The contents of this article, including all figures, are the property of the Crown and are provided to the CSEG with its permission. Copyright remains with the Crown and may not be transferred.


The University of Victoria (UVic) offers a geophysics undergraduate degree through a joint program between the School of Earth and Ocean Sciences (SEOS) and the Department of Physics and Astronomy. At the graduate and research levels, geophysics at UVic is concentrated in SEOS and focuses largely, although not entirely, on topics in earthquake seismology, active tectonics, and geodynamics. This befits our unique location in Canada atop a major tectonic plate boundary, the Cascadia Subduction Zone (CSZ), located offshore western North America extending from Vancouver Island to northern California. At the CSZ, subduction of the oceanic Juan de Fuca and Explorer Plates below the continental North American Plate results in complex deformational processes and is the source of seismicity at all scales. This includes fault ruptures that occur within the crust of the over-riding plate, within the subducting slab, and on the megathrust interface between plates, all of which can cause damaging earthquakes; megathrust and crustal ruptures can also generate tsunamis. At smaller scales, an intriguing phenomenon is Episodic Tremor and Slip (ETS), representing periodic slow slip on/near the subduction interface accompanied by seismic tremor events. ETS occurs at a number of subduction zones worldwide, but was first recognized at the CSZ (Rogers and Dragert, 2003; Kao et al., 2005).

Specific geophysics research areas at UVic include ETS and other subduction-zone processes, active tectonics, earthquake mechanics, crust/mantle rheology, seismic inversion and tomography, induced seismicity, and earthquake and tsunami hazard analysis. Geophysical data acquisition involves marine cruises, land-based fieldwork, and aerial and drone surveys, as well as national and international seismic and geodetic networks and satellite radar imagery.

Geophysics research at UVic involves SEOS faculty members Stan Dosso, Lucinda Leonard, and Edwin Nissen, and a number of adjunct faculty members located at the nearby Pacific Geoscience Centre of the Geological Survey of Canada, including John Cassidy, Honn Kao, and Kelin Wang, among others. SEOS faculty and adjuncts work extensively with graduate students and post-doctoral scholars in their research, with typically about 15-20 such personnel involved at any one time. This article provides an overview, written by each group, of selected topics in geophysics research at UVic.

Seismic and tsunami hazard

Lucinda Leonard and her students and collaborators work on various aspects of earthquake and tsunami hazard assessment. Some on-going and recent research is briefly outlined below.

Study of active faults on Vancouver Island

We contribute to paleoseismic, remote-sensing, and near-surface geophysical investigation of active faults on Vancouver Island, British Columbia (BC). Paleoseismic trenching to about 2-3 m depth has revealed that, during the Holocene (within the last 11,700 years), at least three large earthquakes have occurred on the Leech River Fault (LRF; Morell et al., 2018; Harrichhausen et al., 2021), one on the XEOLXELEK-Elk Lake Fault (X-ELF; Harrichhausen et al., 2023), and five on the Beaufort Range Fault (Lynch et al., 2022). The first two of these faults pass through or close to the city of Victoria, while the third passes close to communities on central Vancouver Island. Figure 1 shows an 81-m long electrical resistivity tomography (ERT) profile, surveyed prior to paleoseismic trench excavation across the X-ELF, that reveals south-side-up reverse displacement beneath a topographic scarp. Slip on the LRF and X-ELF accommodates right-lateral transpression, consistent with inferred strain on related/connected offshore faults along strike to the east. Geodetic data analysis plays an important role in elucidating the style, rates, and drivers of deformation (e.g., Finley et al., 2019). Determination of the potential rupture extent, rupture style, and frequency of past and future earthquakes is critical, to enable assessment of both the seismic hazard and the tsunamigenic potential of these onshore-offshore fault systems in the northern Cascadia forearc.

Figure 1. (a) ERT profile across the north-facing scarp adjacent to the eastern shore of Elk Lake, near Victoria. (b) Interpretation of (a), with the paleoseismic trench outlined with the thin black line. The red line shows a shallow-dipping reverse fault observed in the trench. High-resistivity layer 1 (HR1) correlates with a colluvial wedge unit, lacustrine sand, and anthropogenic fill observed in the trench. Low-resistivity layer 1 (LR1) correlates with a glacial-marine clay. High-resistivity layer 2 (HR2) is interpreted to represent basalt bedrock or consolidated sediments that are vertically offset beneath the topographic scarp along a linear feature that dips between 45° and 55° toward the southwest. Low-resistivity unit 2 (LR2) is interpreted to represent a buried pipe aligned perpendicular to the ERT profile. (Modified from Harrichhausen et al., 2023.)

Tsunami hazard assessment

Active faulting is not confined to onland portions of the Cascadia forearc. There is a particular need to assess submarine crustal faults as potential sources of localized, but potentially hazardous, tsunami waves that could arrive onshore with little to no warning. We are carrying out geophysical data analysis for mapping and characterization of onshore-offshore and submarine fault systems beneath the Salish Sea (the marginal sea surrounding southern Vancouver Island consisting of the Strait of Juan de Fuca, Strait of Georgia, and Puget Sound) for earthquake and tsunami hazard assessment (e.g., Caston et al., 2020).

Subaerial and submarine landslides present an additional and poorly-understood source of tsunami hazard for much of the BC coast. This hazard is demonstrated by numerical modelling of a large, potential subaerial landslide off Orcas Island (Washington) in the Salish Sea, that would trigger damaging tsunami waves on both sides of the Canada/US border, including the Boundary Bay region south of Vancouver (Nemati et al., 2023; Figure 2). Seafloor landslide deposits attest to past events, which may have been triggered by large earthquakes on the nearby Skipjack Island fault zone.

Figure 2. Tsunami wave elevations (metres above reference level) simulated at various time steps following initial wave generation close to the Orcas Island (WA) subaerial landslide source (a-f), and in the Boundary Bay (BC) region (g-i). (Modified from Nemati et al., 2023.)

Remote sensing of earthquakes and faulting

Edwin Nissen and his group use innovative measurements of earthquakes and active faulting to better understand the mechanics of seismic rupture and the tectonics of regions of interest, most notably the Canadian Cordillera and the Arabia-Eurasia collision zone. We specialize in geodetic remote sensing tools that illuminate Earth’s surface using radar or laser pulses from a satellite, airplane, or drone, and map ground deformation or topography from the back-scattered returns. These data are combined with field-based geologic and geophysical measurements (working, for example, with Lucinda Leonard), seismogram analyses (with Honn Kao and others), and numerical models (with Kelin Wang).

Mapping earthquake deformation with satellite radar

Interferometric Synthetic Aperture Radar (InSAR) involves generating images of Earth’s surface on successive passes of a satellite in low-Earth orbit. By differencing radar phase measurements taken before and after an earthquake, the surface deformation can be mapped out, revealing the location, kinematics, and slip distribution of the responsible faulting. We have studied large earthquakes of interest around the world, but focus much of our work on Turkey and Iran, which occupy the Arabia-Eurasia collision zone and together account for about one sixth of all earthquake fatalities. We published the first study of the 2020 magnitude (M) 6.8 Elazığ earthquake in eastern Turkey, which was, at the time, the largest earthquake ever recorded on the East Anatolian Fault (Pousse-Beltran et al., 2020). In this earthquake, slip propagated across one fault segment boundary, but stopped well short of another, which, considered alongside nearby paleoseismic records, implies that the fault sometimes hosts far larger, multi-segment earthquakes. We were tragically proven correct on 6 February, 2023, with the M 7.8 East Anatolian Fault earthquake, which broke through several segment boundaries, killing 60,000 people (Tan et al., 2024). Much of our work in Iran has focused on the Zagros Mountains, one of the world’s most seismically-active continental mountain belts. Here, we documented the first clear case of anthropogenic earthquakes, induced by natural gas extraction (Jamalreyhani et al., 2021). The Zagros fold-and-thrust belt holds 90% of Iran’s gas reserves, so an awareness of this risk is crucial there.

Mapping fault geomorphology with airborne and drone lidar

The advent of sub-metre resolution topographic surveying has revolutionized active fault mapping, and our work has been at the forefront of showcasing these capabilities and driving the technology forward. Light detection and ranging (lidar) data, collected from crewed aircraft, can provide coverage of entire fault systems, even when the terrain is densely forested. Working with a wealth of newly-available lidar data, we have uncovered clear evidence for surface-rupturing earthquakes on the Denali and Tintina Faults in the Yukon and the Rocky Mountain Trench in BC, respectively (e.g., Finley et al., 2022). However, airborne lidar remains expensive to procure, and so we have also developed, at UVic, a state-of-the-art drone lidar system (Figure 3) which overcomes these limitations to provide cost-effective surveys of fault topography, at stunning spatial resolutions, even amongst dense forest cover (Salomon et al., 2024). An example of the exquisite geomorphology imaged with our system is shown in Figure 4, along the Denali Fault at Duke River, Yukon, where a major earthquake would threaten a critical international transportation corridor, the Alaska Highway. We have also applied stereo-photogrammetry to legacy air photo datasets, to map surface deformation in historical ruptures such as the 1971 San Fernando, California earthquake (Gaudreau et al., 2023).

Figure 3. Annotated photograph of the drone platform and lidar instrumentation developed at SEOS/UVic, from Salomon et al. (2024). INS stands for inertial navigation system, which comprises an IMU (inertial measurement unit) and GNSS (global navigation satellite systems) unit, used to track the location and orientation of the laser scanner to high precision.

Figure 4. (A) Hillshaded 0.5 m-resolution drone lidar digital terrain model along the Eastern Denali fault, amended from Salomon et al. (2024). Landforms, indicative of surface-rupturing earthquakes, are marked by white triangles. (B) Inset showing a 2-m high northeast facing scarp along a river terrace. The lower inset shows cross section B-B’ through the classified lidar point cloud, with green for vegetation, pink for ground returns, and yellow for unclassified points.

Earthquake hazard analysis

John Cassidy and Stan Dosso and their students are involved in many aspects of earthquake hazard analysis and mapping earth structure using passive-source seismology, with the goal to contribute to improvements in earthquake building codes across Canada. Three current/recent examples are summarized below.

Earthquake site response

Estimating the soil-column response to earthquake shaking at a site, including amplification and resonance effects, is an important component of seismic hazard evaluation. For example, soft soils often produce longer and stronger shaking than sites with bedrock or firm soil. One approach to assessing earthquake site response is based on using passive-array measurements of ambient seismic noise to estimate the shallow shear-wave velocity (Vs) profile (Molnar et al., 2013; Gosselin et al., 2017, 2022), which characterizes soil-column rigidity and behaviour in earthquake shaking. Our Vs estimation is based on Bayesian (probabilistic) inversion methods, which provide quantitative uncertainty analysis. As an example, Figure 5 shows Vs profiles computed using three different model parameterizations for a measurement site on the Fraser River Delta near Vancouver that is characterized by thick (> 100 m), unconsolidated sediments above the basement. The inversion results are compared to Vs measurements obtained using much more expensive, invasive methods involving boreholes and seismic cone penetration tests. Probabilistic inversion results, as shown here, can provide uncertainty analysis for geotechnical seismic-hazard site assessments required by building codes.

Earthquake building response

In addition to estimating earthquake site response, it is important to consider how buildings respond to earthquake shaking, as well as potential interactions between the site and building responses. One example is a collaborative project with earthquake engineers at the University of British Columbia (UBC) to record ambient-noise seismic data to evaluate the earthquake response of tall wood buildings, a new and increasingly-popular building type. Figure 6 illustrates this response in terms of resonant deformational modes of the 18-storey Brock Commons building on the UBC campus (Leishman et al., 2024). This work has revealed important differences between numerical model predictions of the fundamental-mode periods of the building response, compared to those observed in situ using ambient seismic data. These results indicate that the dynamic behaviour of tall wood buildings requires further research, including the potential for significant soil/structure interactions.

Figure 5. Bayesian inversion results for a Fraser River Delta site. (a), (c) and (e) show Vs marginal probability profiles for power-law, Bernstein-polynomial, and trans-dimensional model parameterizations, respectively, with zoom-ins of the grey rectangles shown in (g), (h) and (i). Circles indicate averages of invasive measurements with one standard-deviation error bars. (b), (d) and (f) show corresponding marginal profiles for the basement interface depth. (Modified from Gosselin et al., 2017.)

Figure 6. The first three deformational mode shapes identified for the Brock Commons building. (a) Mode 1 is translation in the y-direction (in/out of the page), (b) mode 2 is translation in the x-direction (along the length of the building), and (c) mode 3 is torsion. (Modified from Leishman et al., 2024.)

Locating episodic tremor and slip

Locating subduction-zone Episodic Tremor and Slip (ETS) events, particularly in depth, is crucial for understanding slip processes and potential connections to earthquakes. However, ETS is notoriously difficult to localize, as it produces emergent rumblings, rather than distinct seismic-wave arrivals, as from earthquakes. A new approach, referred to as Differential Travel-time Bayesian Inversion (DTBI), provides significantly-improved ETS depth location, as well as rigorous uncertainty quantification (Bombardier et al., 2023). This approach uses cross-correlations between seismic-array station recordings, to extract observed differential travel-times, travel-time predictions with a local 3D velocity model, and grid-based Bayesian localization. The DTBI approach was validated by considering small local earthquakes and comparing results with standard catalogue locations based on S-P wave arrival-time differences (a more informative method that does not apply to ETS due to the lack of clear impulsive arrivals). An example of ETS location for the CSZ is shown in Figure 7 for an ETS episode from 8-25 July, 2004, computed using both the new DTBI method and the well-established Seismic Scanning Algorithm (SSA; Kao and Shan, 2004). The DTBI results display a significantly greater degree of clustering in both epicentres and depth, which is interpreted as evidence of improved spatial accuracy.

Figure 7. Spatial and temporal distributions of 2024 ETS events located by DTBI are shown in (a) and (c), and by SSA in (b) and (d). Events are coloured by origin time and projected onto a margin-normal transect [red line in (a) and (b)] from within ±10 km, and shown in profiles (c) and (d). Histograms to the right of (c) and (d) show relative tremor density according to depth. The dashed grey line indicates the plate interface. Panels (c) and (d) involve no vertical exaggeration (Modified from Bombardier et al., 2023.)

Observational seismology

Honn Kao collaborates with Stan Dosso, Ed Nissen, and Lucinda Leonard on research projects in observational seismology that generally correspond to four themes: methodology development, earthquake source studies, tomographic inversion, and seismo-tectonics. Some current and recent focus areas are briefly described below.

Innovative seismological methodologies

The delineation of earthquake source parameters is a fundamentally important goal in seismology. Taking advantage of rapid advances in observational hardware and computational power, we are developing innovative methods and algorithms to revolutionize automated earthquake detection and location, providing unprecedented accuracy and efficiency. Building on the well-established Seismic Scanning Algorithm (SSA) (Kao and Shan, 2004), we developed the method of Seismicity-Scanning based on Navigated Automatic Phase-picking (S-SNAP; Tan et al., 2019) that combines SSA’s advantage of detecting small events and the high accuracy of locating earthquakes with precise arrival times of seismic phases. S-SNAP is capable of mapping complex spatiotemporal distributions of local and regional seismicity, as illustrated in Figure 8 for the 2019 Ridgecrest California earthquake sequence. Using artificial-intelligence (AI) based 3D image segmentation, we take earthquake detection and location to the next level with the Source Untangler Guided by AI image Recognition (SUGAR) method (Tan et al., 2024), which can outperform human analysts in accuracy, efficiency, and completeness, by a wide margin. We have also developed a sophisticated algorithm to resolve the long-standing challenge of determining accurate focal depths for small earthquakes without seismic stations in the immediate vicinity. The depth-scanning algorithm (Yuan et al., 2020) constructs synthetic waveforms of various depth phases at individual seismic stations, and searches for the best source-depth solution that simultaneously maximizes the number of depth-phase matches and minimizes the arrival-time residuals between synthetic and observed depth phases. This method is particularly useful for shallow earthquakes, where traditional location methods often fail due to the lack of data resolution.

Figure 8. The 2019 Ridgecrest California earthquake sequence (July 1-16) detected and located by S-SNAP. (a) Seismicity rate comparison among different catalogues. Time 0 corresponds to 0:0:0, July 1. S-SNAP’s performance (black line) is significantly better than other automatic methods (red and blue lines), and comparable to skilled analysts (TriNet in Hours ~190–280). (b) Spatiotemporal distribution of the epicentres. Circle size and colour indicate event magnitude and origin time, respectively. In total, 19827 events are detected and located, almost five times that of the Seiscomp3 catalogue. (Modified from Tan et al., 2020.)

Earthquake source processes and monitoring induced seismicity

Induced seismicity has become a focus of public concern in recent years, particularly in the continental interior where historical seismicity rates are relatively low. We study earthquake source processes, focusing specifically on the transition from the aseismic regime to seismic slip in response to fluid injections in the subsurface. The recent discovery of a new type of seismic signals that consist of both high- and low-frequency waveforms (referred to as Earthquakes characterized by Hybrid-frequency Waveforms, EHW, illustrated in Figure 9) provides new insight into the seismogenesis of induced earthquakes. We systematically investigate the relationship between induced-seismicity seismogenesis and industrial operations in Western Canada, as illustrated in Figure 10 (Kao et al., 2018a). Further, we study the source characteristics of induced earthquakes, including their focal mechanisms, Coulomb stress transfer, geomechanical response to fluid injection, and stress drop, to better understand the physical conditions of induced rupture (Gao et al., 2022; Wang et al., 2021a). Finally, we apply state-of-the-art AI analysis to identify the most influential factors that control induced seismicity in various areas of the Western Canada Sedimentary Basin (e.g., Wang et al., 2024). The results of this research contribute to amendments and improvement of regulations on industrial activities, including hydraulic fracturing, wastewater disposal, enhanced hydrocarbon recovery, and carbon sequestration, to mitigate the induced seismic risk from various injection operations (Kao et al., 2018b).

Figure 9. Representative examples of seismic waveforms generated by (a) typical injection-induced events and (b) earthquakes characterized by hybrid waveforms (EHW). The comparison of P and S phases demonstrates the relatively wider pulses of EHWs. (c) Scaling between corner frequency and seismic moment. Coloured and gray dots correspond to values estimated based on spectral ratio fitting of S-phase for EHWs and typical injection-induced events, respectively. Inset: stress drop as a function of well distance. The stress drop of an EHW generally is one order of magnitude lower than typical induced events. (Modified from Yu et al., 2021.)

Figure 10. Distribution of injection-induced earthquakes (IIE) in western Canada. (a) Background colour corresponds to the regional tectonic moment rate estimated from strain rate and thickness of the seismogenic layer. Squares with red outlines mark areas where local seismicity and injections co-exist, forming a ~150-km wide IIE band. (b) Number of IIE in each 20-by-20 km area as a function of injection volume and tectonic moment rate. A high number of IIE occur where either the injection volume is large, the tectonic moment rate is high, or both. (Modified from Kao et al., 2018.)

Seismic tomography

Seismic tomography is one of the most effective means to probe the internal 3D structure of the earth. We utilize ambient seismic noise data to study seismic velocity structures, ranging from the continental scale (Figure 11; Kao et al., 2013) to regional scale and local scale. The temporo-spatial distribution and characteristics of velocity anomalies provide observational constraints for a variety of important issues, including the configuration and evolution of tectonic motions, migration and flow patterns of subsurface fluids, and preliminary evaluation of critical minerals and energy resources. Tomography also provides an important tool in the newly-established field of environmental seismology, to study the impacts on the environment of various natural and anthropogenic activities (Kuponiyi et al., 2021).

Figure 11. Seismic tomography of Canada and adjacent regions using ambient seismic noise. The distribution of shear-wave velocity at the depths of (a) 5 km, (b) 25 km, and (c) 50 km is displayed in colour, with red and blue corresponding to low and high values, respectively. Dashed white lines on the 50 km image mark the location of cross sections shown in (d) and Kao et al. (2013). Small red circles and crosses correspond to the velocity profiles presented in Kao et al. (2013). (Modified from Kao et al., 2013.)


Seismological observations provide critical constraints on the study of kinematics and dynamics of plate tectonics. We conduct detailed seismo-tectonic studies of several significant earthquake sequences, along major plate boundary zones, including the 2012 M 7.8 Haida Gwaii earthquake (Kao et al., 2015) and the 2014 M 6.4 Nootka earthquake (Hutchinson et al., 2020), both off the BC coast, as well as the 2016 M 7.8 Kaikoura (New Zealand) earthquake (Tan et al., 2024a), the 2019 M 7.9 Ridgecrest (California) earthquake (Tan et al., 2020), and the 2023 M 7.8 Kahramanmaras (Turkey) earthquake (Tan et al., 2024b). Results of these studies not only fill critical knowledge gaps in our understanding of the rupture process of major earthquakes, but also contribute to the improvement of mitigation strategies for seismic and tsunamic hazards.

Subduction zone geodynamics

Kelin Wang and his team study the geodynamics of subduction zones and lithosphere, at both global and regional scales, using numerical models constrained by multi-disciplinary observations. We are particularly interested in learning how tectonics and Earth rheology control earthquakes and tsunamis. Together with collaborators, we have made a number of breakthrough findings over the past two decades. Recent/current work consists mainly of the following three areas, with contributions from (former) students, visiting students, and research associates.

Stress and strain in earthquake cycles

In earlier studies, we established the conceptual framework of subduction earthquake deformation cycles consisting of seismic rupture, postseismic stress relaxation of the viscoelastic mantle, afterslip of fault areas around the rupture zone, and relocking of the megathrust fault (e.g., Wang et al., 2012). A distinct feature of the postseismic phase of the cycle is the opposing motion of the offshore rupture area and other areas farther inland, with the dividing boundary of the opposing motion gradually migrating landward. In a subsequent global study, we found that the duration of the postseismic phase, characterized by how long it takes for this migrating boundary to arrive at a common reference location, scales with the size of the earthquakes (Figure 12; Sun et al., 2018). Our more recent studies focused on the vertical deformation of the postseismic phase, the observed complexity of which has puzzled the research community for decades. With global syntheses, we recognized that the pattern of vertical deformation associated with viscoelastic relaxation is common to all earthquakes, but vertical deformation associated with afterslip is specific to individual events. Using simple combinations of the two mechanisms, we explained the complex observations using rather simple models (Luo and Wang, 2022), including how the ubiquitous presence of a cold forearc mantle wedge causes near-arc postseismic uplift (Figure 13; Luo and Wang, 2021). We have also conducted various regional studies, including coseismic and postseismic deformation of several large megathrust earthquakes and interseismic deformation at the Cascadia, Chile, and Kuril subduction margins. In addition to subduction zones, we have studied fundamental physical processes of interseismic strain accumulation along strike-slip faults (Wang et al., 2021b) and the general mathematical theory of faulting in a viscoelastic Earth.

Figure 12. The “reference time” for modelled subduction earthquakes, defined as the time when the landward migrating dividing boundary of opposing motion passes R-50. R-50 is the map view location of the 50-km depth contour of the subduction interface. For a large earthquake, the reference time may vary along strike, so a bar is used to represent the range of values.

Figure 13. Model postseismic vertical deformation (curves) for four subduction earthquakes compared with observations (symbols). For the Chile event, only 8 years of postseismic data from coastal geological observations are available, but model uplift for 5 years after earthquake is also shown for comparison. Data for the other events are from Global Navigation Satellite Systems (GNSS). Data shown here are from within 200 km of display profile for the 2007 Bengkulu earthquake but 100 km for the other events. Circles for the Tohoku-oki event are seafloor GNSS measurements.

Thermal and petrologic processes in subduction zones

Our earlier works established a conceptual framework for the thermal regime of subduction zones (e.g., Wada and Wang, 2009), in which slab subduction gives rise to a cold forearc but the viscous flow of the mantle wedge induced by the subduction gives rise to a hot volcanic arc and backarc. In this framework, a 70-80 km maximum depth of decoupling (MDD) between the slab and the mantle wedge reconciles many thermal, petrologic, and seismological observations and inferences. One of the key consequences of the MDD and the resultant thermal transition is the presence of the cold and partially serpentinized forearc mantle wedge, of which the geodetic signature was examined in the aforementioned postseismic deformation study of Luo and Wang (2021). In a global study of the geological condition of ETS around the tip of this cold mantle wedge (Figure 14), Gao and Wang (2017) explained that the abundance of ETS is associated with the shallow dehydration of the slab in warm subduction zones such as Cascadia, Nankai, and Mexico. Related petrologic processes result in very high pore fluid pressure around the mantle wedge tip, responsible for the ETS. In this study, an important recognition is that the megathrust seismogenic zone and the ETS zone are spatially separated because of the thermally-controlled rheology, putting a strong constraint on the maximum depth of megathrust earthquakes in these subduction zones. In colder subduction zones, ETS is absent around the mantle wedge tip, and the megathrust rupture in great earthquakes can extend to much greater depths. As a case study of these colder subduction zones, Wang et al. (2020) examined the thermal regime of the region of the 2010 M 8.8 Maule, Chile, megathrust earthquake. Combining thermal modelling, petrologic knowledge, and seismic observations, we showed how the serpentinization of the cold forearc mantle wedge controls megathrust seismogenesis in the area. If the serpentinites along the base of the cold mantle wedge are entrained into the megathrust fault zone, the resultant serpentine fault gouge affects seismogenesis. Different types of serpentine minerals are stable to different temperatures, and thus depths, and as a fault gouge they exhibit different frictional behaviours. We found that the depth variation of the frictional behaviour associated with different serpentine minerals caused a depth variation in the distribution of small aftershocks following the 2010 Maule earthquake. We are presently expanding this research to investigate how similar processes may apply to large earthquakes.

Figure 14. Downdip variations in slip behaviour and relationship of ETS with the mantle wedge tip. The inset shows the inferred hydrogeological conditions around the mantle wedge tip. (From Gao and Wang, 2017.)

Subduction zone tsunami sources

Guided by the knowledge of tectonics and fault mechanics, we construct models of megathrust tsunami sources to investigate fundamental processes and to conduct hazard analyses. For the investigation of fundamental processes, Carvajal et al. (2022) conducted a systematic model study of tsunamigenic seafloor deformation at subduction zones, with variable megathrust rupture depths and degrees of trench-breaching. We demonstrated that any real earthquake exhibits a combination of two primary modes of tsunamigenic seafloor deformation (Figure 15). One is the seaward motion (rigid-body translation) of the sloping seafloor, enhanced by trench-breaching rupture. The other is the bulging of the seafloor (elastic thickening), enhanced by buried rupture. We found a fundamentally important offset relationship between the two primary modes, that is, an increase in one mode is always at the cost of the other. Consequently, given the same peak fault slip, as long as the rupture is mostly offshore, tsunami generation is insensitive to rupture depth and whether the rupture breaches the trench. The key to tsunami generation is the size of the fault slip, not the depth and trench-breaching nature of the rupture. This conclusion not only clarifies widely-held misconceptions but also has important implications for assessing tsunami hazards and risk. For assessing tsunami hazard, we have constructed megathrust tsunami source models specifically for the Cascadia margin (Gao et al., 2018), including buried rupture, splay faulting, and frontal thrusting as a variation of trench-breaching rupture in the situation of very thick trench sediments. We have substantively expanded this earlier work and are currently constructing thousands of source models for a U.S. national probabilistic tsunami hazard analysis project along the Cascadia margin.

Figure 15. Cartoon illustrating the trade-off between the two primary mechanisms that control tsunamigenic seafloor deformation. (a) Dominance of rigid-body translation in trench-breaching rupture. The blue shaded area illustrates the enhanced near-trench seafloor uplift due mainly to the horizontal motion of the sloping seafloor. (b) Dominance of elastic thickening in buried rupture. The blue shaded area illustrates the enhanced seafloor uplift due mainly to the horizontal contraction and resultant thickening of the upper plate. Dashed line indicates pre-earthquake seafloor geometry. (From Carvajal et al., 2022.)


Geophysics faculty at the University of Victoria, together with adjunct faculty from the Pacific Geoscience Centre, and the students they supervise, carry out leading research in many aspects of earthquake geophysics. This befits the university’s location proximal to the Cascadia Subduction Zone, representing the highest earthquake and tsunami hazard in Canada. Research includes diverse topics such as mapping active faults; using advanced remote sensing and numerical modelling to study stress/strain, deformation, and fault rupture processes including tsunamigenic potential; seismic-noise inversions for tectonic structure and for site and building earthquake-hazard assessments; high-precision location of earthquakes and ETS to study subduction-zone processes and induced seismicity from fluid injection; among many others. The overall objective of this broad program of research is to better understand earthquake processes and earthquake-related hazards, and to thereby contribute to mitigating the associated societal risks.

Bombardier, M., Dosso, S.E., Cassidy, J.F., and Kao, H., 2023. Tackling the challenges of tectonic tremor localization using differential traveltimes and Bayesian inversion, Geophysical Journal International, 234(1), 479–493,

Carvajal, M., Sun, T., Wang, K., Luo, H., and Zhu, Y., 2022. Evaluating the tsunamigenic potential of buried versus trench-breaching megathrust slip. Journal of Geophysical Research: Solid Earth, 127, e2021JB023722.

Caston, M., Leonard, L.J., Wang, K., Amouzgar, R., and Grivault, N., 2020. Investigating the tsunamigenic potential of crustal faults in the Strait of Georgia. American Geophysical Union Annual Fall Meeting, 2020AGUFMNH0140009C, Online, Dec 7-11.

Dokht, R., Smith, B., Kao, H., Visser, R., and Hutchinson, J., 2020. Reactivation of an intraplate fault by mine-blasting events: Implications to regional seismic hazard in Western Canada. Journal of Geophysical Research: Solid Earth, 125(6), e2020JB019933. doi:10.1029/2020JB019933.

Farahbod, A.F., Kao, H., Walker, D.M., and Cassidy, J., 2015. Investigation of regional seismicity before and after hydraulic fracturing in the Horn River Basin, northeast British Columbia, Canadian Journal of Earth Sciences, 52(2), 112-122, doi:10.1139/cjes-2014-0162.

Finley, T., Morell, K., Leonard, L., Regalla, C., Johnston, S.T., and Zhang, W., 2019. Ongoing oroclinal bending in the Cascadia forearc and its relation to concave-outboard plate margin geometry. Geology 47(2), 155-158,

Finley, T., Salomon, G., Nissen, E., Stephen, R., Cassidy, J. and Menounos, B., 2022. Preliminary results and structural interpretations from drone lidar surveys over the Eastern Denali fault, Yukon. In: Yukon Exploration and Geology 2021, K.E. MacFarlane (ed.), Yukon Geological Survey, p. 83–105.

Gao, X., and Wang K., 2017. Rheological separation of the megathrust seismogenic zone and episodic tremor and slip, Nature, 543, 416–419, 10.1038/nature21389.

Gao, D., Wang, K., Insua, T.L., Sypus, M., Riedel, M., and Sun, T., 2018. Defining megathrust tsunami sources for northernmost Cascadia, Natural Hazards, 94, 445-469,

Gao, D., Kao, H., Wang, B., Visser, R., Schultz, R., and Harrington, R.M., 2022. Complex 3D migration and delayed triggering of hydraulic fracturing-induced seismicity: A case study near Fox Creek, Alberta, Geophysical Research Letters, 49, e2021GL093979, doi:10.1029/2021GL093979.

Gaudreau, É., Hollingsworth, J., Nissen, E., and Funning, G.J., 2023. Complex 3‐D surface deformation in the 1971 San Fernando, California earthquake reveals static and dynamic controls on off‐fault deformation. Journal of Geophysical Research: Solid Earth, 128(3), p.e2022JB024985.

Gosselin, J.M., Dosso, S.E., Cassidy, J.F., Quijano, J.E., Molnar, S., and Dettmer, J., 2017. A gradient-based model parameterization using Bernstein polynomials in Bayesian inversion of surface-wave dispersion, Geophysical Journal International, 211(1), 528–540,

Gosselin, J.M., Dosso, S.E., Askan, A., Wathelet, M., Savvaidis, A., and Cassidy, J.F., 2022. A review of inverse methods in seismic site characterization, Journal of Seismology, 26, 781–821,

Harrichhausen, N., Finley, T., Morell, K.D., Regalla, C., Bennett, S.E.K., Leonard, L.J., Nissen, E., McLeod, E., Lynch, E.M., Salomon, G., and ISethanant, I., 2023. Discovery of an active forearc fault in an urban region: Holocene rupture on the XEOLXELEK-Elk Lake fault, Victoria, British Columbia, Canada. Tectonics, 42(12), e2023TC008170,

Harrichhausen, N., Morell, K.D., Regalla, C., Bennett, S.E.K., Leonard, L.J., Lynch, E.M., and Nissen, E., 2021. Paleoseismic trenching reveals late Quaternary kinematics of the Leech River fault: implications for forearc strain accumulation in northern Cascadia. Bulletin of Seismological Society of America, 111(2), 1110-1138,

Hutchinson, J., Kao, H., Riedel, M., Obana, K., Wang, K., Kodaira, S., Takahashi, T., and Yamamoto, Y., 2020. Significant geometric variation of the subducted plate beneath the northernmost Cascadia subduction zone and its tectonic implications as revealed by the 2014 Mw 6.4 earthquake sequence, Earth and Planetary Science Letters, 551, 116569, doi:10.1016/ j.epsl.2020.116569.

Jamalreyhani, M., Pousse‐Beltran, L., Büyükakpınar, P., Cesca, S., Nissen, E., Ghods, A., López‐Comino, J.Á., Rezapour, M., and Najafi, M., 2021. The 2019–2020 Khalili (Iran) earthquake sequence—Anthropogenic seismicity in the Zagros simply folded belt? Journal of Geophysical Research: Solid Earth, 126(12), p.e2021JB022797.

Kao, H., and Shan, S.-J., 2004. The Source-Scanning Algorithm: mapping the distribution of seismic sources in time and space, Geophysical Journal International, 157, 589-594.

Kao, H., Shan, S.-J., Dragert, H., Rogers, G., Cassidy, J.F., and Ramachandran, K., 2005. A wide depth distribution of seismic tremors along the northern Cascadia margin, Nature, 436, 841-844.

Kao, H., Behr, Y., Currie, C.A., Hyndman, R., Townend, J., Lin, F.-C., Ritzwoller, M.H., Shan, S.-J., and He, J., 2013. Ambient seismic noise tomography of Canada and adjacent regions: Part I. Crustal structures, Journal of Geophysical Research: Solid Earth, 118, 2013JB010535, doi:10.1002/2013JB010535.

Kao, H., Shan, S.-J., and Farahbod, A.M., 2015. Source characteristics of the 2012 Haida Gwaii earthquake sequence, Bulletin of the Seismological Society of America, 105, 1206-1218, doi:10.1785/0120140165.

Kao, H., Eaton, D.W., Atkinson, G.M., Maxwell, S., and Babaie Mahani , A., 2016. Technical meeting on the traffic light protocol (TLP) for induced seismicity: Summary and recommendations, Geological Survey of Canada, Open File 8075, 20 pp., doi:10.4095/299002.

Kao, H., Hyndman, R., Jiang, Y., Visser, R., Smith, B., Babaie Mahani, A., Leonard, L., Ghofrani, H., and He, J., 2018a. Induced seismicity in western Canada linked to tectonic strain rate: Implications for regional seismic hazard, Geophysical Research Letters, 45, 11104-11115, doi:10.1029/2018GL079288.

Kao., H., Visser, R., Smith, B. and Venables, S., 2018b. Performance assessment of the induced seismicity traffic light protocol (IS-TLP) for northeastern British Columbia and western Alberta, The Leading Edge, 37(2), 117-126, doi: 10.1190/tle37020117.1.

Kuponiyi, A. P., Kao, H., van Staal, C.R., Dosso, S.E., Cassidy, J.F., and Spence, G.D., 2017. Upper crustal investigation of the Gulf of Saint Lawrence region, eastern Canada using ambient noise tomography, Journal of Geophysical Research: Solid Earth, 122(7), 5208-5227, doi:10.1002/2016JB013865.

Kuponiyi, A.P., and Kao, H., 2021. Temporal variation in cultural seismic noise and noise correlation functions during COVID-19 lockdown in Canada, Seismological Research Letters, 92 (5), 3024-3034, doi:10.1785/0220200330.

Leishman, T., Ventura, C., Motamedi, M., Cassidy, J.F., and Dosso, S.E., 2024. Ambient vibration testing of Canada’s tallest wood frame building, International Operational Modal Conference Proceedings, 8 pp. (submitted).

Luo, H., and Wang, K., 2021. Postseismic geodetic signature of cold forearc mantle in subduction zones. Nature Geoscience, 14, 104–109.

Luo, H., and Wang, K., 2022. Finding simplicity in the complexity of postseismic coastal uplift and subsidence following great subduction earthquakes. Journal of Geophysical Research: Solid Earth, 127, e2022JB024471.

Lynch, E.M., Regalla, C., Bennett, S.E.K., Morell, K.D., Leonard, L.J., Harrichhausen, N., and Nissen, E., 2022. Holocene paleoseismicity in the northern Cascadia forearc: non-steady slip on the right-lateral oblique Beaufort Range fault, Vancouver Island, Canada. American Geophysical Union Annual Fall Meeting, 2022AGUFM.G35B0333L, Chicago, IL, Dec 12-16.

Molnar, S., Dosso S.E., and Cassidy, J.F., 2013. Uncertainty of earthquake site amplification via Bayesian inversion of surface seismic data. Geophysics, 78, WB37-WB48.

Morell, K.D., Regalla, C., Amos, C., Bennett, S., Leonard, L., Graham, A., Reedy, T., Levson, V., and Telka, A., 2018. Holocene surface rupture history of an active forearc fault redefines seismic hazard in southwestern British Columbia, Canada. Geophysical Research Letters 45(21), 11,605-11,611,

Nemati, F., Leonard, L.J., Thomson, R., Lintern, G., and Kouhi, S., 2023. Numerical modeling of a potential landslide-generated tsunami in the southern Strait of Georgia. Natural Hazards,

Ojo, A.O., Kao, H., Visser, R., and Goerzen, C., 2021. Spatiotemporal changes in seismic velocity associated with hydraulic fracturing-induced earthquakes near Fox Creek, Alberta, Canada. Journal of Petroleum Science and Engineering, 208, 109390, doi:10.1016/j.petrol.2021.109390.

Pousse‐Beltran, L., Nissen, E., Bergman, E.A., Cambaz, M.D., Gaudreau, É., Karasözen, E., and Tan, F., 2020. The 2020 Mw 6.8 Elazığ (Turkey) earthquake reveals rupture behavior of the East Anatolian Fault. Geophysical Research Letters, 47(13), p.e2020GL088136.

Rogers, G., and Dragert, H., 2003. Episodic tremor and slip on the Cascadia Subduction Zone: The chatter of silent slip. Science, 300(5627), 1942-1943.

Salomon, G., Finley, T., Nissen, E., Stephen, R., and Menounos, B., 2024. Mapping fault geomorphology with drone-based lidar. Seismica (in review).

Sun, T., Wang, T., and He, J., 2018. Crustal deformation following great subduction earthquakes controlled by earthquake size and mantle rheology, Journal of Geophysical Research: Solid Earth., 123, 5323–5345.

Tan, F., Kao, H., Nissen, E., and Eaton, D., 2019. Seismicity-scanning based on navigated automatic phase-picking, Journal of Geophysical Research: Solid Earth, 124(4), 3802-3818, doi:10.1029/2018JB017050.

Tan, F., Kao, H., Nissen, E., and Visser, R., 2020. Tracking earthquake sequence in real time: Application of Seismicity-Scanning based on Navigated Automatic Phase-Picking (S-SNAP) to the 2019 Ridgecrest, California Sequence, Geophysical Journal International, 223(3), 1511-1524, doi:10.1093/gji/ggaa387.

Tan, F., Kao, H., Yi, K.M., Nissen, E., Goerzen, C., Hutchinson, J., Gao, D., and Farahbod, A.M., 2024a. Next generation source detection by computer vision: Untangling the complexity of the 2016 Kaikōura earthquake sequence, Journal of Geophysical Research: Solid Earth (in revision).

Tan, F., Kao, H., and Nissen, E. 2024b. The Kahramanmaras (Turkey) earthquake multiplet sequence revealed by deep learning computer vision, Abstract presented at the 2024 Seismological Society of America Annual Meeting.

Wada, I. and Wang, K., 2009. Common depth of slab-mantle decoupling: Reconciling diversity and uniformity of subduction zones, Geochemistry, Geophysics, Geosystems, 10(10), 1-36.

Wang, B., Verdecchia, A., Kao, H., Harrington, R.M., Liu, Y., and Yu, H., 2021a. A study on the largest hydraulic fracturing induced earthquake in Canada: numerical modeling and triggering mechanism, Bulletin of the Seismological Society of America, 111, 1392-1404, doi:10.1785/ 0120200251.

Wang, B., Kao, H., Yu, H., Visser, R., and Venables, S., 2022a. Physical factors controlling the diverse seismogenic behavior of fluid injections in western Canada, Earth and Planetary Science Letters, 589, 117555, doi:10.1016/j.epsl.2022.117555.

Wang, B., Kao, H., Dokht, R.M.H., Visser, R., and Yu, H., 2022b. Delineating the controlling factors of hydraulic fracturing-induced seismicity in the Northern Montney Play, northeastern British Columbia, Canada, with machine learning, Seismological Research Letters, doi:10.1785/0220220075.

Wang, B., Kao, H., Yu, H., Visser, R., and Verdecchia, A., 2023. Quantitative evaluation of the competing effects of wastewater disposal and hydraulic fracturing on causing induced earthquakes: A case study of an M3.1 earthquake sequence in western Canada, Journal of Geophysical Research: Solid Earth, 128, e2022JB025048, doi:10.1029/2022JB025048.

Wang, B., Kao, H., Yu, H., Dokht, R.M.H., and Visser, R., 2024. Unveiling key factors governing seismogenic potential and seismogenic productivity of hydraulic fracturing pads: Insights from machine learning in the southern Montney Play, Earth and Planetary Science Letters, 626, 118511, doi:10.1016/j.epsl.2023.118511.

Wang, K., Hu, Y., and He, J., 2012. Deformation cycles of subduction earthquakes in a viscoelastic Earth, Nature, 484, 327-332.

Wang, K., Huang, T., Tilmann, T., Peacock, S. M., and Lange, D., 2020. Role of serpentinized mantle wedge in affecting megathrust seismogenic behavior in the area of the 2010 M=8.8 Maule earthquake. Geophysical Research Letters, 47, e2020GL090482. doi:10.1029/2020GL090482.

Wang, K., Zhu, Y., Nissen, E., and Shen, Z.-K., 2021b. On the relevance of geodetic deformation rates to earthquake potential, Geophysical Research Letters, 48, e2021GL093231.

Yu, H., Harrington, R.M., Kao, H., Liu, Y., and Wang, B., 2021. Fluid-injection induced earthquakes characterized by hybrid-frequency waveforms manifest the transition from aseismic to seismic slip, Nature Communications, 12, 6862, doi: 10.1038/s41467-021-26961-x.

Yu, H., Kao, H., Wang, B., and Visser, R., 2022. Long-term fluid injection can expedite fault development: Riedel shear structure illuminated by induced earthquakes in Alberta, Canada, Journal of Geophysical Research: Solid Earth, 127, e2022JB025126, doi:1029/e2022JB025126.

Yuan, J., Kao, H., and Yu, J., 2020. Depth-Scanning Algorithm: Accurate, automatic and efficient determination of focal depth for local and regional earthquakes, Journal of Geophysical Research: Solid Earth, 125, e2020JB019430, doi:10.1029/2020JB019430.

About the Authors

Dr. Stan Dosso ( obtained BSc and MSc degrees in physics and applied mathematics from the University of Victoria in 1982 and 1985, respectively, and his PhD in geophysics from the University of British Columbia in 1990. He was a Defence Scientist in Arctic Ocean Acoustics with the Defence Research Establishment Pacific from 1990-1995, before taking up an Ocean Acoustics Research Chair with the School of Earth and Ocean Sciences, University of Victoria, where he is a professor and was the Director of the School from 2016-2021. His research interests involve geophysical inverse problems in ocean geoacoustics and earthquake seismology, with a focus on the development and application of quantitative Bayesian inference methods. Dr. Dosso is a Fellow of the Acoustical Society of America (ASA), which he currently serves as President (2022–2025), and was the recipient of the ASA Silver Medal and Medwin Prize in Acoustical Oceanography, and recipient of the Science Teaching Excellence Award from the University of Victoria.

Dr. Lucinda Leonard ( received her B.A. (Honours) in Geology from Trinity College Dublin, Ireland, in 2001, and her PhD in Geophysics from the University of Victoria in 2006. She was an NSERC Visiting Fellow with Natural Resources Canada from 2008 to 2011, and a contract Research Scientist from 2011 to 2013. She has been a Limited-term Assistant Professor at the School of Earth and Ocean Sciences, University of Victoria, since 2013, and is currently partially seconded to Ocean Networks Canada. Her research interests include the study of active tectonics using geodetic and seismological data, earthquake and tsunami hazard assessment and mitigation, and the use of shallow geophysical methods to investigate the subsurface for neotectonic and groundwater applications.



Dr. Edwin Nissen ( is a Professor and Canada Research Chair in Geophysics at University of Victoria. After growing up in London (England), he got hooked on geology during his B.A. and M.Sc. at Cambridge (2000-2004) and honed in on seismology during his D.Phil. at Oxford (2004-2008). Here, he specialized in mapping earthquake deformation using satellite radar, with a regional focus on Iran and Mongolia. Following a three-year post-doc back in Cambridge, he hopped the Atlantic in 2011, working first as an Exploration Fellow at Arizona State University and subsequently as an Assistant Professor in Geophysics at Colorado School of Mines (2012-2016). In the U.S., he developed an interest in airborne laser scanning data and helped to pioneer the growing field of differential lidar. Ed has served on advisory boards for the international geodesy consortium UNAVCO, the U.S. National Center for Airborne Laser Mapping (NCALM), and the new Cascadia Region Earthquake Science Center (CRESCENT), as well as on review panels for the U.S. Geological Survey, the U.S. National Science Foundation, and Canada’s NSERC. He also won UVic’s Faculty of Science Award for Research Excellence in 2022. Ed is passionate about outreach and leads a “citizen seismology” project ( that is instrumenting BC schools with miniaturized seismometers for use in the classroom as well as for science.

Dr. John Cassidy ( is an earthquake seismologist and Senior Research Scientist with Natural Resources Canada in Sidney, BC. He leads the Geological Survey of Canada’s national “Assessing Earthquake and Volcanic Geohazards Project” and is an Adjunct Professor at the University of Victoria, School of Earth and Ocean Sciences where he teaches courses and supervises graduate students. Dr. Cassidy specializes in earthquake hazard studies and earth structure studies, and during the past 30 years has published more than 240 scientific articles. John serves as Co-Chair of the British

Columbia Seismic Safety Council, and in 2010 was invited to travel through the hardest-hit parts of Chile following the M8.8 earthquake and tsunami as a member of the Canadian Association of Earthquake Engineers Chile Earthquake Reconnaissance Team. John was elected to the Board of Directors for the Canadian Association for Earthquake Engineering in 2021 and he works closely with the engineering community and emergency management organizations that utilize the results of earthquake science to help reduce the impacts of future earthquakes.


Dr. Honn Kao (P.Geo.) ( obtained his BSc in Geophysics from the National Central University, Taiwan, in 1985, and his MSc and PhD in Geophysics from the University of Illinois at Urbana-Champaign in 1991 and 1993, respectively. He worked at the Institute of Earth Sciences, Academia Sinica, Taiwan, as an Assistant Research Fellow (1993-1996), and was promoted to Full Research Fellow in 2000. He joined the Geological Survey of Canada (GSC) as a Research Scientist in 2002. He has been an Adjunct Professor at the School of Earth and Ocean Sciences, University of Victoria since 2006. Currently, he is the Section Head of GSC-Pacific’s Section of Seismology, Tectonophysics, and Volcanology and the Leader of Natural Resources Canada’s Induced Seismicity Research Project.

Dr. Kelin Wang ( obtained his BSc in Geology in 1982 from Peking University and PhD in geophysics in 1989 from the University of Western Ontario. He joined the Pacific Geoscience Centre, Geological Survey of Canada, in 1990. He has been an Adjunct Professor at the University of Victoria since 1997 and was named Honorary Research Professor in 2017. Most of his current research is on the geodynamics of subduction zones and related earthquake and tsunami hazards, but he has also worked on a range of other topics regarding the thermal, mechanical, and hydrogeological processes of Earth’s lithosphere. Dr. Wang is a Fellow of the American Geophysical Union and recipient of the J. Tuzo Wilson Medal of the Canadian Geophysical Union. He was also the Birch Lecturer of the American Geophysical Union in 2015.