Subducting Juan De Fuca Plate Research Articles

The fate of the Juan de Fuca plate: Implications for a Yellowstone plume head

Beneath the Pacific Northwest the Juan de Fuca plate, a remnant of the Farallon plate, continues subducting beneath the North American continent. To the east of the Cascadia subduction zone lies the Yellowstone Hotspot Track. The origins of this track can be traced back to the voluminous basaltic outpourings in the Columbia Plateau around 17 Ma. If these basalts are the result of a large melting anomaly rising through the mantle to the base of the North American continent, such as a mantle plume head, the anomaly would need to punch through the subducting Juan de Fuca slab. Here, we use teleseismic body-wave travel-time tomography to investigate the fate of the subducted slab and its possible interaction with a plume head. Our dataset is derived from the Oregon Array for Teleseismic Study (OATS) deployment in Oregon and all other available seismic data in this region during the same period. In our JdF07 models, we image the subducted Juan de Fuca plate in the mantle east of the Cascades beneath Oregon, where the slab has not been imaged before, to a depth of 400 km but no deeper. The slab dips ∼ 50°E and has a thickness of ∼ 75 km. Immediately beneath the slab, we image a low velocity layer with a similar geometry to the slab and extending down to at least ∼ 575 km depth in the V s model. The total length of the high velocity slab is ∼ 660 km, about 180 km longer than the estimated length of slab subducted since 17 Ma. Assuming similar slab geometry to today, this 180 km length of slab would reach ∼ 60 km depth, comparable to the thickness of continental lithosphere. We propose that the absence of the slab below 400 km today is due to the arrival of the Yellowstone plume head ∼ 17 Ma, which destroyed the Juan de Fuca slab at depths greater than the thickness of the continental lithosphere. Given this scenario, the low velocity anomaly beneath the slab is likely the remnant plume head material which has been pulled down by traction with the subducting plate. The amplitude of the observed low velocity anomaly is comparable with that expected for plume head material 100–300 °C hotter than the surrounding asthenosphere.

Surface wave tomography of the western United States from ambient seismic noise: Rayleigh wave group velocity maps

We have applied ambient noise surface wave tomography to data that have emerged continuously from the EarthScope USArray Transportable Array (TA) between October 2004 and January 2007. Estimated Green's functions result by cross‐correlating noise records between every station‐pair in the network. The 340 stations yield a total of more than 55,000 interstation paths. Within the 5‐ to 50‐s period band, we measure the dispersion characteristics of Rayleigh waves using frequency‐time analysis. High‐resolution group velocity maps at 8‐, 16‐, 24‐, 30‐, and 40‐s periods are presented for the western United States. The footprint of the TA encloses a region with a resolution of about the average interstation spacing (∼70 km). Velocity anomalies in the group velocity maps correlate well with the dominant geological features of the western United States. Coherent velocity anomalies are associated with the Sierra Nevada, Peninsular, and Cascade Ranges, Great Valley, Salton Trough, and Columbia basins, the Columbia River flood basalts, the Snake River Plain and Yellowstone, and mantle wedge features associated with the subducting Juan de Fuca plate.

Regional P wave velocity structure of the Northern Cascadia Subduction Zone

This paper presents the first regional three‐dimensional P wave velocity model for the Northern Cascadia Subduction Zone (SW British Columbia and NW Washington State) constructed through tomographic inversion of first‐arrival traveltime data from active source experiments together with earthquake traveltime data recorded at permanent stations. The velocity model images the structure of the subducting Juan de Fuca plate, megathrust, and the fore‐arc crust and upper mantle. Beneath southern Vancouver Island the megathrust above the Juan de Fuca plate is characterized by a broad zone (25–35 km depth) having relatively low velocities of 6.4–6.6 km/s. This relative low velocity zone coincides with the location of most of the episodic tremors recently mapped beneath Vancouver Island, and its low velocity may also partially reflect the presence of trapped fluids and sheared lower crustal rocks. The rocks of the Olympic Subduction Complex are inferred to deform aseismically as evidenced by the lack of earthquakes within the low‐velocity rocks. The fore‐arc upper mantle beneath the Strait of Georgia and Puget Sound is characterized by velocities of 7.2–7.6 km/s. Such low velocities represent regional serpentinization of the upper fore‐arc mantle and provide evidence for slab dewatering and densification. Tertiary sedimentary basins in the Strait of Georgia and Puget Lowland imaged by the velocity model lie above the inferred region of slab dewatering and densification and may therefore partly result from a higher rate of slab sinking. In contrast, sedimentary basins in the Strait of Juan de Fuca lie in a synclinal depression in the Crescent Terrane. The correlation of in‐slab earthquake hypocenters M > 4 with P wave velocities greater than 7.8 km/s at the hypocenters suggests that they originate near the oceanic Moho of the subducting Juan de Fuca plate.

Low Potential for Large Intraslab Earthquakes in the Central Cascadia Subduction Zone

The Cascadia subduction zone (csz) can be divided into three distinct sections based on the characteristics of intraslab seismicity. Based on a 150-year historical record, no moderate-to-large intraslab earthquakes of moment magnitude ( M ) 5.5 or greater have occurred within the subducting Juan de Fuca plate of the central csz from south of the Puget Sound in northwestern Washington to the Oregon–California border. Also very few intraslab earthquakes as small as M 3 have been instrumentally located within the central csz since 1960, and a Wadati–Benioff zone is not apparent. In the southern csz beneath northwestern California, a Wadati–Benioff zone is present to a depth of about 40 km, but no large Gorda block earthquakes have been observed in the downgoing slab, although large events have occurred near the trench axis. In contrast, the Puget Sound region within the northern csz has been repeatedly shaken by large intraslab earthquakes of M ≥6.5 in the depth range of 40 to 60 km, such as the recent 2001 M 6.8 Nisqually event. A critical question addressed in this article is what is the potential for such large, shallow intraslab earthquakes in the central csz beneath western Oregon and southwestern Washington? I have evaluated the available information on the thermal and physical properties, geometry, and historical and contemporary seismicity of the central csz, and performed thermal modeling. Based on these analyses and comparisons with other subduction zones worldwide, the lack of shallow intraslab earthquakes in the central csz is not unusual. The hot temperatures (>500°C) within the Juan de Fuca plate, particularly below a depth of 40 km where large events are expected, are not conducive to earthquake generation, resulting in either the complete absence of M ≥6.5 shallow intraslab earthquakes or long recurrence intervals (hundreds of years) between such events. Temperatures appear to be sufficiently high in the central csz so that no Wadati–Benioff zone can exist even at shallow depths (<40 km). The young plate age, slower convergence rate, and the insulating effect of the Siletz terrane above the plate are factors that probably lead to the hot temperatures in this portion of the slab. The variability in the maximum depth of the Wadati–Benioff zone along the csz, 60 km beneath the Puget Sound, 40 km within the subducting Gorda block, and essentially zero in the central csz, reflect the differing temperature conditions, that is, the cutoff temperature varies with depth and rock composition, and also the potential for large shallow intraslab earthquakes. In addition to the effects of temperature, the level of tectonic stresses, which vary along the length of the csz, must also be a factor in controlling the occurrence of large intraslab earthquakes. Large events can occur in the Puget Sound region, probably because of cooler intraslab temperatures and possibly because of a stress concentration or zone of weakness along the pronounced arch in the Juan de Fuca plate. A previous study has suggested an intraslab subduction zone origin for a M 7.3 earthquake that occurred in 1873 near the town of Brookings, in southernmost Oregon. However, analysis of its seismotectonic setting and comparison with other historical earthquakes in northernmost California suggest that the event probably had a very shallow origin within the Gorda block (southern csz) and was not a deep intraslab earthquake in the central csz.

Systematics of halogen elements and their radioisotopes in thermal springs of the Cascade Range, Central Oregon, Western USA

This study quantifies the cycling of halogen elements through the Cascadia subduction zone based on the chemistry of thermal springs in the Central Oregon Cascade Range and of a mineral spring in the forearc (Willamette Valley). Considerations based on mass balances, element ratios, and 36Cl/Cl and 129I/I ratios suggest that halogens discharged through the thermal springs in the Cascade Range are probably derived from magma degassing. Our results indicate that < 35% of the subducted Cl and < 20% of the subducted Br and I could be transported through arc volcanism and the thermal springs, a considerably lower percentage than estimated for other volcanic arcs along the Pacific Rim. A likely explanation for this difference is that a large fraction of the halogens is released from the slab at shallow depths into the serpentinized sub-forearc mantle because of the relatively high temperatures in the subducting Juan de Fuca plate. The small fraction of halogens subducted to depth probably also indicates a low rate of water transport, which is consistent with the observation that the Cascade Range sub-arc mantle is relatively dry and has a low degree of volcanic vigor, compared with other arcs.

New constraints on subduction zone structure in northern Cascadia

SUMMARY A detailed passive seismic experiment was carried out across southwestern British Columbia and northwestern Washington to investigate the structure of the subducting Juan de Fuca plate and mantle wedge in Cascadia and its relation to intraslab seismicity. As part of the POLARIS project, 31 three-component broad-band stations were deployed in an approximately linear array spanning southern Vancouver Island, the Gulf and San Juan Islands, Watcom county and the British Columbia lower mainland. P-wave coda from 41 teleseismic events have been employed in formal inversions for fine-scale shear-velocity structure. Our results indicate a structure very similar to that identified across a comparable profile in central Oregon. The continental Moho is evident at the eastern end of the profile near 35-km depth but disappears towards Georgia Strait/Puget Sound. A prominent low-S-velocity zone is clearly evident below southern Vancouver Island dipping eastwards through Georgia Strait/Puget Sound, and coincides with the E-reflection zone originally identified in LITHOPROBE studies. Structure below the E-layer is much less prominent and varies intermittently along the array. Based on the observations and interpretations of similar structures beneath Oregon, Alaska and South America, and its projection to mantle depths, we suggest that the low-velocity E-layer represents the dehydrating oceanic crust of the subducting Juan de Fuca plate. This interpretation is consistent with recent seismicity studies that place shallow Wadati‐Benioff events within the oceanic mantle, and implies that the oceanic crust is 6‐8 km shallower beneath Vancouver Island than previously assumed. As in Oregon, we interpret the diminished signature of oceanic crust below a depth of 45 km to signal the presence of eclogitization, which in turn supplies water to serpentinize the overlying forearc mantle.

Forearc structure beneath southwestern British Columbia: A three‐dimensional tomographic velocity model

This paper presents a three‐dimensional compressional wave velocity model of the forearc crust and upper mantle and the subducting Juan de Fuca plate beneath southwestern British Columbia and the adjoining straits of Georgia and Juan de Fuca. The velocity model was constructed through joint tomographic inversion of 50,000 first‐arrival times from earthquakes and active seismic sources. Wrangellia rocks of the accreted Paleozoic and Mesozoic island arc assemblage underlying southern Vancouver Island in the Cascadia forearc are imaged at some locations with higher than average lower crustal velocities of 6.5–7.2 km/s, similar to observations at other island arc terranes. The mafic Eocene Crescent terrane, thrust landward beneath southern Vancouver Island, exhibits crustal velocities in the range of 6.0–6.7 km/s and is inferred to extend to a depth of more than 20 km. The Cenozoic Olympic Subduction Complex, an accretionary prism thrust beneath the Crescent terrane in the Olympic Peninsula, is imaged as a low‐velocity wedge to depths of at least 20 km. Three zones with velocities of 7.0–7.5 km/s, inferred to be mafic and/or ultramafic units, lie above the subducting Juan de Fuca plate at depths of 25–35 km. The forearc upper mantle wedge beneath southeastern Vancouver Island and the Strait of Georgia exhibits low velocities of 7.2–7.5 km/s, inferred to correspond to ∼20% serpentinization of mantle peridotites, and consistent with similar observations in other warm subduction zones. Estimated dip of the Juan de Fuca plate beneath southern Vancouver Island is ∼11°, 16°, and 27° at depths of 30, 40, and 50 km, respectively.

Extent and duration of the 2003 Cascadia slow earthquake

Inversion of continuous GPS measurements from the Pacific Northwest show the 2003 Cascadia slow earthquake to be among the largest of ten transients recognized here. Twelve stations bracketing slow slip indicate transient slip propagated bi‐directionally from initiation in the southern Puget basin, reaching 300 km along‐strike over a period of seven weeks. This event produced, for the first time, resolvable vertical subsidence, and horizontal displacement reaching six mm in southern Washington State. Inverted for non‐negative thrust slip, a maximum of 3.8 cm of slip is inferred, centered at 28 km depth near the sharp arch in the subducting Juan de Fuca plate. Nearly all slip lies shallower than 38 km. Inverted slip shows a total moment release equal to Mw = 6.6 and a high degree of spatial localization rather than near‐uniform slip. This suggests rupture concentrated along asperities holds for slow earthquakes as well as conventional events.

Fault Parameters of the Nisqually Earthquake Determined from Moment Tensor Solutions and the Surface Deformation from GPS and InSAR

The magnitude 6.8 Nisqually earthquake occurred on 28 February 2001 at a depth of 50-60 km within the subducting Juan de Fuca plate. The fault parameters are estimated from moment tensor solutions and by comparing the surface displacements from Global Positioning System (GPS) data and from satellite interferometry, with predictions from elastic deformation models. The simple deformation model calculates the coseismic surface displacements caused by an earthquake on an extensional rectangular fault in a uniform half-space. Continuous GPS stations within 200 km of the epicenter resolved horizontal displacements as great as 9 mm (210°) and vertical displacements as great as 13 mm of subsidence near the epicenter. The near-vertical displacements were also determined from a differential interferogram created from synthetic aperture radar (InSAR) data from RADARSAT , the only satellite data available for the event. A maximum vertical subsidence of approximately 20 mm is observed 30 km east of the epicenter. The GPS, InSAR, and moment tensor solutions provide a consistent solution for the rupture parameters of the Nisqually earthquake. The fault has a strike of 180° and a dip of 20° and is centered at 47.10° N and 122.67° W (4 km east and 6 km south of the epicenter) or has a strike of 360° and a dip of 70° and is centered at 47.10° N and 122.69° W (2 km east and 6 km south). The trade-off between fault area and rupture displacement is not resolved by our data, but a good fit is found with the main rupture having an area of 230 km2 and an along-strike length of 23 km and downdip width of 10 km. The rupture area is centered at a depth of 51 km with a scalar moment of 2.0 × 1019 N m and, with the given area, an average rupture displacement of 1.4 m. To refine the interpretation of the GPS and InSAR data, a 3D heterogeneous numerical model was generated having realistic shear moduli structure, a 3D model of the subducting slab, and a spherical Earth. The results are similar to those from the half-space model, but there is significant refinement with a deeper rupture center at 60 km. Manuscript received 9 April 2003.

Possible emplacement of crustal rocks into the forearc mantle of the Cascadia Subduction Zone

Seismic reflection profiles shot across the Cascadia forearc show that a 5–15 km thick band of reflections, previously interpreted as a lower crustal shear zone above the subducting Juan de Fuca plate, extends into the upper mantle of the North American plate, reaching depths of at least 50 km. In the extreme western corner of the mantle wedge, these reflectors occur in rocks with P wave velocities of 6750–7000 ms−1. Elsewhere, the forearc mantle, which is probably partially serpentinized, exhibits velocities of approximately 7500 ms−1. The rocks with velocities of 6750–7000 ms−1 are anomalous with respect to the surrounding mantle, and may represent either: (1) locally high mantle serpentinization, (2) oceanic crust trapped by backstepping of the subduction zone, or (3) rocks from the lower continental crust that have been transported into the uppermost mantle by subduction erosion. The association of subparallel seismic reflectors with these anomalously low velocities favours the tectonic emplacement of crustal rocks.

A Comparison of Ground Motions from the 2001 M 6.8 In-Slab Earthquakes in Cascadia and Japan

Are earthquake source and propagation properties for in-slab earthquakes in the Cascadia region (British Columbia and Washington State) similar to those in the subduction zone of southwest Japan? To answer this question, we compare ground motions for two recent earthquakes of similar magnitude and depth. The comparisons are made for horizontal-component 5% damped response spectral acceleration at frequencies from 0.5 to 10 Hz. The selected Cascadia event is the Nisqually earthquake of 28 February 2001, which occurred within the subducting Juan de Fuca plate with a moment magnitude of M 6.8 at a depth of 51 km. The Japan event is the Geiyo earthquake of 24 March 2001 with M 6.8, at a depth of 60 km within the subducting Philippine Sea plate. Generally, the response spectral amplitudes from the two events overlay each other, when amplitudes are plotted against distance. An interesting relationship is observed however. At low frequencies (0.5-1 Hz) the Nisqually data tend to have higher amplitudes, whereas at high frequency (5-10 Hz) the Nisqually data have lower amplitudes, in comparison to the Geiyo data. The observed “drop through” of the Nisqually data relative to the Geiyo data as frequency increases could be due to either source effects or site effects. (It is not a path effect because there is no indication of differences in the slope of the attenuation curve.) If the observed frequency-dependent amplitude differences are due to source effects, this could imply that in-slab earthquakes in Japan are fundamentally different from those that occur in the Cascadia subduction zone. This would have important implications, as it would cast doubt on the applicability of lessons learned in Japan to hazard assessment in Cascadia. On the other hand, if the observed differences are attributable to site effects, then Japanese observations and results are applicable to Cascadia once we have accounted for differences in site effects. To address the origin of the observed differences in ground motion between the Nisqually and Geiyo earthquakes, we evaluate the influence of site effects on the recorded motions. We correct all response spectra to an equivalent rock motion, using amplifications derived from quarter-wavelength site response calculations for representative regional soil profiles. The site-corrected rock motions from Nisqually and Geiyo are in reasonable agreement with each other at all frequencies, with the exception of some Nisqually observations from West Seattle. We conclude that the observed differences between the Nisqually and Geiyo ground motions are attributable to regional differences in site response.

Evidence for both crustal and mantle earthquakes in the subducting Juan de Fuca plate

We investigate the relationship between high‐precision hypocentres in the subducting Juan de Fuca plate beneath southwest British Columbia, and the velocity structure from receiver function analysis and reflection seismic data. 108 earthquakes at depths of 40 to 65 km were relocated using a double‐difference algorithm to obtain precise earthquake locations. Correlation of the relocated seismicity with structural information shows a concentration of earthquakes near the top of the subducting oceanic crust, and a deeper layer of seismicity in the uppermost mantle of the subducting Juan de Fuca plate. The strong correlation of the seismicity with the structure of the subducting plate suggests that seismic failure is caused by changes in mechanical strength within the plate, supporting the hypotheses that phase transformations and thermal‐petrological conditions play an important role in the seismogenesis of double‐seismic zones.

Intraslab earthquakes: dehydration of the Cascadia slab.

We simultaneously invert travel times of refracted and wide-angle reflected waves for three-dimensional compressional-wave velocity structure, earthquake locations, and reflector geometry in northwest Washington state. The reflector, interpreted to be the crust-mantle boundary (Moho) of the subducting Juan de Fuca plate, separates intraslab earthquakes into two groups, permitting a new understanding of the origins of intraslab earthquakes in Cascadia. Earthquakes up-dip of the Moho's 45-kilometer depth contour occur below the reflector, in the subducted oceanic mantle, consistent with serpentinite dehydration; earthquakes located down-dip occur primarily within the subducted crust, consistent with the basalt-to-eclogite transformation.

Multiparameter two‐dimensional inversion of scattered teleseismic body waves 3. Application to the Cascadia 1993 data set

This is the third paper in a three‐part series that examines formal inversion of the teleseismic P wave coda for discontinuous, two‐dimensional (2‐D) variations in elastic properties beneath dense arrays of three‐component, broadband seismometers. In this paper, the method is applied to data from the Incorporated Research Institutions for Seismology‐Program for Array Seismic Studies of the Continental Lithosphere (IRIS‐PASSCAL) Cascadia 1993 experiment undertaken across central Oregon. Two major features are imaged in the resulting model. The continental Moho becomes evident ∼150 km from the coast beneath the Western Cascades and extends through the eastern end of the profile at 35–40 km depth. In the western portion of the model, oceanic crust of the subducting Juan de Fuca plate dips shallowly (12°) at the coast and more steeply (27°) below the Willamette Valley and is evident to depths of &gt;100 km beneath the High Cascades. The abrupt increase in plate dip at ∼40 km depth coincides with an apparent thickening of the oceanic crust followed by a diminution in its signature. Building on previous work, we argue that these results are consistent with the consequences of prograde metamorphic reactions occurring within the oceanic crust. Progressive dehydration at lower‐grade facies conditions culminates in the transformation to eclogite, producing a pronounced increase in the seismic velocity, density and dip of the subducting plate, and structural complexity in the overlying wedge.

A regional study of shear wave splitting above the Cascadia Subduction Zone: Margin‐parallel crustal stress

Recordings of local earthquakes from 16 three‐component broadband seismic stations in southwestern British Columbia, Washington, and northern Oregon are used to study regional variations of shear wave anisotropy in the North American plate above the subducting Juan de Fuca plate. There is evidence for shear wave splitting at all sites, with good agreement of fast polarization directions and travel time delays at adjacent stations. Most stations exhibit fast directions parallel to the strike of the margin, with anisotropy of 1–2%. These fast polarization directions are consistent with earthquake focal mechanisms and borehole stress studies, indicating that the observed anisotropy is likely due to crustal stresses (i.e., extensive dilatancy anisotropy theory). The margin‐parallel stresses may be due to oblique subduction of the Juan de Fuca plate. However, at the station closest to the coast (OZB), the fast direction shows a more margin‐normal orientation that may be associated with the proximity of the locked portion of the underlying subduction thrust fault.

Geochemical and 3He/4He evidence for mantle and crustal contributions to geothermal fluids in the western Canadian continental margin

Isotopic and geochemical evidence together with helium isotopes are used to identify contributions of deep crustal to upper mantle volatile components in thermal waters at three sites across the western North American plate margin: (1) the low heat-flow forearc of the Cascadia subduction zone; (2) the high heat-flow volcanic arc; and (3) the interior crystalline terrain of the ancestral continental margin.Western continental margin hotsprings issue 50°C, low salinity Na–Cl water and N2 gas with 0.25% CH4. Stable isotopes and 14C indicate local meteoric recharge during the early Holocene. Redox is buffered by sulphate reduction, suggesting that the methane originates from a deeper source. The waters have high helium contents (He/Heair=30) and a 3He-excess (R/Rair=0.27; 3He/3Heair=8), representing a mixture of radiogenic 4He production in crystalline rock with >10% He from subducted oceanic crust. Geophysical data indicate fluid-filled discontinuities in the subduction zone that may provide a pathway for He, and possibly for CH4 and a saline fluid component from depth.In the Garibaldi belt of Quaternary arc-volcanism, 60°C Na–Cl hotsprings and 200°C geothermal well waters discharge from fractures in the basement rocks. δ18O and δ2H show the thermal waters to be a mixture of local recharge with up to 8% “andesitic” water from the upper mantle. He isotopes indicate a mantle origin (R/Rair=6.0), with a minor crustal signature, consistent with observations in the Cascadia range to the south and at other circum-Pacific volcanic arc settings. High PCO2, an enriched δ13CDIC, elevated 3He/CO2 ratios and elevated Cl− are likely to be derived from subducted Juan de Fuca plate sediments and pore waters.Thermal Na–SO4 waters (up to 58°C) from the Omineca Crystalline Belt are locally recharged and have an unusually rapid circulation time of less than 40years. This contrasts with their high radiogenic He content (176×10−7cc/g) with minimal mantle input (R/Rair=0.06). The underlying Columbia River Fault Zone is the likely pathway for helium migration from depth.

Deformation across the forearc of the Cascadia subduction zone at Cape Blanco, Oregon

Over the interval 1992–1999 the U.S. Geological Survey measured the deformation of a geodetic array extending N80°E (approximate direction of plate convergence) from Cape Blanco on the Oregon coast to the volcanic arc near Newberry Crater (55 and 350 km, respectively, from the deformation front). Within about 150 km from the deformation front, the forearc is being compressed arcward (N80°E) by coupling to the subducting Juan de Fuca plate. Dislocation modeling of the observed N80°E compression suggests that the main thrust zone (the locked portion of the Juan de Fuca‐forearc interface) is about 40 km wide in the downdip direction. The transverse (N10°W) velocity component of the forearc measured with respect to the fixed interior of North America decreases with distance from the deformation front at a rate of about 0.03 mm yr−1 km−1. That gradient appears to be a consequence of rigid rotation of the forearc block relative to fixed interior North America (Euler vector of 43.4°±0.1°N, 120.0°±0.4°W, and −1.67±0.17 ° (m.y.)−1; quoted uncertainties are standard deviations). The rotation rate is similar to the paleomagnetically measured rotation rate (−1.0±0.2 °(m.y.)−1) of the 15 Ma lava flows along the Columbia River 250 km farther north. The back arc does not appear to participate in this rotation but rather is migrating at a rate of about 3.6 mm yr−1 northward with respect to fixed North America. That migration could be partly an artifact of an imperfect tie of our reference coordinate system to the interior of North America.

Relative sea-level change inferred from intertidal sediments beneath marshes on Vancouver Island, British Columbia

Abstract Relative sea-level change at the time of, and since, the most recent great earthquake at the Cascadia subduction zone is estimated from intertidal sediments at three marshes on western Vancouver Island, British Columbia. We compare the elevation of the pre-earthquake surface, which is marked by a tsunami sand sheet, with the modern depositional elevation range of the sediment type upon which the sand was deposited. At a site south of the Nootka fault zone, which is the northern boundary of the subducting Juan de Fuca plate, tidal mud overlies the pre-earthquake marsh surface. The stratigraphy at this site indicates 0.2–1.6 m of coseismic submergence and 1.1 m of subsequent emergence. In contrast, two sites to the north lack obvious stratigraphic evidence for coseismic land-level change and record between 0.1 and 1.7 m of post-earthquake submergence. These results indicate a difference in tectonic environment across the Nootka fault zone and suggest that plate-boundary rupture during the last great Cascadia earthquake probably did not extend north of central Vancouver Island.

The northern limit of the subducted Juan de Fuca Plate system

Analysis of data recorded at an array of three‐component broadband seismograph stations deployed on northern Vancouver Island and the adjacent British Columbia mainland, at the northern end of the Cascadia subduction zone, provides the first constraints on the S wave velocity structure of this region and permits us to define the northern limit of the subducted Juan de Fuca plate system. During a 2‐year period, more than 80 teleseisms were recorded at our five stations. The method of receiver function analysis was used to constrain the S velocity structure to upper mantle depths. Beneath the northern three stations, a relatively simple continental crust is interpreted with a well‐defined Moho near 37–39 km depth. An upper crustal S velocity discontinuity at these stations is interpreted as the top of the high‐velocity rocks of the Wrangellia terrane. In contrast, more complicated structure dominated by pronounced low‐velocity zones dipping to the NE are interpreted beneath our southern two stations. The shallower low‐velocity zone is 6–8 km thick, has an S velocity contrast of 0.6–1.1 km/s, and lies within the continental crust. This feature is similar to a pronounced low‐velocity layer (the E zone) imaged beneath southern Vancouver Island. The deeper low‐velocity zone is interpreted as the subducted oceanic crust. We interpret the pronounced change in S velocity structure that we observe as the northern limit of the subducted oceanic plate beneath Vancouver Island. This change coincides with significant changes in topography, heat flow, gravity, and geochemistry.

Evidence for large earthquakes at the Cascadia Subduction Zone

Large, historically unprecedented earthquakes at the Cascadia subduction zone in western North America have left signs of sudden land level change, tsunamis, and strong shaking in coastal sediments. The coastal geological evidence suggests that many of the earthquakes occurred at the boundary between the overriding North American plate and the subducting Juan de Fuca plate. This hypothesis is consistent with geodetic measurements and the results of geophysical modeling, which indicate that part of the plate boundary is locked and accumulating elastic strain that will be released during a future large earthquake. Arguments based on potential amounts of seismic slip and likely rupture areas suggest that most or all of the plate boundary earthquakes were magnitude 8 or larger events. The last earthquake or series of earthquakes, about 300 years ago, ruptured the entire 1000‐km length of the subduction zone; if it was a single quake, it probably exceeded magnitude 9. Other earthquakes may have ruptured one or more segments of the subduction zone or may have occurred on faults in the North American plate. Recurrence intervals are uncertain because of difficulties in identifying and dating earthquakes. In southwestern Washington state, intervals for the seven most recent earthquakes average about 500 years but range from less than 200 years to 700–1300 years. Future research on Cascadian plate boundary earthquakes will probably focus on (1) the relation between plate boundary and crustal earthquakes, (2) earthquake magnitude, (3) the areal extent and severity of seismic ground motions, (4) ages and number of past plate boundary earthquakes, and (5) land level changes preceding earthquakes.