A new perspective on fractures - using unmanned aerial vehicles to analyze tectonic structures in high resolution

• Main Author: Weismüller, Christopher
• Format: Thesis
• Language: English
• Published: RWTH Aachen University 2021
• Subjects: fractures; Iceland; Bristol Channel; UAV; drone; photogrammetry
• Online Access: https://dx.doi.org/10.18154/rwth-2021-04963; https://publications.rwth-aachen.de/record/819297

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2021; Aachen : RWTH Aachen University 1 Online-Ressource : Illustrationen, Diagramme (2021). = Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2021 : Fractures are ubiquitous structures in the brittle crust of the Earth and occur at a variety of scales. Better understanding them contributes to a wide field of implications such as water and geothermal energy supply, natural hazard assessment, paleostress analysis, basin modelling, hydrocarbon exploration or reservoir geology. Numerous methods are used to study the evolution and geometry of fractures. These methods consist of classical ground-based field work and mapping, large scale remote sensing studies based on satellite data, analog models and numerical models, or combinations of different methods. In this work, I present fracture analyses based on high-resolution digital elevation models (DEMs) and orthorectified mosaics that I created using unmanned aerial vehicle (UAV) photogrammetry techniques. The objective of using these novel methods is to bridge the gap between classical ground-based field work and analyses using aerial and satellite imagery for the investigation of the structure and geometry of fractures and fracture networks. Photogrammetric models containing real world data serve as benchmark for analog models of faults and associated fractures and are used to develop and utilize workflows for the digital extraction and interpretation of wide-ranging sets of fracture data. To investigate the structure and evolution of massively dilatant faults on Iceland, I created a comprehensive dataset of fractures associated with dilatant faults. I mapped the fractures and extracted geometrical measurements based on surface models. Furthermore, the models and results were used as benchmark for analog models. The identification of surface structures in the real-world data and the corresponding structures in the analog models aid the prediction of the subsurface structure of massively dilatant faults based on their surface expressions. I show that massively dilatant faults can be described as different endmembers of their surface manifestation with respect to their opening width and vertical offset. All endmembers are part of a continuum with smooth transitions. Furthermore, the analog models show that dilatancy and geometry of dilatant faults systematically changes with the obliquity of the basement fault. Models predict a general fault structure with large, subvertical open fractures and a zone of hybrid failure between opening mode in shallower depth and shear failure at greater depth. Often described but poorly understood surface ramps in normal fault zones are found to be the manifestation of tilted blocks, that are further classified according to their kinematics and the location of their hinge. Beneath these tilted blocks, so far barely acknowledged large tectonic cave systems are predicted. I created high-resolution orthomosaics of the world-class outcrop of the 100.000 m² fractured carbonate rocks in the Bristol Channel near Lilstock, UK. In five selected domains of this outcrop, I manually traced and interpreted fractures using the orthomosaic as basemap and assigned fracture generations based on abutting and crosscutting criteria. The manually traced fracture network was compared to a network generated with an automatic tracing technique based on ridge-detection. The comparison shows that the automatic method can produce similar fracture maps in shorter time, however, with a bias towards closely spaced Y over X-nodes. Automatic assignment of fracture generations is not yet possible. Quantitative analyses of the network connectivity, fracture density and intensity reveal a spatial variety of these parameters over short distances. Overall, the network is well connected and has slightly different characteristics in the different domains. An evolutionary model of the network connectivity based on the manually interpreted fracture generations was introduced. It reveals an overall increase of connectivity with additional fracture generations, that varies in its degree depending on the number and geometry of fractures within a certain generation. The findings of both projects show that UAV photogrammetry is a powerful tool in structural geology and neotectonics. Large amounts of data can be acquired in a short time and can be used to create high-resolution models of the fractures and their surroundings. I present qualitative analyses based on these models, but also show that high numbers of quantitative measurements can be taken, what would barely be possible to do so manually. This high density of measurements helps to reveal and quantify so far unrecognized geometries and structural details. : Published by RWTH Aachen University, Aachen