The presence of active fault traces in proximity to any new infrastructure project is a major concern for the design process. The relative displacements that can be experienced in surface fault rupture during a seismic event must be either entirely avoided or mitigated in some way. Blind faults present a significant challenge to engineers attempting to identify these hazards. Current standards of practice employed to locate these features are time consuming and costly. This work investigates the geophysical methods of refraction microtremor (ReMi) and seismic refraction with regard to their applicability in this task. By imaging a distinct lateral variation in the shear wave velocity (Vs) profile across a short horizontal distance, these methods may provide a means of constraining traditional investigation techniques to a more focused area. The ReMi method is still very new, but holds key advantages over other geophysical methods in its ease of application and ability to achieve good results in highly urban settings. It is one of the few geophysical techniques that does not suffer in the presence of high amplitude ambient vibrations. The seismic refraction method is here applied in an attempt to corroborate data obtained through the ReMi analysis procedure. Sensitivity, precision parametric studies are carried out in order to learn how to best apply the ReMi method. Both tests are then applied at a previously trenched fault trace to determine whether the data can be matched to the subsurface information. Finally, the methods are deployed at a location with an inferred fault trace where little to nothing is known about the subsurface. The precision study indicates a coefficient of variation for the ReMi method on the order of 7%. At the known fault trace both methods generally agree qualitatively with available subsurface data and each other. Using the ReMi method, a marked shift is observed in the Vs profile laterally across the fault trace. In the case of the inferred fault trace, the same type of lateral variation in the V­­s profile is observed using the ReMi method. The seismic refraction at this site does not agree with the ReMi data, but seems reasonable given the visible geomorphology. Receiver arrays placed in close proximity to the inferred fault trace recorded erratic signals during seismic refraction testing, and displayed abnormal response modes after transforming the ReMi data to frequency-slowness space. These anomalies may possibly be attributed to the presence of abnormal subsurface structural geometry indicative of faulting.

Η παρούσα διπλωματική εργασία περιγράφει την έρευνα της θαλάσσιας γεωφυσικής διασκόπησης η οποία εκτελέστηκε στον Πατραϊκό κόλπο και παρουσιάζει τα αποτελέσματα της ερμηνείας των γεωφυσικών στοιχείων που συλλέχθηκαν με την βοήθεια του τομογράφου υποδομής πυθμένα. Τα στοιχεία συλλέχθηκαν από το Εργαστήριο Θαλάσσιας Γεωλογίας και Φυσικής Ωκεανογραφίας του τμήματος Γεωλογίας του Πανεπιστημίου Πατρών. Η συλλογή, επεξεργασία και ερμηνεία του συνόλου των σεισμικών γραμμών επέτρεψε την χαρτογράφηση των υποθαλάσσιων ρηγμάτων του Πατραϊκού κόλπου. Ο χάρτης με τα υποθαλάσσια ρήγματα αποτελεί έναν τροποποιημένο χάρτη από τον ήδη διαθέσιμο χάρτη ρηγμάτων του Πατραϊκού κόλπου του 1985 από τους Ferentinos et al., 1985. / The present study describes the submarine geophysical survey which was carried out in the Gulf of Patras and presents the results of the geophysical data analysis using a subbottom profiler system. The data were collected by the Laboratory of Marine Geology & Physical Oceanography, department of Geology, University of Patras. A detailed fault map was produced by the data analysis of the collected seismic profiles of the Gulf of Patras. The present fault map is a modified map from a former map that has been produced in a 1985 survey by Ferentinos et al., 1985.

In this thesis, I constrain rupture scenarios of active crustal faults in the southern Strait of Georgia and Boundary Bay in order to assess their tsunamigenic potential. The NW-SE-trending Drayton Harbor, Birch Bay, and Sandy Point faults had been previously identified on the southern side of Boundary Bay from aeromagnetic, LiDAR, and paleoseismic data; all show evidence of abrupt vertical Holocene displacements. South of Boundary Bay, the E-W-trending Skipjack Island fault zone was recently mapped on the basis of multibeam sonar imagery and seismic reflection data, with evidence for Holocene offsets of the seafloor and subsurface sediments. In addition, the Fraser River Delta fault had been hypothesized on the basis of a line of pockmarks and fluid seeps. Since these faults have only been recently mapped and identified as active, there is little information available on their structure, rupture style, and past large earthquakes. This makes it difficult to constrain rupture models to predict how fault slip could displace the seafloor during a large earthquake, for input to tsunami models. I analyzed relocated earthquake hypocentres, earthquake mechanisms, bathymetry, topography, and aeromagnetic, seismic reflection, and magnetotelluric data, to constrain the location, strike, dip, and rupture width of each fault. Correlations between datasets enabled mapping of northwestward extensions of the Sandy Point and Birch Bay faults, as well as delineating the previously unmapped Fraser River Delta fault. These offshore faults appear to be associated with infilled basement valleys in the subsurface, perhaps due to differential glacial erosion of weakened fault zone material. The Drayton Harbor fault could not be definitively mapped across Boundary Bay, so was excluded from the rupture modelling. Rupture styles were constrained using a combination of earthquake mechanisms, stress orientations, other evidence of regional compression, and vertical paleoseismic offsets. Where possible, paleoseismic displacements in past earthquakes were used to constrain the amount of fault slip for scenario earthquakes; empirical relations between fault slip and fault length or area were used to estimate displacements for the Skipjack Island and Fraser River Delta faults. The Birch Bay, Sandy Point, Skipjack Island, and Fraser River Delta faults all pose a significant tsunami risk to communities surrounding the southern Strait of Georgia and Boundary Bay. Considering both the originally mapped and extended lengths, the Birch Bay and Sandy Point faults could rupture in reverse-faulting earthquakes up to Mw 6.7-7.4 and 6.8-7.5, respectively, with seafloor uplift up to 2-2.5 m triggering damaging tsunami waves (up to at least 2.5 m) that could arrive onshore with little to no warning after the shaking begins. Similarly, the Fraser River Delta fault could host reverse or dextral-reverse slip earthquakes up to Mw 7.0-7.6, with seafloor uplift of 0.6-3.5 m. Ruptures on the Skipjack Island fault would likely have a larger strike-slip component; earthquakes of Mw 6.9-7.3 produce modelled seafloor uplift of 0.5-1.9 m. These results suggest that large tsunamigenic earthquakes on crustal faults in the southern Strait of Georgia should be included in future seismic and tsunami hazard assessments on both sides of the international border. / Graduate