T53A-1106
Forearc Structure and Fault Slip Near the Epicenter of the April 1, 2007, Megathrust Earthquake (Mw 8.1) and Tsunami in the Solomon Islands
We reprocessed three multichannel seismic-reflection (MCS) lines collected by the USGS (1982, 1984) near the epicenter of the 2007 Mw-8.1 megathrust earthquake that struck the Solomon Islands. Near the epicenter, several bathymetric and tectonic elements, including an active spreading ridge and a transform fault, are being subducted at the New Britain Trench. These subducted elements affected fault slip during the earthquake, as indicated by two finite fault models (Yagi, 2007; Ji, 2007). Slip began around the epicenter, southeast of where the spreading ridge enters the subduction zone. Slip was reduced directly over the ridge, and northwest of the ridge, slip resumed with increased, possibly maximum amplitude. Fault rupture propagated northwestward at about 1.95 km/s (Yagi, 2007). The Woodlark spreading ridge, with its irregular bathymetry and probable high heat flow, injects a strong three-dimensionality into the analysis of fault slip along the interplate thrust. MCS data show smooth reflections from the interplate decollement that can be followed for about 40 km east of the trench, and hypocenters locally recorded during 1998 (Yoneshima et al., 2005) trace the plate interface deep into the subduction zone. The downgoing plate dips ~30° northeast through the zone of highest 1998 seismic activity, which occurs below 20 km depth. Although young oceanic crust is being subducted eastward along the New Britain Trench, the subducting plate bends sharply downward and dips steeply (30° to 45°) into the mantle. Teleseismic data place the 2007 earthquake epicenter near the trench axis, close to the up-dip limit of the seismogenic zone indicated by the 1998 seismicity. Under the upper slope of the island arc south of the epicenter, strong reflections suggest a mixed volcanic and carbonate-rock framework of the island arc. Down slope of where the strong reflections end, seaward-verging thrust faults deform several small forearc basins. Deformation occurred episodically, as indicated by stacked angular disconformities within the basins. Listric normal faults document local collapse of the upper slope that may have been driven by subduction of high-standing oceanic features. Modeling the tsunami produced by the 2007 earthquake indicates that some islands are so close to the epicenter that tsunami waves arrived within 5 min. after shaking began, allowing people scant time to react.
T53A-1107
Crustal structure around the asperity regions of large earthquakes along the southernmost Kuril trench revealed by two Airgun-OBS seismic profilings
In the southeast off Hokkaido, Japan, large earthquakes have occurred repeatedly with temporal and spatial regularities along the Kuril trench due to the subduction of Pacific plate at a rate of 80 mm/year (DeMets et al., 1990) [e.g. the 1952 Tokachi-oki earthquake (Mw=8.2), the 1973 Nemuro-oki earthquake (Mw=7.8) and the 2003 Tokachi-oki earthquake (Mw=8.2)]. It is considered that the next large earthquake will occur at the source region of the 1973 Nemuro-oki earthquake in the near future because a low seismic activity has been found in the off shore region of the Nemuro peninsula. In order to clarify the relation between asperity and the recurrence of large earthquake, it is necessary to determine a detailed crustal structure running across the two asperities of the 2003 Tokachi-oki earthquake and 1973 Nemuro-oki earthquake. Therefore we conducted a wide-angle survey across the coseismic rupture areas of the 2003 Tokachi-oki and 1973 Nemuro-oki earthquakes and the afterslip area of the 2003 Tokachi-oki earthquake parallel (profile-A) and perpendicular (profile-B) to the Kuril trench using Ocean Bottom Seismometers (OBSs). In profile-A, 19 OBSs were deployed at a spacing of about 10km and three 25 liter air-guns were fired every 90 seconds which corresponds to a shot interval of about 230m. In profile-B, 11 OBSs were deployed at a spacing of about 11km and two 25 liter airguns were fired every 60 seconds which corresponds to a shot interval of about 150m. In this presentation, we report on the crustal structure using the data obtained in the profile A and B and compare our results with past researches around this region. This study is founded by the Ministry of Education, Culture, Sports, Science and Technology, Japan.
T53A-1108
Seismotectonics of accretive versus erosive subduction zones - insights from analog seismic cycle simulation
Accretive and erosive subduction zones differ both in their forearc structure and seismic release character. For instance the greatest historical megathrust earthquakes concentrated along accretive margins, tectonically characterized by forearc shortening (e.g. Sumatra, Southern Chile, Alaska), whereas erosive margins, tectonically characterized by forearc extension (e.g. Peru, Kuriles), have often been the locus of tsunami earthquakes (i.e. slow and shallow events). Here we investigate the implied link between internal forearc deformation and megathrust seismogenesis and its implications for seismic hazard in subduction zones. We interpret quasi two- dimensional plastoelastic (allowing deformation to localize, permanent shortening dominates) and elastoplastic (elastic deformation dominates, minor internal deformation) granular wedge models as analogs of accretive and erosive subduction forearcs, respectively, overlying a rate-state frictional plate interface which represents a seismogenic megathrust. Experimental observations support current hypotheses that internal forearc deformation is controlled by stress changes associated with the megathrust seismic cycle: Consistent with the theory of dynamic Coulomb wedges, coseismic compression at the updip limit of great earthquakes triggers shallow postseismic forearc deformation. Plastic shortening of the outer forearc wedge and shallow afterslip both are interpreted as transient postseismic relaxation mechanisms with the first being dominant in plastoelastic/accretive settings and the second being dominant in elastoplastic/erosive settings. Interseismically, permanent crustal shortening localizes in both settings above the downdip limit of great earthquakes and may lead to uplift of a coastal cordillera. Longterm coastal uplift rates at elastoplastic/erosive margins are about one order of magnitude lower than in plastoelastic/accretive settings, and associated with permanent crustal extension above the seismogenic zone. Extension here appears as an effect of postseismic relaxation coeval to interseismic reloading. Internal shortening of plastoelastic/accretive margins reduces the subduction velocity stretching the recurrence intervals. We found, however, no evidence for that longer recurrence intervals lead to larger earthquake slip (e.g. due to stronger healing effects). Consequently, the seismic moment release rate tends to be smaller at accretive than at erosive margins, at least in the 2D case presented here. Since megathrust earthquakes of magnitude 8 and higher gain their size not primarily by increasing coseismic slip but by unzipping larger parts of the plate interface along strike of the subduction zone, the historical concentration of great events along accretive margins, if statistically significant, is probably a 3D effect associated with lateral rupture propagation. These findings may have important implications for seismic hazard in subduction zones since (1) forearc anatomy is suggested to reflect the seismogenic structure at depth, for instance in accretive settings, (2) erosive settings tend to release elastic energy by shallow megathrust afterslip and thus have a generally higher risk of tsunami earthquakes than accretive margins and (3) erosive margins are characterized by more frequent megathrust earthquakes than accretive margins. Very great earthquakes along accretive margins may be triggered either by a longer wavelength along-strike-segmentation of the subduction interface or by more synchronized seismic cycles of neighboring segments compared to erosive margins thus boosting lateral rupture propagation during a single giant event.
T53A-1109
Lesser Antilles Subduction Zone Investigation by a Cluster of Large Seismic Experiments in the Forearc Region
Thales LAST stands for Lesser Antilles Subduction zone Team which gathers the scientific teams of a cluster of surveys and cruises that have been carried out in 2007 and coordinated under the European Union THALES WAS RIGHT project (Coord. A. Hirn). This cluster is composed by the German cruise TRAIL with the vessel F/S Merian (PI E. Flueh and H. Kopp, IFM-GEOMAR), the French cruise SISMANTILLES 2 with the IFREMER vessel N/O Atalante (PI M. Laigle, IPG Paris and JF. Lebrun, Univ. Antilles Guyane), and French cruise OBSANTILLES with the IRD vessel N/O Antea (PI P. Charvis, Geoazur, Nice, France). During these cruises and surveys, 84 Ocean Bottom 3-components Seismometers (OBS) and 20 Hydrophones (OBHs) have been brought together from several pools (Geoazur, INSU, IPGP, IFM-GEOMAR, AWI,), with up to 30 land stations (CSIC Barcelone, IPG Paris, INSU-RLBM and -LITHOSCOPE) in addition to the permanent onshore arrays of IPGP and SRU. The deployment of all these instruments has been supported principally by ANR Catastrophes Telluriques et Tsunamis (SUBSISMANTI), by the EU SALVADOR Programme of IFM-GEOMAR, as well as by the EU project THALES WAS RIGHT on the Antilles and Hellenic active subductions to which contribute IPGP, Geoazur, IFM-GEOMAR (Germany), ETH Zurich (Switzerland), CSIC Barcelona (Spain), Univ. Trieste (Italy) and NOA Athens (Greece). The main goal of this large seismic investigation effort is the understanding of the behaviour of the seismogenic zone and location of potential source regions of mega-thrust earthquakes. Specific goals are the mapping of the subduction interplate in the range where it may be seismogenic along the Lesser Antilles Arc from Antigua to southern Martinique Islands, as a contribution to identification and localisation in advance of main rupture zones of possible future major earthquakes, and to the search for transient signals of the activity. The forearc region, commonly considered as a proxy to the seismogenic portion of the subduction mega-thrust fault plane, and which is here the main target has been localized along 3 transects to the Arc thanks to a preliminary survey in 2001, the French SISMANTILLES cuise. We will present the first results obtained during these experiments dedicated specifically to image at depth the seismic structure and activity of this region. To image faults at depth and the detailed upper-crustal structure, 3700 km of multi-beam bathymetry and multi-channel reflection seismic profiles have been collected along a grid comprising 7 strike-lines of up to 300 km long and spaced by 15 km and 12 transects of up to 150 km long and spaced by 25 km (SISMANTILLES 2). All these airgun shots dedicated to deep penetration have been recorded by the 84 OBSs and 20 OBHs deployed by the F/S Merian and N/O Atalante on the nodes of this grid of profiles. It will permit to get Vp constraints on the deep forearc region and mantle wedge by wide-angle refraction studies, as well as constraints on the updip and downdip limits of the seismogenic part of the mega-thrust fault plane. Two of these transects have been extended across the whole arc during the TRAIL survey, with up to 50 OBSs deployed along both 240 km long profiles. All these OBSs remained several months beyond the shot experiments for local earthquakes Vp and Vp/Vs tomography. They have been recovered and partly redeployed by N/O Antea during the OBSANTILLES survey. A significant number of those instruments had broadband seismometers, a notable originality in the case of the OBSs to detect low-frequency transient signals.
T53A-1110
Effects of Initial Stress State on Splay Fault Activation During Dynamic Rupture Propagation
Critical tapered wedge concepts suggest that as material is added to the accretionary prism or removed from the forearc, the material overlying the plate interface must deform to maintain a wedge structure. This internal deformation is achieved by slip on splay faults branching from the main detachment. During major thrust earthquakes, splay faults are possibly activated as part of the seismic event. As a rupture propagates updip along the plate interface, it will reach a series of junctions between the shallowly dipping detachment and more steeply dipping splay faults. The amount and distribution of slip on these surfaces will determine the seafloor deformation and the tsunami waveform. Numerical studies by Kame et al. [JGR, 2003] of fault branching during dynamic slip-weakening rupture in 2D plane strain showed that branch activation depends on the initial stress state, rupture velocity at the branching junction, and branch angle. They found that for a constant initial stress state with the maximum principal stress at shallow angles to the main fault, branch activation is favored on the compressional side of the fault for a range of branch angles. By extending the part of their work on modeling the branching behavior in the context of subduction zones, where the angle \Psi that the principal stress makes with the main fault is shallow, we hope to better understand the conditions for splay fault activation and the criteria for significant moment release on the splay. In aid of that, we conduct similar dynamic rupture analyses to those of by Kame et al., but use explicit finite element methods, and take fuller account of overall structure of the zone (rather than focusing just on the branching junction). Recent studies suggest that splay faults may have been activated during the 2004 Sumatra Andaman event [Araki et al., 2005; Sibuet et al., 2007]. Thus, we use a geometry representative of Sumatra [Banerjee et al., 2007] for our modeling. The aim is to understand what stress states and rupture parameters predict seismic splay fault activation, and what not. Based on Kame et al., the orientation \Psi of the most compressive stress at the branch junction is one important parameter. We use the wedge mechanics concepts to guide the choice of initial stress state in the modeling, take the stresses to be depth dependent, and allow for different elastic properties in the wedge and subducting seafloor. We also investigate how the stress state and other model parameters control the partitioning of slip between main and splay faults and what that means for the tsunami waveform.
T53A-1111
Focal mechanism distribution of main- and after- shock of the 2005 off Miyagi Earthquake (M7.2) by using radiation pattern of P- and S- wave
Interplate earthquakes of magnitude of 7.5 have occurred along the subduction plate boundary of the Miyagi-Oki region, middle part of the Japan Trench area, repeatedly on about 40 years intervals. In 2005, interplate earthquake with magnitude of 7.2 occurred in this area, considered to be the rupture of one of the asperities of the 1978 earthquake (M7.4) (Okada et al., 2005: Yaginuma et al., 2007). Hino et al. (2007) and Suzuki et al. (2007) determined the focal mechanism distribution of main- and after- shock of the 2005 earthquake by using ocnean bottom seismometers (OBSs) and coastal stations. They classified the aftershocks into two types according to their focal mechanism solutions: 1) mainshock type having focal mechanism similar to the mainshock and 2) non-mainshock type with focal mechanism dissimilar to the mainshock. Based on this classification, Suzuki et al. (2007) discovered that the non-mainshock type events tend to occur near the edge of the areas with large co- and post- seismic moment release. However, they used only the polarities of the first P wave arrivals for determining focal mechanisms. Therefore the estimation errors are large for earthquakes which are not located very beneath the seismic network due to poor station coverage, and some of the aftershocks could be classified into another groups due to the large estimation errors. In this paper, we improve the accuracies of focal mechanisms by using P- and S- wave radiation pattern to discuss whether the systematic variation of the mechanism of the aftershocks are substantial. Our new result, in which we use the radiation pattern to calculate the focal mechanisms, suggests that focal mechanism patterns show more distinguished relationship with the co- and post-seismic slip area of the mainshock than previous studies: Non-mainshock-type events are distributed near the edge of the large co- and post-seismic slip areas; mainshock-type events occur within post-seismic slip area. We consider that the non- mainshock type events occurred reflecting the stress change due to the fault motion, both the mainshock and the afterslip. The high seismicity of the non-mainshock type events on the peripheries of the areas of the co- and post- seismic slips may reflect the existence of barriers prohibiting rupture propagation and giving rise to stress concentration.
T53A-1112
Imaging Deep Structure of Sumatra Megaquake Fault Plane Using Long Offset Marine Seismic Reflection Data
Seismic reflection imaging in subduction zone environment is hampered by poor penetration of seismic energy in deformed sediments, seafloor scattering and severe water bottom multiples. To image deep structure, we require a very large source rich low frequencies and a very long streamer. We use the data acquired by WesternGeco marine vessel Geco Searcher in July 2006 towing two long streamers (5.5 and 12.5 km) and 10170 cubic inch airgun array source. Different offset data contain information on different structures. Therefore, instead of using conventional stacking procedure, we have developed offset optimisation and horizon consistent velocity analysis techniques where data from different offsets are stacked for different horizons, which provide a higher quality image, particularly of deep structures that are generally hindered by noise. Since deep reflectors are generally broken up due scattering noise from near surface structures and do not have diffraction tails, conventional post-stack migration of data leads to severe smiles at the extremities of deep reflections and scattered noise. We iteratively build a migration velocity based on constant velocity migration that provide the best migrated image of deep structure and then performed a 2D Kirchhoff migration technique leading to a better quality image. Our re-processed image clearly shows the top and bottom of the down going oceanic plate and deep penetrating faults that seem to cut through the oceanic crust and go down in the mantle.
T53A-1113
Crustal features along the southern Kuril Trench, Japan, obtained by a refraction/reflection seismic survey
The seismogenic zone in the southern Kuril Trench can be divided into two segments by the Kushiro Canyon, the Nemuro segment to the east and the Tokachi segment to the west. Except for the giant compound earthquake in 17th century, [e.g. Sawai et al., 2002], M8 class earthquakes have occurred repeatedly within each of these segments. The 1952 and 2003 Tokachi earthquakes are considered to be repeated rupture of the asperity of the Tokachi-oki segment. In order to reveal the seismic velocity structure related to the rupture propagation or suspension along the plate boundary, we made a seismic survey across the segment boundary between the Nemuro and Tokachi segments. In the experiment, we deployed 16 OBSs along a seismic line with about 180 km length and shot 75 liter airgun to correct wide-angle seismic data, and MCS survey was also made simultaneously. The profile ran through the focal areas of the 2003 Tokachi and the 1973 Nemuro earthquakes along the strike of the Kuril Trench. The first arrival times observed by the OBSs are inverted for 2-D P-wave velocity distribution and locations of major reflectors are imaged by using traveltime mapping method (TMM) [Fujie et al., 2005]. In the obtained crustal velocity model, sedimentary layers with Vp < 4.8 km/s shows significant variation along the profile. In the rupture area of the 2003 Tokachi earthquake, their total thickness is about 8 km, it decrease to about 4 km in the segment boundary zone around the Kushiro Canyon. In the Vp model obtained by Nakanishi et al [2004], the layer with Vp of about 5~6 km/s was interpreted as the upper crustal layer of the Kuril arc. But the present result of the TMM shows that there is a distinct reflective boundary within the layer, which separating the layer into upper and lower units. Judging from its large vertical velocity gradient, the upper unit may be old sedimentary unit. Wells et al [2003] pointed out the correlation between the low gravity anomaly (LGA) zones and areas of large coseismic slip. Based on this relation, they discussed that sedimentary basins are developed above locked portions of the plate boundaries due to basal erosion, including the Tokachi segment. Our structure model demonstrates that a thick sedimentary pond is actually developed in the LGA corresponding to the asperity of the Tokachi segment.
T53A-1114
Observations of Shallow Thrust Seismicity in the Northern Mariana Subduction Zone From Broadband Ocean-bottom and Land Seismometers
The Mariana arc is considered the type `aseismic' or `decoupled' subduction zone, since the concept was first proposed by Uyeda and Kanamori [1979]. The historical seismic record going back to about 110 years shows that the largest events in the Northern and Central Mariana were Ms 7.4 and 7.1 events in 1902 and 1934 respectively, but the locations and depths are not well constrained and it is thus unclear whether they represent shallow thrust faulting events. Recent large (Mw > 7.0) earthquakes are confined to the southernmost arc. Thus the largest confirmed shallow thrust faulting event along a 1000 km section of the Central and Northern Mariana is Mw 6.4, a remarkable seismic deficit considering the convergence rate of 4 cm/yr. We study the seismicity of the Mariana forearc using records from the 2003-2004 Mariana Subduction Factory Imaging Experiment to explore the mode of seismic release on the shallow thrust zone. The deployment of 20 broadband land stations and 58 semi-broadband ocean-bottom seismometers (OBS) ran for 11 months. Locations obtained using the genloc algorithm within the Antelope software package and further refined using a local velocity model from Takahashi, et al [2007] indicate sparse seismicity within regions immediately beneath the Big Blue and Celestial serpentinite seamounts, but do show dense clusters of seismicity at depths of 25-50 km immediately to the west of these seamounts. Focal mechanisms of the largest events occurring in these clusters were studied using a regional waveform inversion method and suggest the events occur along the shallow thrust zone. A preliminary comparison of microearthquake seismicity rates to the CMT record shows a disparity between occurrence of teleseismic events and smaller events recorded by our deployment, suggesting a small b-value along the shallower part of the thrust zone and a larger b-value along the deeper part. Observations of low microseismicity at shallow depths (0-25 km) and prolific microseismicity at deeper depths have been interpreted as evidence of a locked shallow subduction interface for several other subduction zones such as Costa Rica, where it is consistent with geodetic data. However, it is not clear whether a similar interpretation is feasible in the Mariana Islands considering the lack of historical seismicity.
T53A-1115
Preliminary Analysis of an Anomalous Bathymetric "Notch" Between the Kumano Forearc Basin and the Slope of the Nankai Trough Accretionary Prism
A 12 km wide, 56 km long, three-dimensional (3-D) seismic volume acquired over the Nankai Trough offshore the SE Kii Peninsula, Japan images the Nankai accretionary prism, forearc basin and the subducting Philippine Sea plate. Preliminary analysis reveals an unusual, trench-parallel ~1200 m deep depression (a "notch") along the seaward edge of the Kumano forearc basin and the upper slope of the accretionary prism, just landward of the shallowest branches of the previously-mapped splay-fault system. The shape of this notch varies along strike, from a single, steep-walled, ~3.5 km wide notch in the northeast, to a broader, ~6 km wide zone with several shallower linear bathymetric lows in the southwest. We have interpreted the shallow notch area in detail and find a large degree of deformation and a number of faults with varying dips and lateral extents. Faults below the central axis (bottom) of the notch are almost vertical, though often sinuous or splayed and therefore may have a translational component. Off-axis faults often dip toward the central axis but most of these faults have no indication of the sense of slip. The Bottom Simulating Reflector (BSR) is strong throughout most of the study area but has distinct gaps beneath steep seafloor slope breaks, such as the landward (northern) edge of the notch. Interpretation below the uppermost 2000 m is hampered by poor imaging due to the complex topography, strong BSR and associated gas deposits, and structural complexities from extensive deformation and fracturing. The structures associated with the notch may become clearer with additional seismic data processing. Several possible causes for this feature include an axial strike-slip fault complex, a thrust or normal fault system or a current-incised channel. The presence of faults along bathymetric anomalies indicates some degree of structural control, but an interaction between structure and currents cannot yet be ruled out.
T53A-1116
Quantifying Geomorphology Associated With Large Subduction Zone Ruptures
For 30 rupture zones, we quantify forearc basin size and subducting seafloor roughness through a simple and traditional semivariance technique. The subduction of geomorphologic features, such as ridges and seamounts, increases basal erosion and subsidence of the accretionary wedge. Forearc basins form through such subsidence. Many subduction zone ruptures have been associated with these basins, where a great portion of the ruptures" asperities collocate with the basins. Mechanisms responsible for this spatial correlation include the initiation of stick-slip sliding beneath the accretionary wedge due to the subduction of bathymetric highs (e.g. seamounts), along-strike variations in crustal thickness and density of the overriding plate, and structural segmentation of the upper plate. Comparing the maximum semivariance (sill), the horizontal distance associated with the sill (range), and the fractal dimension of each rupture zone to the same parameters of the mirrored area on the subducting seafloor, we see a relationship between forearc and seafloor geomorphology, where rougher seafloor correlates to rougher forearc. Seafloor range, seafloor fractal dimension, and rupture zone fractal dimension all increase with earthquake moment magnitude. Additionally, we observe an increase in tsunami magnitude with range, suggesting some geomorphologic mechanism for tsunami earthquake generation.
T53A-1117
Asperities of the 1703 Genroku and 1923 Kanto earthquakes and the purpose of the Kanto Asperity Project
Earthquakes along the Sagami trough, where the Philippine Sea slab is subducting, have repeatedly occurred. The 1703 Genroku and 1923 (Taisho) Kanto earthquakes (M 8.2 and M 7.9, respectively) are known as typical ones, and cause severe damages in the metropolitan area. The recurrence periods of Genroku- and Taisho-type earthquakes inferred from studies of wave cut terraces are about 200-400 and 2000 years, respectively(e.g., Earthquake Research Committee, 2004). We have inferred the source process of the 1923 Kanto earthquake from geodetic, teleseismic, and strong motion data (Kobayashi and Koketsu, 2005). Two asperities of the 1923 Kanto earthquake are located around the western part of Kanagawa prefecture (the base of the Izu peninsula) and around the Miura peninsula. The strong motion data provided us with much more information on the source time function than the teleseismic data (Kobayashi and Koketsu, 2005). We adopted an updated fault plane model, which is based on a recent model of the Philippine Sea slab, and the asperity around the Miura peninsula moves to the north (Sato et al., 2005). In this study, we have investigated the slip distribution of the 1703 Genroku earthquake. Since no seismic waveform data were recorded, we cannot infer the time-dependent rupture process. We used crustal uplift and subsidence data investigated by Shishikura (2003), and inferred the slip distribution with the same geometry of the fault as the 1923 Kanto earthquake. An additional asperity is located the southern part of the Boso Peninsula and the maximum slip is over 16 m. In the case of subduction-zone earthquakes, the asperity is considered not to be changed. In an interseismic period, coupling between plates is strong in an asperity region, and is weak in a non-asperity region (e.g., Coordination Committee for Earthquake Prediction Research in Universities, 2006). This framework can be applied to the subduction zone along the Sagami trough. The difference between the Sagami and Nankai troughs, where the Philippine Sea slab is also subducting, is that slow slip events occur at the same depths as the asperity. Many studies on the asperity along the Sagami trough have been carried out. Seismicity during these five years is very low in these asperities (Kobayashi and Koketsu, 2005). This suggests strong seismic coupling, which is also inferred from the geodetic study of the slip deficit distribution (Sagiya, 2005) Around the asperities, small repeating earthquakes are observed as well as slow slip events. Kimura (2005) shows that the small repeating earthquakes occurred during the slow slip events. We have proposed seismic monitoring in the Sagami Bay and off Boso region as a part of the Kanto Asperity Project, which is an IODP drilling plan, to characterize asperity and non-asperity regions. Nine borehole sites and inland ones can cover the asperity and non-asperity regions. Good coverage and very low noise in boreholes can catch smaller earthquakes, improve hypocenter and focal mechanism determinations. If seismic surveys with OBS are performed, we can obtain shallow structures of the Philippine Sea slab and knowledge on the relation between amplitude of reflection and asperity. In future, we can get a chance of deep drilling reaching the plate boundary in the asperity region, which cannot be included in this project, if we successfully show good results.
T53A-1118
Uplift and deformation of the Kumano forearc basin: Preliminary results from Nankai Trough 3D seismic imaging offshore Kii Peninsula, Japan
Forearc basins contain excellent records of tectonic events that influence the subduction zones they overlie. The Kumano forearc basin, offshore the Kii Penninsula, Japan, overlies the Nankai Trough subduction zone, accretionary prism, and mega-splay fault believed to have slipped during the 1944 Mw8.0 seismic event. In 2006, an academic 3D seismic survey acquired by Petroleum GeoServices (PGS) and processed through prestack time migration by Compagnie Générale de Géophysique (CGG) resulted in unparalleled images of the outer Kumano forearc basin. We map key unconformities and sequences within this basin to look for evidence of significant events in the geologic history of the basin that might yield clues as to the processes related to the formation or evolution of the mega-splay system. Sequence 1 noncomformably overlies seismically-opaque, accretionary prism lithologies, averages less than 0.5 s twtt in thickness, and is generally flay-lying. These strata may have been deposited in a slope basin environment and then uplifted to form the base of the forearc basin. The flat-lying geometry is in stark contrast to all strata above this sequence, which have a pronounced landward tilt. Sequences 2-6 total over 1 s twtt in thickness and unconformably overlie Sequence 1. The geometry of the deepest of these landward tiled sequences, Sequence 2, seems to suggest a seaward sediment source. Sequences 3-6 exhibit diminishing landward tilt, increasingly landward-shifted sediment depocenters, and onlap relationships. The unconformity that separates Sequences 2 and 3 may mark a significant tectonic event. The Sequence 2 sediments below this unconformity have undergone wholesale tilting and form a pronounced bathymetric high running parallel and proximal to the current outer-arc high. Sediments above this surface show infilling of the sediment trap created by the uplift of the seaward 15 km of the forearc basin. Numerous clearly imaged normal faults persist in both the older, highly titled strata of Sequence 2 and to a lesser extent in the younger sequences. Several different processes could account for the uplift recorded in the forearc basin including accretion of large volumes of sediment to the prism changing the prism taper, subduction of a seamount or ridge, and/or changes in the fault (mega-splay) architecture. The waning tilt in the uppermost sediments suggests that there is a transient nature to whatever process caused the uplift.
T53A-1119
Coseismic vertical deformation during the great 2007 Solomon Islands megathrust rupture
A joint US-Solomon Islands team visited the epicentral area of the 1 April 2007 Mw 8.1 Solomon Islands earthquake a few weeks after the event. We used coral microatolls, satellite imagery, and displaced geomorphic and cultural features to map the coseismic deformation pattern in the region directly above and adjacent to the subduction megathrust rupture. Among our main findings is that most slip occurred on the shallow portion of the megathrust and that slip appears to have reached the deformation front at Ranongga island, which was uplifted as much as ~2.5 m when the rupture propagated across the subducting Simbo ridge transform. Simbo island, which sits on the downgoing Australian plate and lies only 8 km across the plate boundary from uplifted Ranongga, subsided ~0.7 m coseismically and experienced only subdued ground motions. The line of zero vertical displacement (hingeline) runs closely along the southwestern coasts of Vella Lavella, Ghizo, and Parara islands, implying a persistent structural relationship between the downdip limit of coseismic slip and these coastlines. A broad, asymmetrical subsidence trough as deep as ~0.7 m extends across Vella Lavella, Kolombangara, Parara, and New Georgia. Uplift of ~0.35 m on the westermost tip of Rendova, along with overall subsidence of Rendova and Tetepare, place a firm limit on the southeastern extent of rupture. Uplift of Mono and subsidence of Fauro and the Shortlands, and no resolvable vertical change on Bougainville, define a rupture length of nearly 250 km between Rendova and the Woodlark rise.
T53A-1120
Spatial Variability in Apparent Stress for Subduction Zone Earthquakes Along the Nicoya Peninsula, Costa Rica
Subduction zones around the world release much of the world's seismic energy in mega- thrust earthquakes. This energy release varies in subduction zones around the world, likely related to heterogeneity in the strength of the mega-thrust zone. Here we examine the heterogeneity along the megathrust at the Middle America Trench, focusing on small magnitude earthquakes along the northwestern margin of Costa Rica, Central America. This area is interesting because seismicity patterns vary along strike of the subduction zone and there is heterogeneous topography on the incoming Cocos plate that could affect the strength of the interface. We compute source spectra and thus seismic moment from coda amplitude measurements for over 2200 well-located earthquakes from the Nicoya Peninsula portion of the 1999-2001 CR-SEIZE project. This method of using the coda wave amplitudes provides more stability for the magnitude estimates because we minimize effects of energy distortion from crustal heterogeneity. These values are then used to compute apparent stress for each earthquake. Preliminary results suggest that earthquakes in the northern and central portions of the peninsula have higher apparent stress than those in the southern portion of the peninsula. These differences seem to coincide with the changes in interface dip as well as the subduction of the Fisher seamount group at the southern tip of Nicoya. These apparent stress results suggest that subduction zone interface in the northern and central portions of the peninsula is either stronger or experiences higher friction along its surface than in the southern region.
T53A-1121
Segmented Forearc Deformation Along the Nicoya Peninsula Seismic Gap, Costa Rica
The Nicoya Peninsula, Costa Rica deforms in response to rapid NE subduction of the Cocos plate at the Middle America Trench (9-10 cm/yr). This emergent outer fore arc peninsula lies 60-80 km inboard of the trench and coincides with a locked segment of the seismogenic zone. The Nicoya segment is a high-potential seismic gap, with a slip deficit of >5 m since the last major earthquake (M7.7, 1950). That event produced widespread damage and up to 1.0 m of coseismic coastal uplift. Net Quaternary deformation on the Nicoya Peninsula is recorded by emergent marine terraces at the coast, and by uplifted alluvial fill within interior valleys. Recent field mapping, surveying, and isotopic dating provide new constraints on deformation patterns and upper-plate faulting. Local uplift anomalies reveal upper plate faults that may accommodate a significant fraction of fore arc deformation (shortening and/or lateral sliver transport). At the peninsula's southern tip (Cabo Blanco), a prominent uplifted marine erosion surface (Cobano surface) encompasses at least three distinct Pleistocene terraces at 30-220 m elevation. Preliminary OSL dating yields terrace ages consistent with OIS 3-5 sea level high stands (30-120 ka), indicating net uplift at 1.0-2.0 m/k.y. A NW- striking thrust fault (Delicias fault) offsets the upper terrace by 40 m, thrusting Cretaceous basalt over Plio- Pleistocene marine sediments. Radiocarbon ages for adjacent Holocene terraces (Cabuya surface) indicate recent uplift at 1.5-3.5 m/k.y. On the peninsula's south-central coastline (Puerto Carrillo to Playa Camaronal) marine terraces and related fluvial straths (Carrillo-Camaronal surface) occur at 20-40 m elevation. Correlations with dated Cobano terraces and Quaternary sea level curves suggest terrace formation between 80-215 ka (OIS 5-7) and net uplift rates of 0.2-0.3 m/k.y. Along the northern Nicoya coast (Tamarindo to Nosara), a 3 km wide wave-cut surface (Iguanazul surface) includes three treads with paleo-shorelines at 10-45 m elevation. Age correlations (as above) suggest terrace formation between 80-215 ka (OIS 5-7) and net uplift rates of 0.1-0.2 m/k.y. Radiocarbon ages for Holocene beachrock horizons are consistent with recent uplift at <0.5 m/k.y. Within the interior Morote Valley, geomorphic anomalies indicate deformation along the NW-striking Morote fault (e.g., uplifted and incised alluvial fill, irregular drainage networks, stream piracy, and abrupt mountain facets). Uplifted Pleistocene alluvium (La Mansion surface) is incised 5-40 m by stream networks. At one site, fluvial gravels overlie a paleosol on fine-grain wetland deposits, 10 m above local base level. In some areas, the paleo- valley gradient is opposite that of modern incised streams, indicating capture and drainage reversal. The Morote fault forms a prominent NW-trending lineament oriented sub-parallel to the plate margin. A composite focal mechanism for shallow seismicity along this trend (CRSEIZE) shows dextral slip for a NW-striking nodal plane, consistent with NW fore arc sliver escape. The observed differences in Quaternary deformation along the Nicoya Peninsula coincide with three contrasting domains of subducting seafloor offshore (EPR, CNS-1, CNS-2). Upper-plate segmentation may reflect along- strike variations in subducting plate roughness, coupling, and seismogenic zone geometry (e.g., dip angle, depth of up-dip and down-dip limits). This segmentation may have implications for understanding the rupture behavior and earthquake hazards along the Nicoya seismic gap.
T53A-1122
Configuration of the Philippine Sea plate in the Kanto district, Japan, estimated from SP and PS converted waves
In the Kanto district, the Philippine Sea plate (PHS) is subducting northwestward beneath the inland plate from the Sagami trough and the Pacific plate (PAC) is subducting westward beneath the PHS from the Japan trench. Because of the complexity of the geological structure, various models for the configuration of the PHS have been proposed (e.g. Ishida, 1992; Kimura et al. 2006; Hirose et al., 2007). In this study, we used SP and PS waves converted at the upper boundary of the PHS to estimate its depth variation. The advantage of the depth estimation utilizing converted waves is that the boundary can be estimated without seismicity near the boundary. On the other hand, small repeating earthquakes are also useful to identify plate boundaries since they are thought to be distributed on the boundaries (e.g. Kimura et al., 2006). We identified small repeating earthquakes in the Kanto district and selected earthquakes that occurred between the PHS and PAC (i.e., bottom of the PHS) based on the waveform similarity and their focal mechanisms. In the seismograms of the earthquakes, we found prominent phases between P and S waves, that are identified as SP and PS waves converted at the upper boundary of the PHS. We measured the SP-P (N=794) and S-PS (N=212) times for the earthquakes and estimated the upper boundary. We expressed the depth of the upper boundary of the PHS as a function of latitude and longitude in the form of power series and the factors of power series were determined by the inversion of the SP-P and S-PS time data, following Horiuchi et al. (1982). The estimated plate boundary was in good agreement with the location of small repeating earthquakes on the upper boundary of the PHS, and thus we also included the depths to the repeaters (N=48) as constraints in the inversion. The depth to the upper boundary of the PHS shows a very small inclination in the depth range between 30-40km beneath northern Chiba prefecture, where no small repeating earthquakes on the top surface of the PHS have been found. We infer the shape is due to the existence of the PAC under the PHS. From the comparison with tomographic study (Hirose et al., 2007), we found the estimated boundary was, in most cases, located near the upper boundary of low Vs, high Vp/Vs layer of the PHS which is interpreted as the oceanic crust of the PHS. Acknowledgements: The waveform data were provided by National Research Institute for Earth Science and Disaster Prevention (Hi-Net), University of Tokyo, Japan Meteorological Agency and Tohoku University.
T53A-1123
Geologic evidence for great earthquakes during the last 1500 years along the eastern Nankai Trough, central Japan
Coastal marsh sequence reveals that great plate boundary earthquakes have caused the coastal area to subside along the eastern Nankai Trough. We analyzed eight drilling cores obtained from a coastal marsh neighboring the plate boundary. The marsh is located on the Philippine Sea plate that subducts beneath Southwest Japan Arc on Eurasian plate in northwestward direction. The plate boundary extends in northeast direction from the Nankai Trough through Suruga Trough to the Fujikawa-kako fault zone in central Japan. The marsh stratigraphy consists of alternation of dark-colored peat and light-colored mud layers, each ranges several tens of cm to 100 cm in thickness. Six couplets of peat and mud layers were recognized from the stratigraphic record formed in the last 1500 years. According to the micro-biological and sedimentological analyses, rapid water-level rise caused the facies change from peat layers to mud layers. Four kinds of criteria strongly suggest that the facies changes reflect the regional coastal subsidence due to the plate boundary earthquakes: 1) wide lateral extent of the facies boundary, 1) the suddenness of facies change, 3) Coincidence of tsunami deposits with the facies boundary, 4) Correspondence to historical earthquakes. Six couplets of peat and mud laterally extend over two km in the marsh. On one hand, the contact of peat layer and overriding mud layer is sharp and often shows scoured surface. On the other hand, the mud layers gradually change into peat layers. These facies changes indicate the rapid water-level rise demolishing the marsh vegetation and gradual recovery of marsh condition. Cross- or parallel laminated sand beds often found at the base of mud layers suggest water-level rise beginning with intrusion of strong sediment flows. Sand beds consisting of multiple fining-upward units, in some cases, suggest the deposition from successive water pulse such as tsunamis or river floods. On the basis of 14C ages and tephrochronology, four peat-mud contacts were correlated to historical earthquakes occurred along the Nankai Trough in AD 684, 1096, 1361 and 1707, respectively. Other two peat-mud contacts, formed in 6th and 8th Century, do not have historical counterparts. They probably indicate the occurrences of earthquakes along the Fujikawa-kako fault zone.
T53A-1124
Seismic Structure Reconnaissance of the Lesser Antilles Subduction Zone
The Lesser Antilles is an active subduction zone which has experienced in 1843 a M > 7.5 probably mega- thrust earthquake that destroyed Pointe-a-Pitre city on Guadeloupe island. But the location in depth of the mega- thrust subduction fault and its geometry along the arc were unknown until recently.A first reconnaissance of the seismic structure and activity has been undertaken by Sismantilles project which has been carried out at a regional scale from Northern Guadeloupe to Martinique islands. The project focused more particularly on the detection, mapping and characterization of the potentially seismogenic part of the interplate subduction fault.For this purpose, wide angle reflection and refraction (WARR) data have been collected by a broad but sparse array of 37 3-components Ocean Bottom Seismometers (OBS) that were deployed offshore Guadeloupe and Martinique islands over severals weeks and recorded continuously the MCS shots from the French N/O Nadir vessel as well as regional and local seismicity.Joint modeling of MCS / wide-angle data on three transects across the arc enables us to identify the contact zone between the forearc crust and the subducting oceanic crust, which is considered as a proxy for the seismogenic part of the interplate. The most striking feature is that its location with respect to the deformation front and the volcanic arc and its downdip size appear significantly variable along the arc.Relocation of recorded earthquakes has been improved by a better network geometry including 3- components OBS just above the sources and thanks to the application of stations corrections for OBS's observations. The recorded earthquakes are mostly located downdip of the contact between the two crusts. Depending on the real geometry of the interplate, which is not yet constrained deeper than 20 km depth, seismicity occured either in the forearc mantle, either at the interplate or in the subducting plate. In the upper plate, a part of microseismicity has been clearly located in the arc crust and mantle.This first study reveals an unexpected strong variation along the arc of the forearc structure and allows to target for finer measurements in the contact zone between the two crusts.
T53A-1125
Widespread presence of serpentinite within the Cascadia megathrust suggested by teleseismic receiver function analysis.
Receiver function analysis of data from long-running observatories located along the West coast of the US in Oregon and Washington is used to characterize the crustal and upper mantle structure of the Cascadian forearc. Our primary goal is to explore along-strike variation in the structure of the crust and the uppermost mantle. Long periods of data collection (5 and more years) allow us to construct well-populated backazimuth and epicentral distance gathers of receiver functions throughout the region. In the area between the coast and the volcanic arc we are able to identify coherent SV-polarized phases consistent with P-S conversion from the top of the subducting Juan-de-Fuca plate. We note that in all cases these phases display significant directional variation, sometimes even switching polarity. We also note that in all cases there is corresponding SH-polarized energy associated with these phases. P-SH scattering has significant amplitude, but does not exhibit polarity changes in the slab dip direction, which means that seismic anisotropy exerts a strong influence on P-S conversion at the top of the Cascadia slab. Beneath Corvallis, Oregon a strongly anisotropic layer on top of the Juan-de-Fuca plate has been related to the presence of serpentinite within the zone of plate contact. Our new results suggest that this feature is wide-spread in Cascadia, although the specific receiver function signature varies considerably. Presence of serpentinized material between tectonic plates of subduction zones has been proposed as an inhibitor to earthquake rupture. We will compare the area of serpentinized plate contact to other constraints on the depth of the seismogenic zone in Cascadia.
T53A-1126
Deformation of the Converging Juan de Fuca Plate Offshore Central Oregon
The Global Positioning System combined with acoustic ranging measured convergence of the Juan de Fuca and the North America plates 150 km offshore central Oregon to be 15.9 mm/yr at N73.6 E from 2000 to 2003. This is 50 % slower and rotated 25 degress more easterly than predicted by geomagnetic anomalies. The oceanic plate may deform here due to a combination of regional plate forces acting within the subduction zone between the obliquely converging, weaker, downgoing slab and the stronger, overidding Siletz Block in the forearc. The seafloor geodesy data and proposed mechanisms will be presented and examimed in the context of regional observations of on-land horizontal and vertical geodesy, forearc structure and faulting, seimsity, and other data.
T53A-1127
First Results from NanTroSEIZE LWD, IODP Expedition 314: Physical and hydrologic properties
The IODP Expedition 314 will be carried out in September through November, 2007 off Kii Peninsula of Japan. It is the first leg of Nankai Trough Seismogenic Zone Experiments (NanTroSEIZE), a multi-year, multi-leg and multi- platform project to investigate fault mechanics and seismogenesis along a subduction megathrust through direct sampling, in situ measurements, and long-term monitoring. During Expedition 314, we plan to drill into 6 NanTroSEIZE sites down to ~1000m below sea floor (bsf). Using the LWD (Logging-While-Drilling) technique, we obtain continous vertical profiles of physical properties such as bulk density, resistivity, porosity, P-wave velocity, and borehole images at all six sites. These sites include the incoming plate trench sediments, the frontal thrust and toe of the accretionary prism, forearc basin deposits, a major out-of-sequence thrust system (the gmegasplayh fault), and highly-deformed rocks of the interior of the accretionary prism. We will show preliminary results of LWD physical properties obtained during Exp.314. One of the anticipated goals is to quantitatively describe the evolution of accretionary prism off Kumano, through porosity reduction with increasing depth and in the landward direction. Abrupt decrease in porosity or other properties is expected across the mega-splay, and its offset is estimated. The activity of mega-splay at shallow depth is inferred from this offset, as well as through analysis of slope failure deposits and young sediments in the forearc basin. Hydrological properties, such as permeability and storage coefficients are estimated as well as elastic properties, especially around fault intervals. They are essential to understand inter- and co-seismic behavior of the faults. Two of six sites will drill into the shallower part of two deep riser holes, targeting the mega-splay and mega-thrust faults. Results from these sites will make an important milestone toward understanding the state and interseismic behavior of the M8 seismogenic zone.
T53A-1128
First Results From NanTroSEIZE Logging While Drilling, IODP Expedition 314: Fault Zone and Thrust Sheet Structural Synthesis
Integrated Ocean Drilling Program (IODP) Expedition 314 is both the maiden scientific voyage of the new drilling vessel Chikyu and the first leg in the Nankai Trough Seismogenic Zone Experiment (NanTroSEIZE), a multi-year effort designed to investigate fault mechanics and seismogenesis along a subduction megathrust through direct sampling, in situ measurements, and long-term monitoring. During Expedition 314, which takes place September through November, 2007, our primary goals are to obtain a comprehensive suite of geophysical logs and other downhole measurements at six sites along a transect focused on the up-dip transition from seismic to aseismic fault behavior, using state-of-the-art logging-while-drilling (LWD) techniques. The sites include sampling to depths of 600 to more than 1000 m of a major out-of-sequence thrust system (the "megasplay" fault), thought to be connected to the seismogenic portion of the plate boundary, as well as incoming plate trench sediments, the frontal thrust of the accretionary prism, forearc basin deposits, and highly-deformed rocks of the interior of the accretionary prism. The principal goals of the LWD program are to document in situ physical properties, lithostratigraphic and structural features, sonic to seismic scale velocity data for core-log-seismic integration, and parameters related to stress, pore pressure, and hydrological properties. A synthesis of results bearing on the structural geology, lithology, and in situ stress conditions of the sediments and faults of the Nankai Trough will be presented. Resistivity image logs will elucidate fault zone structure as well as possible borehole breakouts, and multi-parameter log interpretation will document lithology and permit integration with 3D seismic reflection imaging. http://www.iodp.org/nantroseize
T53A-1129
3D Interpretation of Active Thrust Systems, Nankai Accretionary Wedge: Relationship Between In-Sequence and Out-of Sequence Thrusts
At the Nankai Trough subduction zone large tsunamigenic earthquakes occur with a recurrence interval of 100 to 200 years. One key parameter governing the severity and characteristics of tsunamis is the amount and location of slip near the sea floor at the up-dip end of the megathrust fault system. As a result, the structure in the uppermost 10 km has received close attention, including a recently acquired 3D seismic survey. A major apparent out of sequence thrust (OOST), termed the mega-splay thrust fault, has been identified underlying the Kumano Forearc Basin and approaching the surface at the outer arc high. This mega-splay system may accommodate a large amount of plate convergence in the form of tsunamigenic slip in great earthquakes (>8). Working with new industry-quality 3D seismic reflection data, we focused on the structural differences between the landward out of sequence mega-splay thrust system and the seaward in-sequence thrusts of the outer prism, to unravel their evolution as the prism has developed. In the upper most few kilometers, the OOST branches into 3 to 4 major splays and numerous minor ones with complex three-dimensional geometry. In places, these shallow splays are high-amplitude reflectors and juxtapose thrust sheets of contrasting internal reflectivity character. Numerous apparently in-sequence faults of the outer Accretionary wedge are also well imaged. Detailed 3D seismic interpretations sheds light on the transition from outer wedge faulting to OOST's of the mega-splay. Interpretation and images of the complex geometry of the faults across this transition will be presented. Determining the slip history of the out of sequence thrusts is of particular interest to understand how tsunamigenic slip has occurred during earthquakes, and how the complex fault patterns in the prism have evolved over time.
T53A-1130
Rupture of the Northern San Andreas Fault and Possible Stress Linkages to Cascadia
We relate the late Holocene Northern San Andreas Fault (NSAF) Offshore/onshore paleoseismic history along the northern California continental margin to a similar dataset from the Cascadia margin. Evidence from stratigraphic correlation and merging of turbidity currents at channel confluences supports synchronous triggering of turbidity currents during the Holocene, when other sources such as storm river flows are less unlikely to reach the abyssal plain. In order to make comparisons between the temporal records from the NSAF and Cascadia, we refine correlations of southern Cascadia great earthquakes using 44 piston/trigger pairs, 7 box cores collected in 1999 and two Kasten cores from 2002, combined with the land paleoseismic record. Stratigraphic correlation is accomplished with P-wave velocity, gamma-ray density, RGB color reflectance, magnetic susceptibility, high-resolution imagery and AMS 14C ages. The late Holocene turbidite record off Cascadia and northern California passes several tests of synchronous triggering. Many turbidites can be correlated stratigraphically between channel sites supported by AMS 14C ages. Paleoseismic work at onshore sites along the Cascadia and NSAF systems shows good correspondence with the offshore record, further circumstantial evidence that the offshore record is primarily earthquake generated. During the last ~2800 years, 15 turbidites including the great 1906 earthquake establish an average repeat time of ~200 years, similar to the onshore value of ~210 years. The combined land and marine paleoseismic record from the southern Cascadia subduction zone, developed using similar methods includes a similar number of events in the past 3000 years. While the recurrence interval for full margin Cascadia events is ~530 years, the southern Cascadia margin has a repeat time of 260-290 years, similar to that of the NSAF. We observe that 11 of the previous 15 NSAF events were preceded by Cascadia events by ~0-80 years, averaging 47 years suggesting a temporal link (as compared to ~150 years if Cascadia follows the NSAF). We model the coseismic and cumulative postseismic deformation from great Cascadia megathrust events and compute related stress changes along the NSAF in order to test the possibility that Cascadia earthquakes triggered the penultimate, and perhaps other NSAF events. The Coulomb failure stress (CFS) resulting from the viscous deformation over ~ 60 years does not contribute significantly to the total CFS on the NSAF. However, the coseismic deformation increases CFS on the NSAF by a maximum of about 9 bars, in the section of the fault offshore of Point Delgada, most likely enough to trigger that fault to fail in north-to-south propagating ruptures. Triggering of Cascadia by the NSAF is also possible, though not favored by the paleoseismic data. http://www.activetectonics.coas.oregonstate.edu/