Methods and uncertainty estimations of 3-D structural modelling in crystalline rocks: a case study
- 1Institute of Geological Sciences, University of Bern, Baltzerstrasse 1 + 3, 3012 Bern, Switzerland
- 2Graduate School AICES, RWTH Aachen University, Schinkelstrasse 2, 52062 Aachen, Germany
- 3Nagra, Hardstrasse 73, 5430 Wettingen, Switzerland
Abstract. Exhumed basement rocks are often dissected by faults, the latter controlling physical parameters such as rock strength, porosity, or permeability. Knowledge on the three-dimensional (3-D) geometry of the fault pattern and its continuation with depth is therefore of paramount importance for applied geology projects (e.g. tunnelling, nuclear waste disposal) in crystalline bedrock. The central Aar massif (Central Switzerland) serves as a study area where we investigate the 3-D geometry of the Alpine fault pattern by means of both surface (fieldwork and remote sensing) and underground ground (mapping of the Grimsel Test Site) information. The fault zone pattern consists of planar steep major faults (kilometre scale) interconnected with secondary relay faults (hectometre scale). Starting with surface data, we present a workflow for structural 3-D modelling of the primary faults based on a comparison of three extrapolation approaches based on (a) field data, (b) Delaunay triangulation, and (c) a best-fitting moment of inertia analysis. The quality of these surface-data-based 3-D models is then tested with respect to the fit of the predictions with the underground appearance of faults. All three extrapolation approaches result in a close fit ( > 10 %) when compared with underground rock laboratory mapping. Subsequently, we performed a statistical interpolation based on Bayesian inference in order to validate and further constrain the uncertainty of the extrapolation approaches. This comparison indicates that fieldwork at the surface is key for accurately constraining the geometry of the fault pattern and enabling a proper extrapolation of major faults towards depth. Considerable uncertainties, however, persist with respect to smaller-sized secondary structures because of their limited spatial extensions and unknown reoccurrence intervals.