Joint inversion of receiver function and magnetotelluric data via a nature-inspired stochastic hybrid technique

This study introduces a novel joint inversion scheme using the variable Particle Swarm Optimization-Grey Wolf Optimization (vPSOGWO) algorithm, specifically designed to improve the integration of Receiver Function (RF) and Magnetotelluric (MT) data. The proposed algorithm incorporates variable inertia and population sizes, enhancing both exploration and exploitation capabilities, along with Probability Density Function addressing limitations in traditional inversion methods. The algorithm was first tested on synthetic RF and MT datasets and subsequently applied over the Garhwal Himalayas. The area consists of an intermediate high-conductivity layer (3.2 ohm m) between 6 and 14 km depth, which is likely caused by fluid release from metamorphic processes. This layer's presence is crucial for understanding seismic phenomena in the region, including the clustering of major earthquakes and the concentration of smaller seismic events along fault zones. The fluid dynamics within the highly porous rocks may play a pivotal role in facilitating seismic activity, further highlighting the need for continuous geophysical monitoring. Thus, the thrust-collision model for the Himalayas and provide insights into the relationship between electrical resistivity, fluid migration, and seismicity. The study demonstrates the effectiveness of the vPSOGWO algorithm in solving complex geophysical inverse problems. Future work aims to expand the algorithm's application to 2D/3D inversion challenges and further enhance computational efficiency for more precise and reliable geophysical models. The proposed method has the potential to improve monitoring and modeling in other tectonically active regions, making it a valuable tool in seismology and geophysical research.