Joint inversion and integration of multiple geophysical data for improved models of near-surface structures

Abstract: Geophysical methods are non-invasive and allow an effective way of understanding subsurface structures and their physical properties. One of the main challenges is the often non-uniqueness of the geophysical models and that several different models can explain a dataset to an agreeable fit. Moreover, noise and limitations in resolution, which are inherent to field data, are additional obstacles for obtaining a true physical property model of the subsurface. Facing all these challenges, geophysicists have dedicated their efforts for decades to recover models that represent, as close as possible, the true subsurface. Joint inversion and integration of multiple geophysical data are two main approaches that I studied to better resolve subsurface structures. I further used these approaches, together with new software and hardware implementations for data acquisition and inversion, for near-surface applications. In this thesis, radio-magnetotelluric (RMT), boat-towed RMT, boat-towed controlled source MT (CSMT), electrical resistivity tomography (ERT), and first-arrival traveltime tomography are jointly used for quick clay investigations and fracture zone delineation under shallow water-bodies. The joint approach, as compared with any individual method, shows a better ability to both resolve the geological targets and to assist in understanding the subsurface geology that hosts these targets. For examples: by performing the joint inversion of lake-floor ERT and boat-towed RMT data, a fracture zone is better delineated with greater details compared with single inversion; by employing boat-towed CSMT measurements and jointly inverting with boat-towed RMT data, the subsurface structures, especially at greater depth, are better resolved than by inverting each dataset alone.During my PhD studies, two types of new implementations were employed. (1) Boat-towed data acquisition system was implemented to expand the RMT and CSMT method from land to shallow-water applications. This is significant since many large-scale underground infrastructures are likely to cross these water zones (for example multi-lane train or bypass tunnels, such as the Stockholm bypass). (2) The modification of a well-structured code EMILIA allows joint inversion of boat-towed RMT and lake-floor ERT datasets, and the modification of another well-structured code MARE2DEM can accurately model high frequency CSMT data and handle joint inversion of boat-towed RMT and boat-towed CSMT datasets. Thus, the code modification as another type of new implementation guarantees the success of near-surface applications using the boat-towed RMT and CSMT data acquisition systems.Studies conducted during my PhD work, included under the SEG-GWB (the Society of Exploration Geophysicists - Geoscientists Without Borders) program and the TRUST (TRansparent Underground STructure) umbrella project, are useful for near-surface applications including, for examples, engineering purposes such as planning of underground infrastructures, site characterization in connection with energy or waste storage, and geohazard investigations.