Zonal Cooperative Inversion of Crosshole Tomographic Data

In many near-surface geophysical applications, it is now common practice to acquire multiple geophysical datasets. The cooperative inversion of the co-located geophysical data sets may result in an improved reliability of the tomographically reconstructed parameter distributions underlying the acquired data. Here, we apply a recently developed zonal cooperative inversion approach to crosshole georadar and P-wave traveltimes collected at a well-studied test side (Horstwalde, Germany). In this inversion approach fuzzy c-means cluster analysis links the regularized least squares inversion algorithms used to invert the two traveltime data sets. This results in a single zonal model of the subsurface, which outlines the major subsurface units and explains both data sets. At the Horstwalde test site a comparison of the zonal model to direct push cone penetration test data demonstrates that the inversion approach delineates the major subsurface lithology.