Resolving the fine‐scale density structure of shallow oceanic mantle by Bayesian inversion of localized geoid anomalies

Abstract

We present a new approach to estimate the density structure of shallow oceanic mantle by inversion of localized geoid anomalies. Our method is based on Bayesian statistics and is implemented by combining forward modeling with Markov Chain Monte Carlo sampling. The inherent nonuniqueness of such inversion is reduced by using spectral localization, reference models, and a priori bounds on the amplitude of density perturbations expected within the convecting mantle. We apply this approach to the geoid anomalies around the Mendocino Fracture Zone that have recently been revealed by wavelet analysis. The depth and vertical extent of density anomalies derived from our inversion indicate that they are intimately related to the structure of the lowermost lithosphere. The amplitude of density perturbations and their spatial organization suggest the occurrence of small‐scale convection induced by a lateral temperature gradient across the fracture zone. As its applicability is not limited to the vicinity of fracture zones, the new inversion method should allow us to resolve the fine‐scale density structure of shallow oceanic mantle beneath the world's oceans.