Assessment of Bayesian PDF for 2D structure using Hybrid Global Inversion of Magnetotelluric datasets.

Non-linear, ill pose and multidimensionality problems have been a challenging task while dealing with geophysical inversion. Solving such problems with conventional inversion techniques give inappropriate solutions which increases from 1D to 2D problem and increases the probability to converge in local minima. Therefore, the study is focused on using hybrid global optimization technique that hybridized Particle Swarm Optimization (PSO) with Grey Wolf Optimizer (GWO) due to their individual capability of balancing the exploitation and exploration properties. The present work aims at applying the hybrid algorithm over 2D Magnetotelluric datasets with the application of Bayesian Probability Density Function for initialization of the inversion. The algorithm is tested on two block model assuming a conductive and a resistive body with 20% noisy datasets. The results indicate that the hybrid inversion algorithm successfully resolve both the conductive and resistive structures.