Joint Gramian inversion of geophysical data with different resolution capabilities: case study in Yellowstone

Joint inversion of multiphysics data is a practical approach to the integration of geophysical data, which produces models of reduced uncertainty and improved resolution. The development of effective methods of joint inversion requires considering different resolutions of different geophysical methods. This paper presents a new framework of joint inversion of multiphysics data, which is based on a novel formulation of Gramian constraints and mitigates the difference in resolution capabilities of different geophysical methods. Our approach enforces structural similarity between different model parameters through minimizing a structural Gramian term, and it also balances the different resolutions of geophysical methods using a multiscale resampling strategy. The effectiveness of the proposed method is demonstrated by synthetic model study of joint inversion of the P-wave traveltime and gravity data. We apply a novel method based on Gramian constraints and multiscale resampling to jointly invert the gravity and seismic data collected in Yellowstone national Park to image the crustal magmatic system of the Yellowstone. Our results helped to produce a consistent image of the crustal magmatic system of the Yellowstone expressed both in low-density and low-velocity anomaly just beneath the Yellowstone caldera.