ALBERT

Description: In recent years, joint inversion of magnetotelluric (MT) and seismological receiver function (RF) methods has provided valuable information about crustal structures and demonstrated a complementary tool. Joint inversion of such datasets that are not well correlated physically has been generally applied as structurally coupled in the literature. These are based on constraining the directions of parameter changes by a spatial regularization operator and placing their spatial changes at similar positions. However, the electrical conductivity and seismic velocity, which are the model parameters for MT and RF, may have different characteristics for the changes in the crustal zone, such as porosity, permeability, and temperature, and the unpredictable relationship between these parameters prevents full correlation. In this study, we jointly inverted the MT and RF data utilizing Pareto-based multiobjective particle swarm optimization (MOPSO) based on a structurally constrained approach without restricting the direction of the model parameters and compared the results with structurally coupled approach. MOPSO, which overcomes the difficulties of traditional inversion such as the dependence on the initial model and the trapping of local minima, achieves a global solution without requiring a regularization operator for either the model parameters or misfit functions. The presented approach was verified on synthetic models and an application from field data obtained from measurements at a single station on the Biga Peninsula in Turkey. The results of these analyses confirm the usefulness of the method as a new approach for joint inversion of geophysical data sensitive to different physical phenomena.