Geological model selection for the Bedretto lab using gravity and magnetic observations

This thesis is written in collaboration with the ETH Zurich and the Bedretto Underground Laboratory for Geosciences and Geoenergies (BULGG). The Bedretto lab is located in an old train supply tunnel in the Swiss central Alps. A better understanding of the subsurface around the lab is of high importance for the simulations carried out in the lab. For the thesis, fieldwork is carried out to collect gravity and magnetic measurements both on the surface of the Bedretto area and inside the tunnel. As a preparation to the fieldwork, a survey design is done to define measurement areas, which are sampled more or less densely. The data is corrected and separately inverted into density and magnetic susceptibility contrast models. A geological model selection is carried out by running the inversion with four different starting models. These are based on prior geological knowledge of the area. The four starting models are: (1) homogeneous starting model of gneiss, (2) starting model with a gneiss - granite transition implemented at tunnel meter (TM) 1134, (3) starting model with 11 major fault systems of the area, (4) starting model with both the gneiss - granite transition and fault systems. The inversion results showed that with an increasing complexity of starting models, a decreasing detail in contrast values was shown and an increase in model misfit parameters was observed. The inversion using a homogeneous starting model, could resolve four distinct geological units found in the area, but the assigned density values were not in line with values found in literature. The four geological units observed in the inversion results are the Rotondo granite on the ridges, the Prato series (gneiss) for the first 1000 TM, the Nufenen zone in the valley along the river and lamprophyre dykes on the passes. The magnetic inversion results did not resolve any geological units, but indicate a high magnetic contrast that aligns perpendicular to lithological changes