Abstract Title: Assessment of Equivalence Problem by Joint Hybrid Global Inversion of 1D MT and DC Resistivity and Using Integral Equations of MT for Cost Function

Abstract Submitted to: NONLINEAR GEOPHYSICS

Abstract Text:

Inversion of geo-electrical data for interpretation of subsurface physical parameters has to deal with suppression and equivalence problem. Individual geo-electrical methods has been failed to solve the inherent problem of suppression and equivalence by various researcher. In order to resolve the problems, the study contain the joint inversion of magneto-telluric (MT) and direct current (DC) resistivity data sets has been carried out using hybrid particle swarm optimization and grey wolves optimizer (PSOGWO) technique, PSOGWO has ability to balance both the explore and exploit property, in which objective function for MT data set has been carried out with integral of alternative equations with the conventional equation for calculating the apparent resistivity. The hybrid technique has been observed over synthetic layered Earth structures over same search space for two different models, that is, Model 1 and 2 (maximum relative deviation is 10% and 50% for DC and MT curves between two models) which is taken from Harinarayana (1999). The inverted result of Model 1 and 2 even having same search range doesn’t interfere with each other model parameters and able to obtain inverted value close to true value of each model shown in Table 1. The apparent resistivity curve and phase for MT, apparent resistivity curve for DC with root mean square error (RMSE) curve for both models are presented in Figure 1 and Figure 2. To check the performance of joint inversion of DC and MT, it has been tested over a model with thin resistive layer in middle has been taken from Vozoff and Jupp (1974), which has successfully able to justify the motive of joint that individual method could not do it shown in Table 2 and its apparent curves for both method with RMSE are shown in Figure 3. The developed joint inversion techniques has ability to overcome the problem of equivalence by joint inversion of DC and MT using the hybrid PSOGWO global optimisation techniques, and able to predict model parameters which agree well with the true parameters.