Nonlinear inverse modeling of gravity data for basement relief using imperial competition algorithm

Gravity inversion is a classical tool in applied geophysics. Inversion of basement relief of sedimentary basins is an important application among the nonlinear techniques. Classically, local deterministic optimization techniques have been employed to solve the non-linear gravity inverse problem. Swarm intelligence algorithms, such as ant colony algorithm or particle swarm optimizers, are promising alternatives to classical inversion methods. In this study, imperialist competitive algorithm (ICA), was designed and utilized for two-dimensional (2D) gravity inversion of basement relief in sedimentary basins. Reliability of this technique was tasted by modeling of gravity data acquired from a synthetic model, and then, the synthetic model parameters were obtained from this modeling approach with acceptable accuracy. Moreover, the results of utilizing this approach on noisy data showed that this approach was robust to the presence of noise in the data. For the case of real data, this approach was applied on a real gravity profile in Atacama Desert (north Chile) and the results were confirmed with previously published works related to this area. Generally, compared to techniques already proposed for 2D nonlinear gravity inversion, the ICA technique, proposed here, appears as a powerful tool for estimating the basement relief of sedimentary basins.