The seismic signal generated by the underground nuclear explosion, Long Shot, has been compared with seismic signals of earthquake origin and found to be similar on a regional scale. Negative Long Shot magnitude residuals are associated with areas of recent tectonic activity as are late arrivals, while positive Long Shot magnitude residuals and early arrivals have been found to be associated with tectonically stable regions. These trends are coincident with those indicated by data from other seismic events. The more detailed comparison of Long Shot and earthquake magnitude residuals at Penticton and Port St. James indicates that the Long Shot residuals also reflect the location of the source. At these stations, earthquakes with distances and azimuths comparable to Long Shot exhibit magnitude residuals that are most similar to those of Long Shot. The magnitude residuals at the University of British Columbia exhibit the same dependence on source parameters although a direct comparison with Long Shot could not be made. An examination of earthquake travel time residuals at Penticton and Fort St. James also indicates the same dependence on source location. Long Shot surface waves indicate an average unified magnitude of 5.1 at Canadian stations as compared with an average unified magnitude of 6.0 from body waves at the same stations. The comparison of the power spectra of Long Shot and earthquakes at Leduc and Victoria indicates relatively more energy at high frequencies from Long Shot than from earthquakes. This variation in spectral decrement is interpreted as an effect of the different source mechanisms. The spectrum of Long Shot at Rocky Mt. House appeared to be anomalous as it had a significantly larger spectral decrement than at the other stations and was indistinguishable from the spectra of earthquakes recorded at Rocky Mt. House. The trend of the power spectra also appear to be partially determined by the crustal and upper mantle structure in the vicinity of the station. The effect of the source parameters and travel path is also indicated by a tendency for the spectral decrement to increase with increased distance to source and with increased depth. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

In this thesis, we present the first detailed, long-term seismotectonic studies of oceanic ridge-transform systems. The proximity of the Juan de Fuca plate to a network of broadband seismic stations in western North America provides a unique synergy of interesting tectonic targets, high seismicity, and recording capabilities. Our main tools are earthquake source parameters, determined by robust waveform modeling techniques, and precise earthquake locations, determined by joint epicenter relocation. Regional broadband data are used to invert for the source moment tensors of the frequent, moderate-sized (M ≥ 4) earthquakes; this analysis began 1994. We include Harvard centroid moment-tensors available since 1976 for larger (M ≥ 5) earthquakes. Two studies comprise the main part of this thesis. In the first, we determine the current tectonics of Explorer region offshore western Canada. Earthquake slip vector azimuths along the Pacific-Explorer boundary require an independent Explorer plate. We determine its rotation pole and provide a tectonic model for the plate's history over the last 2 Ma. Plate motion changes caused distributed deformation in the plate's southeast corner and caused a small piece in the southwest corner to transfer to the Pacific plate. Capture of the plate fragment indicates that preserved fragments not necessary represent entire microplates. In the second study, we investigate seismicity and source parameters along the Blanco Transform Fault Zone (BTFZ). The deformation style-- strike slip and normal faulting-- correlates well with observed changes in BTFZ's morphology. We infer that Blanco Ridge probably consists of two fault segments, that several parallel faults are active along BTFZ's west part, and that Cascadia Depression possibly is a short spreading center. The slip distribution along the BTFZ is highly variable, although seismicity could account for the full plate motion rate along the entire BTFZ. The final part is a short study where we locate an earthquake in the tectonically active Mendocino triple junction region offshore northern California using land and offshore data. The precise location may be useful as a master event for relocating other earthquakes. / Graduation date: 1999

A comprehensive seismotectonic model is developed to explain the seismicity of British Columbia. In order to do this extensive revisions are made to location and magnitude parameters in the Canadian Earthquake Data File. Fault plane solutions are calculated for all earthquakes possible and all mechanisms previously calculated are examined and upgraded where necessary. It is proposed that the subcrustal suite of earthquakes in the Puget Sound and southern Vancouver Island region are a result of strain caused by phase changes in the descending oceanic lithosphere of the subducting Juan de Fuca plate. The crustal earthquakes above the deeper seismicity can be explained with an oblique subduction model. The distribution of seismicity, the amount of seismicity and the focal mechanisms support these interpretations. The large earthquakes of central Vancouver Island are probably a result of the interaction of the Explorer Plate with the overriding America Plate. In the southern Queen Charlotte Islands thrusting components in the fault plane solutions confirm there is an element of convergence across the pacific/America boundary in this region. The distribution of seismicity suggests all relative plate motion is presently occurring along the Queen Charlotte fault. The Quaternary volcanoes of British Columbia show little correlation with the seismicity pattern except for the eastern end of the Anahim volcanic belt. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

The Queen Charlotte Islands are located east of the seismically active Queen Charlotte transform fault zone which separates the Pacific and North American plates. The fault zone is the locus of significant seismic activity and is distinguished bathymetrically by two steep scarps bounding a 15 to 25 km wide terrace. To better define regional seismicity characteristics, 16 portable seismographs and 6 ocean bottom seismographs were operated for 9 weeks and 5 days, respectively, during June to August 1983. Three hundred and seventeen events were detected; 130 events that were recorded on 3 or more stations have been located. Twenty' events were identified as possible blasts. Ninety-two of the located earthquakes lie along the Queen Charlotte transform fault zone, most within the 1949, Ms = 8.1, earthquake rupture zone along the inner scarp of the terrace. However, several earthquakes are located on the terrace and a few are aligned with the outer scarp where no activity has previously been observed. Most of the activity is well constrained to be less than 15 km in depth. Two areas of low seismicity were observed along the fault zone. Only two earthquakes occurred in the documented seismic gap bound on the north by the 1949 rupture zone and on the south by the 1970 M = 7.0 earthquake. They were both at -the southern tip of the gap. A similar region of low activity was observed for the fault along Graham Island. No major seismicity (M > 4.0) has been located in this region since the 1949 earthquake. Composite fault plane mechanism solutions were determined for five clusters of events along the fault zone. Events to the northwest of Graham Island are consistent with strike-slip motion along a fault in the direction of the Queen Charlotte transform fault. The four remaining clusters were located along Moresby Island. The mechanisms for these are dominated by thrust faulting with a component of compressional stress trending north-south. These events are interpreted as the result of oblique convergence between the Pacific and North American plate. Significant seismicity was located east of the main Queen Charlotte transform fault zone. Eighteen earthquakes, the largest ML = 3.8, were located in northeastern Graham Island and adjacent Hecate Strait - Dixon Entrance area. None could be associated with known faults. The focal depth of these events is well constrained within the crust so they could not be associated with a subducted plate. A composite fault plane mechanism solution determined for some of these earthquakes indicates a thrusting mechanism with north-south trending compressional stress. One event with a well constrained solution at a shallow focal depth occurred in southeastern Hecate Strait. This event could be associated with crustally pervasive faults identified in Hecate Strait. A magnitude scale based on the coda length of the earthquake signal was determined. Magnitudes were calculated for 265 of the events recorded during the study. For the complete data set a b-value of 0.55 ± 0.05 was determined. This value is significantly lower than values from other studies in the Canadian Cordillera, indicating that a greater percentage of the total number of earthquakes occurs at the higher magnitudes. However, the short period of recording and large magnitude seismic activity (4 earthquakes with ML > 3.8 in 9 weeks) might have biased the estimate toward a low value. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

The implications of the acceleration data recorded during the September 19, 1985 Mexican earthquake for seismic design in Canada are investigated by determining if the deep deposits of the Fraser Delta could cause large amplification of earthquake motions. The conditions for amplification of low level incoming ground motions at deep sites are identified; in particular, the critical role of variation in shear modulus with shear strain. The current procedure for determining site specific ground motions is evaluated and major sources of uncertainty in the results identified. Criteria for selecting representative input motions for site response studies are recommended based on analyses of Mexico City sites. A comparative study of sites in the Fraser Delta area of British Columbia showed that offshore subduction earthquakes should be considered when developing design spectra for deep sites in the Delta. / Applied Science, Faculty of / Civil Engineering, Department of / Graduate

It is being increasingly realized that there exists an interrelationship between seismicity and rock slope failures. Possible chronological clustering of rock avalanches in the Fraser River corridor was investigated to determine if a common seismic trigger existed. It was determined that the events occurred throughout the Holocene indicating that either these slides were not seismically triggered or that seismic triggers were chronologically unrelated. Cascadia Subduction Zone earthquakes are believed to have occurred throughout the Holocene (Adams, 1989; Atwater, 1987; Hull, 1987). The ages of the earthquakes were compared to the ages of rock avalanches in the Fraser River corridor, but no distinct correlation could be made. The lack of distinct correlation between large rock avalanches in Fraser Corridor and paleoseismicity, and the absence of event clustering, indicated either seismicity was not a factor, or that these rock avalanches may not have been susceptible to seismic triggering. A stability study of Cheam Slide was performed to investigate the susceptibility of large rock avalanches to earthquake triggering. The results suggested that the seismic susceptibility of a slope is closely linked to the displacement the slope must undergo for failure to take place. A large critical displacement may render the slope relatively insensitive to seismic triggering, while a low critical displacement may result in high seismic susceptibility. A comparison was made between the effects of seismic and pore pressure related triggering. The results indicated that a high critical displacement slope, which is close to failure, may be more likely to fail by high pore pressures than by seismic loading. Low critical displacement slopes which are stable enough to surviving hydrodynamic loading may, because of their susceptibility to seismic triggering, pose the greatest hazard. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

A study of the dispersion of long-period Rayleigh waves was proposed in order to acquire additional knowledge about the lithospheric structure of the Vancouver Island region. Three portable, long-period seismographs were designed and built to operate in field conditions. An array was established with stations at Victoria, Vancouver and Quadra Island; during six weeks of operation, several earthquakes were recorded, of which two were aligned conveniently with the array. Another network with stations at Ucluelet, Quadra Island and Victoria was set up later; no usable data were recorded. the earthquake record was chosen for analysis, and group-velocity dispersion calculations were begun. Echo resolution on these data and a lack of ether usable records prevented the continuation of the data processing. No conclusions are drawn about the lithospheric structure; however, recommendations are made regarding possible future experimentation with the existing apparatus. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

Over the past decade, patterns of horizontal crustal motion observed along the Cascadia subduction zone (CSZ) from Global Positioning System (GPS) measurements have been used to derive locked subduction zone models with varying geometry and coupling factors. Although vertical crustal deformation estimates have been less abundant and less accurate than horizontal component observations, they provide key constraints to the models for estimating the extent of rupture for the next subduction thrust earthquake. In order to provide updated model constraint estimates, the contemporary vertical deformation pattern across the northern Cascadia margin was investigated through the combined application of GPS, repeated leveling, precise gravity, and monthly mean sea level measurements across southern Vancouver Island and repeated leveling on the mainland. To the first order, these estimates are consistent with across-margin tilt predictions from current dislocation models for the region. In their details, however, they reflect a more complex system than suggested by the simple models. Minor landward tilt across the margin at Tofino determined from the re-analyses of -8 years of continuous vertical GPS positions, -40 years of monthly mean sea levels and long-term time (decadal) intervals of repeat leveling surveys is distinctly different than the -3 mm yr'l of landward tilt observed at Neah Bay. While this difference may be minimized by allowing for a small amount of tilt induced at the southern stations from northward migration of the Cascadia forearc, differences in tilting of 3-4 mm yr'l between short- and long-term estimates of repeat leveling at Bamfield are attributed to transients. To a lesser degree, elevation changes across the margin at Tofino may also illustrate transients. As well, distinct differences in the magnitude of vertical deformation for stations to the north and south of Barkley Sound suggest that differential deformation may be occurring along the margin. Similarly, while repeat relative gravity measurements across the margin at Tofino indicate 3-7 mm yr-' of seaward tilt (at odds with results from all other methods), a temporal dependence of vertical deformation might also be evident from the long-term versus short-term tilt rates. However, although repeat absolute gravity estimates between 1995 and 2002 indicate little across-margin tilt, consistent with continuous GPS results, differences between the time series at the Ucluelet absolute gravity and GPS stations indicate that gravity observations could be influenced by episodic mass redistribution beneath western Vancouver Island. This suggests that gravity results might not be directly comparable to estimates from other geodetic methods in determining uplift rates. Extension of the vertical deformation profile eastward into the backarc using repeat leveling surveys indicates a broad region of uplift in the Pemberton area with respect to the coast, which is consistent with the vertical component at the continuous GPS station WSLR. Current dislocation models cannot account for the observed deformation. Therefore, modification of one model was attempted in which a weaker crustal zone, coincident with high heat flow near the Garibaldi Volcanic Arc, was included. A poor fit to the observed deformation rates indicates that further refinements must be made to such a model. Nonetheless, these results suggest a complex system of strain accumulation in the northern CSZ, which may result from a greater 3- dimensionality of the tectonic controls than current dislocation models of the region employ.

The 1980 Vancouver Island Seismic Project (VISP) was conducted to investigate lithospheric structure associated with the underthrusting oceanic Juan de Fuca plate and the overriding continental America plate. The principal components of the survey were: (l) an onshore-offshore refraction line, which was approximately perpendicular to the continental margin (line 1), and (2) a refraction line which ran along the length of Vancouver Island approximately parallel with the continental margin (line IV). Lines I and IV were originally interpreted by Spence el a.1. (1985) and McMechan and Spence (1983), respectively. However since the original interpretations of these lines, deep multichannel seismic reflection data have been obtained on southern Vancouver Island as part of the 1984 LITHOPROBE project and off the west coast of the island during a marine survey in 1985. This study was undertaken to resolve differences between the subsurface structures proposed in the original interpretations of lines I and IV and those suggested by the more recently acquired deep reflection data. The vertical two-way traveltimes to prominent reflectors, observed in the onshore-offshore deep reflection data, were used as a constraint in constructing velocity models which are consistent with both the reflection and refraction data. The traveltimes and amplitudes observed in the VISP refraction data were modeled using a two-dimensional raytracing and asymptotic ray theory synthetic seismogram routine. The principal difference between the model originally interpreted for line I and the revised model involves the introduction of a twice repeated sequence of a low velocity zone (≈ 6.4 km/s) above a thicker high velocity zone (≈ 7.1 km/s) for the underplated region directly above the subducting Juan de Fuca plate in place of the single high velocity block underlain by a thick low velocity zone. The revised model for line IV is significantly different from the originally interpreted model. The two low-high velocity zones of line 1 are continued along the length of the island at depths between 10 and 35 km. Below this, the structure of the subducted plate is included to maintain consistency with the revised model developed for line 1. Additional features of the revised onshore-offshore model corresponding to line 1 include an oceanic lithosphere that dips approximately 3° beneath the continental slope, then 14° to 16° beneath the continental shelf and Vancouver Island, and an average velocity for the upper oceanic mantle of 8.22 km/s. Two separate two-dimensional models were needed to explain the data collected along line IV as a result of considerable azimuthal coverage due to a 30° change in profile direction. The revised models developed for line IV are consistent with the revised model developed for line 1. The velocity in the upper 10 km ranges from 5.5 km/s to approximately 6.7 km/s. Below 10 km the velocity structure is consistent with that interpreted for line 1 and shows some variations along strike of the subduction zone. Several possible interpretations can be made for the origin of the sequence of layers directly above the subducting plate beneath Vancouver Island. The two favored interpretations are: (1) a. three stage tectonic process consisting of: stage 1 — offscraping of sediment from the top of the subducting plate forms the uppermost low velocity layer in the sequence; stage 2 — an imbricated package of mafic rocks derived by continuous accretion from the top of the subducting oceanic crust forms the first high velocity layer; and stage 3 — stages 1 and 2 repeat themselves with stage 2 currently occurring; or (2) remnant, pieces of oceanic lithosphere left stranded above the current subducting plate during two previous episodes of subduction in which the subduction thrust jumped further westward isolating the remnant. The revised model along line IV indicates that this process of subduction underplating could have been a pervasive feature of this convergent margin. / Science, Faculty of / Physics and Astronomy, Department of / Graduate