideal result. We have also made inversion experiments respectively on the insulative and electric models involving the improved simulated annealing algorithm. The result shows that this method is feasible, the inversion is relatively stable and the convergence is fairly fast.