3-D joint inversion of surface wave dispersion and receiver functions

Abstract

Joint inversion of both surface wave dispersion and receiver functions using Markov chain Monte Carlo parameter space searches has become popular for revealing crustal structure with uncertainty estimates. Typically, joint inversion is performed in two steps: first, invert for 2-D surface wave velocity maps, then invert for the 1-D structure below each station using the surface wave dispersion extracted from the 2-D maps and receiver functions to construct a 3-D velocity structure by interpolation. However, this approach may fail to preserve lateral spatial variations and dipping discontinuities, leading to biased 3-D structures. Furthermore, the interpolation approach requires dense station spacing to be valid. A one-step 3-D direct inversion using reversible jump Markov chain Monte Carlo and 3-D Voronoi tessellation addresses these issues by inverting 3-D structures directly from frequency-dependent traveltime measurements and receiver functions. Our method can optionally account for dipping interfaces when forward modelling the receiver functions. Synthetic tests show that separate inversions of surface waves and receiver functions have limitations, while the joint solution provides better subsurface estimates and uncertainty assessments. Subsequently, we validate our method using real data from a case study in Sri Lanka, showcasing its practical applicability and effectiveness in real-world contexts. Compared to the classical two-step method, the direct 3-D inversion yields more reasonable results and interpretable uncertainties, preserving valuable lateral structure variations.