Structurally based joint inversion of electromagnetic, seismic and potential field data for improved classification and characterization: Field examples from the vadose zone

Abstract

A key strategy to characterize subsurface materials is the integrative analysis of tomograms of several physical properties provided by different geophysical data. Among the recent strategies to produce these tomograms is the joint inversion. Unfortunately, most joint inversion techniques are based on hypothetical property relationships that hinder the site-specific property correlations that characterize individual materials. Differently, a recently developed cross-gradient joint inversion technique is underpinned by the hypothesis of structural resemblance and permits the coexistence of natural property correlations that provide the key to unravel the signatures of subsurface materials. In the present work, electromagnetic, seismic and potential field data are jointly inverted for several field sites with near surface targets. The results show not only a clearer disposition of the mapped units in structural terms, but also a sound evidence of their actual fluid-mineral characteristics. In general, the results demonstrate the power of analyzing multiple-property images when they are structurally consistent.