De Meyer, Cédric

Master thesis supervised by Sintubin, Manuel; Van Noten, Koen (KULeuven)

Abstract: The World Stress Map (WSM) compiles accurate stress regime data from indicators such as earthquake borehole overcoring, hydro-fracking, borehole slotting, the analysis of fault slip rates, borehole breakouts and most importantly, focal mechanism analysis. The stress regime data is used in scientific studies such as seismic hazard assessment, ore-forming geology and geodynamics. Accurately determining the local stress regime is crucial for improving and extending the World Stress Map. In Belgium, the Royal Observatory of Belgium is the research institution responsible for acquiring and processing seismological data. The ROB has an in-house program (DP) that determines the focal mechanism solutions for earthquakes in and around Belgium. This code is based on the FOCALMEC program developed in the PhD dissertation of Camelbeeck (Camelbeeck, 1993), which was written in the older programming language of FORTRAN. The main goal of this thesis is to develop a focal mechanism solution Python program based on the analysis of P-wave first motions, similar to the ROB FORTRAN code but following modern standards in seismology. A second objective is to determine the regional stress regimes in and around Belgium using the obtained focal mechanisms of the earthquakes that have sufficient P-wave arrivals in the ROB database. The developed program is based on an extensive literature study that provides the basic concepts of P-wave first motion focal mechanism analysis. In total, 33 earthquakes are studied, which are geographically categorised in the Ruhr Valley Graben, the London-Brabant Basement, the Ochtendunger Fault Zone and the western Ardennes. Comparing the obtained results for these earthquakes with focal mechanism solutions found in literature or provided by the ROB shows that the developed program provides similar solutions. However, the P-wave first motion polarity method has an important limitation. It is only possible to determine if a seismic station is located in the correct quadrant of the focal sphere qualitatively. As a result, the fitting method used provides all solutions for which most stations are located in the correct quadrant. The variability in possible solutions changes depending on the distribution of the seismic stations around the earthquake’s epicentre. When the number of seismic stations for which a P-wave polarity could be determined is high, and their spatial distribution around the hypocentre is low, the program can provide accurate focal mechanism solutions. When both these criteria are lacking, the variability in provided solutions can be quite high and it becomes difficult to determine an accurate focal mechanism solution. Based on the focal mechanism solutions for each earthquake, the most likely nodal plane responsible for the earthquake is discussed and generally shows coherence with other publications. The determination of this most likely nodal plane is based on the geological context of the different regions and is not obtained from the focal mechanism solution itself. For each earthquake, the corresponding stress regime is discussed. Combining the stress regimes for each determined geographical region allows for the determination of the regional stress regime. The seismologically-active Ruhr Valley Graben is mainly characterised by NNW-SSE normal faults along which predominantly ENE-WSW extension takes place. However, some earthquakes could also have taken place on antithetic faults with similar strikes. In addition, some earthquakes show transtension along WNW-ESE faults that result from a Late Oligocene rifting phase. Lastly, a possibly anomalous compressional to transpressional stress regime is observed along a likely NNW-SSE fault in the eastern part of the Ruhr Valley Graben in the Campine Basin. This stress regime is obtained from the focal mechanism of an induced earthquake in the region. This anomalous character could be due to the erroneous determination of the earthquake’s depth or location. It is also possible that the stress regime is not anomalous compared to the predominant stress regime in Ruhr Valley Graben and that in the CampineBasin, the stress regime has a compressionalcomponent. Therefore, additional research should focus on this issue. In the region of Court-Saint-Etienne, the London-Brabant Basement is characterised by a uniform strike-slip stress regime, which induces sinistral strike-slip earthquakes along shallow faults. The Ochtendunger Fault Zone (Germany) is characterised by a more or less uniform strike-slip stress regime in which fault movements are likely dextral. Lastly, in the western Ardennes, some earthquakes show a ENE-WSW extensional stress regime along NNW-SSE striking normal faults, corresponding to the neighbouring region of the Ruhr Valley Graben. However, other earthquakes show NNW-SSE extension or strike slip. It is thus believed that the stress regime in the Ardennes is complex due to the complexity of its subsurface. In addition, a sensitivity analysis is conducted that takes into account a change in depth. For every earthquake, the focal mechanism is determined for depths ranging between -6 km and +6 km the calculated focal depth of the earthquake. The influence of a change in depth strongly depends on the earthquake and, more importantly, the distribution of seismic stations around the earthquake’s hypocentre. If the distribution of stations is adequate small changes in nodal plane orientations with depth can be observed. This results from the low variability in possible solutions for well-determined focal mechanisms. If a good distribution of the stations is lacking, the change of the focal mechanism solution will be very small because the variability in solutions is very large. However, in general, the focal mechanism solutions do not change drastically, and as a result, the determined stress regime largely remains the same.

Keyword(s): Focal Mechanism ; Stress-state ; Seismotectonics