Graph-based fracture network analysis to integrate structural geology properties and identify preferential flow pathways in the aquifer system of Tsanfleuron, Swiss Alps

Abstract

Graph theory has emerged as a promising method for analyzing fracture networks and complementing traditional geometrical descriptions. It emphasizes the network’s topological aspects, highlighting the importance that fractures have in enhancing permeability and influencing flow anisotropy. However, integrating different structural geology analyses in a single framework remains a challenge. We propose a graph-based fracture network analysis combining geometry, topology, kinematics, age relationships, and geomechanics to identify the most important connections within a network. We apply it to the karstic aquifer system of Tsanfleuron, in the Western Helvetic Alps domain, where fractures play a crucial role in groundwater circulation and karstification. We acquired new structural data from a high-resolution digital elevation model (DEM), a 3D point cloud digital outcrop model (DOM), fieldwork, and one scanline. We interpreted and measured lineaments in 1:2.500 (DEM) and measured attitudes (DOM). Verification was done in the field, adding more measurements and the kinematics. The fracture network was transformed into a graph, and measurements were stored as attributes. Four main subvertical fracture sets were identified in chronological order, for the study site: E-W (85°), NE-SW (57°), N-S (176°), and NW-SE (117°). Censored fractures were identified to unbiased length, and abutting relationships were defined for pairs of sets to define age relationships. The E-W and NE-SW are the most persistent and longer fractures through all the site. The N-S set is localized in the central area, and the set NW-SE superimposes the others. The displacement is mostly dextral for the NE-SW and E-W faults, and sinistral for the N-S and NW-SE sets. Paleo tensor analysis results in NW-SE directed maximum compression. Fractures are opened and often enlarged by dissolution. Connectivity is moderate in terms of percolation. Betweenness and percolation centralities highlight the preferential water flow towards NE, ENE, and E.