T41C-1209 0800h
Rapid Kinematic and Tectonic Variations Along the 1400-km-long Australia-Woodlark Plate Boundary Zone, Papua New Guinea and Woodlark Basin
Previous GPS studies have shown the wide variability in present-day plate motions across the highly arcuate, 1400-km-long Australia-Woodlark plate boundary extending from Papua New Guinea to the Solomon Islands. GPS-determined motions range from orthogonal oceanic spreading in the Woodlark basin, to continental transtension in the 2.5-km-high core complex area of easternmost Papua New Guinea, to continental strike-slip and transpression in 4-km-high mountains of the Papuan Peninsula. We use imagery, earthquake focal mechanisms, coral reef uplift data, and structural mapping studies to establish the along-strike continuity of the active plate boundary fault. Systematic angular changes in the direction of the plate vector along this continuous fault explain its varied tectonic geomorphology, Holocene uplift history, and geologic structure. We use a series of plate reconstructions to illustrate the longer term, Cenozoic evolution of this boundary including: its formation as an arcuate, N- and NE-dipping ophiolitic suture zone during Paleogene time, the progressive "unzippering" of this thrust over the past 6 Ma along a N- and NE-dipping, low-angle normal fault in easternmost Papua New Guinea, and its "zippering" or continued shortening on the suture thrust in the Owen Stanley Ranges of the Papuan Peninsula. Over the 1400-km-length of the fault, the length of segments of oceanic spreading, transtension, and transpression is 250-500 km; the time period separating one tectonic style from the succeeding style encroaching from the east is several million years. This systematic spatial and temporal superposition of tectonic styles, leads to complex - but predictable - along-strike variations in geologic history.
T41C-1210 0800h
Microplate rotation leads to rapid exhumation of a subduction complex in eastern Papua New Guinea
The Solomon Sea microplate (SSM) formed in response to oblique Australian (AUS)-Pacific convergence. Since the latest Miocene counterclockwise rotation of the SSM has led to rifting, metamorphic core complex (mcc) formation, eclogite exhumation at plate tectonic rates, and seafloor spreading along the SSM-AUS boundary. Aeromagnetic data provide additional constraints on the nature of this plate boundary. The Papuan Peninsula north of the Owen Stanley Fault (OSF) is characterized by low frequency magnetic lows associated with outcrops of mafic and ultramafic rocks of the Papuan Ultramafic belt (PUB). The magnetic signature of the OSF is a sharp contact separating the low frequency magnetic lows of the PUB from the high frequency magnetic highs associated with greenschist grade AUS metasediments and basalts. A similar aeromagnetic pattern is found on the Fergusson and Goodenough Island mccs. Low frequency magnetic lows occur on the north sides of the islands where mafic and serpentinized ultramafic rocks occur in the upper plates. The lower plates consist of gneisses, migmatites and eclogites, bounded by mylonitic shear zones, and intruded by magnetic granodiorites and volcanic rocks; these are characterized in general by high frequency magnetic highs. Reconstructions based on magnetic anomalies in the western Woodlark Basin agree with GPS crustal motion studies and predict sinistral motion along the OSF on the Papuan peninsula and extension in the D'Entrecasteaux Islands from the Pliocene to the Present. We interpret the southward decrease in metamorphic grade of AUS metasediments and basalts (i.e., eclogite to prehnite-pumpellyite facies) as regional isograds and infer that rifting has reactivated structures within a primary plate boundary. Major extensional shear zones exhumed previously subducted AUS crust from beneath mafic and ultramafic rocks of the PUB. This was accompanied by ascent of migmatitic crust, intrusion of granodiorite plutons and rift-related volcanism that variably reset mineral assemblages and thermochronometers as the Woodlark Basin spreading center rift tip propagated westward.
T41C-1211 0800h
Thermochronologic and Geochemical Investigations of High Pressure Metamorphism and Pliocene Exhumation of the D'Entrecasteaux Islands, Southeastern Papua New Guinea
Evidence for high pressure (HP) metamorphism in eastern PNG exists as variably retrogressed eclogites and blueschists in the lower plates of the D'Entrecasteaux Island (DI) metamorphic core complexes (MCCs). Eclogites from Fergusson and Goodenough Islands record P-T conditions up to 24 kbar, 930$\deg$C and 21 kbar, 890$\deg$C, respectively. In situ U-Pb analyses on zircons from these variably retrogressed eclogites yielded ages from $\sim$7.9 to 4.3 Ma on Fergusson Island, and from 2.9 to 2.1 Ma on Goodenough Island. Zircon analyses from each sample yielded what is interpreted to be a single age population containing no inherited zircons. In-situ trace and rare earth element (REE) analyses on adjacent zircon and garnet from one Fergusson Island sample suggest that zircons grew at 4.3 $\pm$ 0.4 Ma at depths of $\sim$75 km, documenting zircon growth under HP metamorphic conditions and exhumation at plate tectonic rates (cm/yr). Additional trace and REE analyses will determine if 7.9-2.1 Ma zircons grew prior to, during, or subsequent to eclogite facies metamorphism and will further constrain P-T-t-D paths followed by these eclogites. The Prevost Range of eastern Normanby Island, the easternmost of the DI, is a Pliocene metamorphic core complex exposing greenschist and blueschist facies metamorphic rocks from shallower structural levels than Goodenough and Fergusson Islands. Rocks exhumed from the deepest structural levels, located in the NE Prevost Range, are retrogressed blueschists which record pressures from 6 to 6.5 kbar. $^{40}$Ar/$^{39}$Ar step heat experiments on white mica from the overprinting retrograde fabric on a blueschist sample yield a weighted mean age of 3.03 $\pm$ 0.08 Ma, giving a minimum age for blueschist facies metamorphism and documenting the timing of white mica growth and/or cooling below closure temperature during MCC exhumation. Step heat analyses on white mica from a calc-silicate in the NW Prevost Range yield a complex spectrum with apparent ages from 7 to 4 Ma, reflecting either partial thermal resetting of white mica or multiple white mica growth stages during metamorphism. $^{40}$Ar/$^{39}$Ar analyses on plagioclase within greenschists from shallower structural levels give complex age spectra with apparent ages from $\sim$100 to 2.4 Ma, suggesting that protolith ages were preserved or only partially reset during metamorphism preceding Pliocene exhumation.
T41C-1212 0800h
Metamorphic Tectonites and Differential Exhumation Reveal 3D Nature of Extension and Lower Crustal Flow in the Active Woodlark Rift, Papua New Guinea
The D'Entrecasteaux Islands metamorphic core complexes (MCCs) occur in the Woodlark rift, a continental region where $\sim$200 km of extension since $\sim$6 Ma has been focused into a relatively small number of normal faults, some dipping at $<$30$\deg$. Flow of a low viscosity lower crust on a time scale of $<$4 m.y. is evidenced by the narrow width of the rift zone, and the apparent large magnitude of extension (Beta $\sim$5) , as contrasted by observed relief on the Moho ($<=$10 km). Gneisses in the footwalls of MCCs, including retrogressed eclogites, have been exhumed from depths of $\sim$30 km since $\sim$3 Ma. Our structural and thermochronometric data covers parts of the D'Entrecasteaux MCCs, including Normanby Island. These are mantled by serpentinized ultramafic rocks of the Papuan ophiolite. Remnants of this upper plate are preserved along the margins of the MCCs. In underlying mylonites, exhumation-related fabrics, microstructures and quartz lattice preferred orientations reveal a regional pattern of lineations and top-north shear inconsistent with diapir tectonics. The gently dipping mylonites were later warped by uplift of the $\sim$30 km-wide domes, perhaps in response to magmatic underplating during intrusion of dolerite dikes and granodiorites at $\sim$2 Ma. Southward across Fergusson Island, muscovite Ar/Ar ages, increase by $\sim$2 m.y. along the detachment, suggesting footwall exhumation by top-to-the-north slip at $>$12 mm/yr. To the east, an MCC on eastern Normanby Island has top-north footwall mylonites that dip gently SW and that were exhumed during the Pliocene as part of a northward progression of normal faulting that did not arrive at the offshore Moresby Seamount until $\sim$1.2 Ma (ODP Leg 180 site), relationships that suggest a rolling-hinge style uplift. Importantly, its detachment exposes no rocks deeper than blueschist-facies. Ductile deformation fabrics in the MCCs reveal patterns of lower crustal motion that can be evaluated against seafloor spreading-derived plate motions. Shear fabrics in MCCs closest to the Woodlark spreading ridges, including Normanby and Misima Islands are parallel to the NNE direction of 0.5-3.6 Ma Solomon Sea-Australia spreading. Farther west, lineations in the lower plates of the D'Entrecasteaux MCCs locally deflect $\sim$40-50$\deg$ clockwise from this direction. This obliquity is interpreted to reflect inhomogeneous lower crustal extension to the west of the Woodlark spreading ridges. A rift corridor extending $\sim$100 km to the north of Goodenough and Fergusson Islands is defined by active normal faulting and subsidence of the Trobriand margin. We infer that a previously subducted, locally eclogite-bearing, slab of thinned Australian lower crust to the north of the islands is being pulled out from beneath this zone as it is being sinistrally sheared along its eastern edge. Today the rift zone steps $\sim$70 km south towards the Papuan Peninsula to define a right-step, an asymmetry that is enhanced by seafloor spreading east of $\sim$151.4$\deg$ E. Published focal mechanisms suggest that N-S sinistral shear along the northern Woodlark rift is continuing. In central Normanby Island, $\sim$2 Ma andesites may have erupted along a transverse fault bounding the deeply exhumed D'Entrecasteaux corridor. Its offshore extension trends N along a major embayment in the Woodlark Rise.
T41C-1213 0800h
High Ductile Shear Strains and Strongly Localised Brittle Slip Associated With a Low-Angle Detachment Fault on Misima Island, Woodlark Rift, Papua New Guinea
The Woodlark rift system forms the boundary between the Solomon Sea Plate and Australian Plate. Misima Island is a metamorphic core complex (MCC), on the Pocklington Rise, $\sim$75 km to the south of the Woodlark spreading centre and $\sim$100 km east of the western tip of sea-floor spreading. Structural data and microstructural analysis reveal that Misima Island is different from other MCCs, including the D'Entrecasteaux Islands MCCs that lie west of the propagating ridge tip. Geological fieldwork on Misima Island has confirmed the presence of a low angle detachment fault that dips $\sim$25$\deg$ to the NE. Amphibolite-facies felsic and mafic gneisses, occur in the lower plate, whereas unmetamorphosed Quaternary-Tertiary volcanic and sedimentary rocks, and low-grade phyllites occur in the upper plate. A flight of 6 coral terraces on the southern coast of the island have been uplifted to a maximum elevation of $\sim$300 m on the lower plate. The eastern side of the island has been uplifted to a lesser extent resulting in $\sim$40 m high coral cliffs at the coast. The juxtaposition of amphibolite-facies rocks in the lower plate against phyllites and unmetamorphosed conglomerates in the upper plate implies that the magnitude of oblique-slip on the fault is on the order of tens of kilometres. Exhumation related ductile deformation has resulted in NNE trending stretching lineations that occur within a $\sim$2 km thick shear zone immediately beneath the detachment. The shear strain gradient increases upwards towards the detachment fault, where there are abundant $<$1 m thick shear zones. The measured NNE slip direction (slickenlines, lineations) is parallel to the Australia-Solomon Sea plate vector between 500 ka and 3.5 Ma, calculated from seafloor spreading data (Taylor et al., 1999). Pre-existing E-W trending stretching lineations show a pattern of progressive rotation towards the detachment fault. We suggest these lineations have been passively rotated by ductile shear and have numerically modelled this process as a function of oblique-slip simple shear parallel to the fault. The results of this model fit the observed lineation pattern implying that ductile deformation in the lower plate did occur chiefly by simple shear. The model suggests that shear strains as high as 10 to 20 are obtained immediately below the detachment in this shear zone and that the zone accommodates a ductile displacement of $>$20 km. Tectonites from the lower plate show little evidence for overprinting of amphibolite-facies deformation fabrics by greenschist-facies mylonitic textures as is commonly seen in other MCCs, including the nearby D'Entrecasteaux Islands MCCs. This implies strongly localised slip on the brittle detachment fault during exhumation of these rocks through the middle parts of the crust. Factors that might contribute to such a situation include; 1) A high geothermal gradient to markedly reduce the envelope of greenschist-facies metamorphic conditions at higher structural levels; 2) A possible weakening effect of high strain-rates on fault gouge material; 3) Deep embrittlement of the crust through prolonged maintenance of high pore-fluid pressures or high strain rates. Preliminary $^{40}$Ar/$^{39}$Ar results, from lower-plate tectonites on Misima Island, yield mica cooling ages of $\sim$8 Ma (Baldwin et al., 2000). Thermochronologic analysis (in progress) will help to determine the relationship between the processes of MCC formation on Misima Island and the sea-floor spreading history of the Woodlark rift.
T41C-1214 0800h
Association between collision, microplate rotation, and back-arc rifting at obliquely convergent margins in the western Pacific: New insights from geodetic data
We present results from kinematic modelling of geodetic data from two obliquely convergent margins in the western Pacific (Papua New Guinea and the North Island, New Zealand). In each of these plate boundary zones, microplates rotate rapidly (3-8 degrees/Myr) relative to the bounding Australian and Pacific plates. These rapid microplate rotations cause dramatic along-strike changes in plate boundary deformation type and rate, and are associated with back-arc rifting in the Manus Basin and the Taupo Volcanic Zone. The rotation of the eastern North Island of New Zealand (NZ) also accommodates much of the margin-parallel component of relative motion that must be taken up in the plate boundary zone, and therefore contributes to the process of slip partitioning. The South Bismarck microplate in Papua New Guinea (PNG), and the eastern North Island of NZ each have angular velocity vectors describing their motion (relative to the subducting plate) that intersect the surface of the Earth where buoyant features (e.g., a continental fragment, oceanic plateau or seamount chain) are colliding with the New Britain Trench (PNG) and the Hikurangi Trough (NZ). We suggest that it is the change from collision to subduction along the obliquely convergent plate margins that causes the rapid microplate rotation and subsequent back-arc rifting in NZ and PNG. At other Western Pacific convergent margins where buoyant indentors subduct and back-arc rifting occurs (Marianas, Vanuatu, and Tonga), our modeling results are very similar to the PNG and NZ cases, in that microplates rotate rapidly about the point where a buoyant indentor intersects the margin. The coincidence of microplate rotations and back-arc rifting with collision of buoyant features at subduction margins suggests there may be a causal relationship between these phenomena. A possible explanation is that the change from collision-related forces to subduction-related forces along the plate boundary exerts a torque on the forearc microplate. This torque causes the plate boundary zone microplates to rotate rapidly about a point (relative to the subducting plate) where the collisional resistance forces are highest.
T41C-1215 INVITED 0800h
Spatial and Temporal Variations Along the New Zealand Plate Boundary: Decoupling, Delamination, and Localization
New Zealand sits astride the Pacific-Australia plate boundary and hosts two fundamental transitions in plate interactions. Subduction of the Pacific plate beneath the North Island along the Hikurangi margin ends and the plate motion is taken up along the Alpine Fault translational plate boundary in the Marlborough-Kaikoura regions of the South Island. This subduction-translation transition has migrated southward since 20-25 Ma. In the south, a small sliver of the Australia plate subducts beneath the Fiordland region, accommodating an $\sim$ 100 km transpressional left-step in the Australia-Pacific plate boundary south of the South Island. At the northern termination of this small subduction zone the Alpine Fault initiates. This subduction-translation-subduction plate boundary structure has undergone significant evolution over the past 15-20 million years, driven by changes in relative plate motion and significant lithospheric deformation focused in the transition zones along the plate boundary. At the northern transition, we argue that the encroaching subducted Pacific slab acts as a chisel on the lower lithosphere of the overriding Australian plate driving the active delamination of much of the mantle lithosphere. This mass removal makes the necessary space to accommodate the slab. Additionally it drives substantial vertical tectonics of the Australia plate producing rapid and localized uplift in the zone of the active delamination. Also a series of ephemeral sedimentary basins have developed and subsequently been exhumed in the wake of the advancing slab edge. At the southern transition, we argue that the localized subduction of Australia beneath Fiordland is enabled by the progressive tearing of a sliver from the Australian plate. This leaves a newly formed edge to the Australian plate that translates northward along the plate margin becoming the western side of the Alpine Fault plate boundary. It is useful to distinguish between the well-described near-surface Alpine Fault (AF) and the less well understood deeper plate boundary shear zone, which we term the Southern Alps plate boundary (SAPB). As a result of the southward migration of the Hikurangi subduction, the SAPB has been shortening in time. Concurrent with the shortening of the SAPB (since $\sim$15 Ma) plate motions between the Pacific and Australia plate have changed, driving a clockwise rotation in the azimuth of motion along the plate boundary through New Zealand. This rotation produces a mismatch between the sense of shear in the ductile lower crust/upper mantle of the plate boundary and the orientation and location of the upper crustal AF. Localization of deformation along the SAPB shear zone can lead to a significant decoupling between the crust and mantle lithosphere. Evidence from upper mantle shear-wave anisotropy (SKS splitting) and deformational modeling suggest that such a decoupling has occurred, and the resulting spatial and temporal variability in crust-mantle coupling across South Island, New Zealand may lead to variability in deformational style along the Southern Alps orogen.
T41C-1216 0800h
Birth of a Plate Boundary: Detailed thermochronological constraint of Cenozoic Plate Boundary Evolution in the South Island of New Zealand
Detailed thermochronological investigations of individual samples collected west of the Alpine Fault zone in New Zealand provide new insight into the spatial distribution of early Australian-Pacific (AUS-PAC) plate boundary evolution that is not preserved elsewhere in the modern orogenic system of the South Island. Internally consistent cooling histories, derived through combination of $^{40}$Ar-$^{39}$Ar, fission track and (U-Th)/He analyses of a range of minerals from individual samples, span the Cenozoic. These records provide direct constraint on the timing and character of three distinct episodes in the evolution of this tectonic boundary. Of particular note, the early Miocene propagation of the coherent AUS-PAC boundary through the South Island is expressed as short-lived, episodes of geographically localized cooling, which we infer to be due to exhumation, beginning at 23-25 Ma. This is consistent with development of this boundary as an initially distributed zone of deformation, progressively localizing onto discrete fault structures developing primacy of strain accommodation over time. Further work is in progress to map out the spatial distribution of this activity in greater detail, and thereby constrain the progressive development of the plate boundary structure. Rapid cooling universally observed from 8 Ma onwards corroborates the association of the modern obliquely convergent character of the AUS-PAC boundary through the South Island with a marked change in Pacific Plate motion at that time. Reconciling this regionally synchronous onset with previous geographically wider-spread studies of the Southern Alps supports recent suggestions that along-strike variation in age reported for individual thermochronometers results largely from structural inhomogeneity within the orogen, rather than diachronous tectonic development of the system.
T41C-1217 0800h
Influence of Cumulative Boundary-Normal Convergence on Topography and Thrust Fault Development in Oceanic Lithosphere: The Australian-Pacific Plate Boundary South of New Zealand Since 10.9 Ma
Individual segments of the submarine Australian-Pacific plate boundary south of New Zealand have evolved with unique tectonic histories since ~10.9 Ma (anomaly 5o). This is the direct result of the variable orientation of the boundary and relatively close proximity of the Australian-Pacific poles of rotation. Interaction along the oceanic extent of the plate boundary involves oceanic lithosphere of broadly similar age (i.e. thermal structure), resulting in morphologic and structural differences that can be attributed primarily to the variability in the angle and rate of convergence with respect to the plate boundary orientation over the last 10.9 Ma. We use swath bathymetry data from three cruises to quantify the amount of deformation that has occurred along ~1500 km of the plate boundary south of New Zealand. We calculate volumes of crust within 75 km of the plate boundary to the east and west that are displaced from the average local seafloor depth, and compare those results to plate boundary-normal convergence along the boundary predicted by stage rotations since 5.9 Ma. Plate boundary-normal convergence and topography are compared considering active faulting observed at the seafloor to characterize the geodynamic evolution of the different regions. Analysis reveals that a boundary-normal convergence of ~100 km marks the transition from strike-slip dominated faulting to partitioned underthrusting and strike-slip faulting. We consider the topography and structural development at the Macquarie and McDougall segments of the MRC to reflect failure to initiate subduction, suggesting that topographic relief (deepest trench to ridge crest) at a convergent oceanic boundary can reach ~5 km before initiation, which could take 5.9 m.y. These estimates provide observational constraints for the deformation processes involved in the poorly understood topic of subduction initiation, and should be incorporated in numerical modeling of such processes in similar settings.
T41C-1218 0800h
New constraints on the Pleistocene slip rate of the Hope Fault, South Island, New Zealand.
The Marlborough Fault Zone (MFZ) is a series of 4 major strike-slip faults which lie along the Australian-Pacific plate boundary in the northern part of the South Island, New Zealand. The MFZ accommodates almost all of the relative plate motion in this region and the most southerly structure, the Hope Fault, is considered to be the most active. Despite its regional significance, the Pleistocene slip rate of the Hope Fault remains poorly constrained and current geological estimates range between 18 and 35 mm/yr for its central segment. More robust slip rate measurements would therefore be important for seismic hazard analysis. At Sawyers Creek a flight of river terraces have been clearly offset by the Hope Fault. These have been used to calculate some of the existing published slip rates, however the complexity of the offset geomorphic features and the dating methods used limit the accuracy of these results. This study revisits the Sawyers Creek area, employing contemporary surveying and dating techniques to resolve these issues. Kinematic GPS height measurements are combined with interpretations of aerial photographs and field observations to produce a 2 m resolution digital elevation model of the offset features, and the detailed nature of this allows a reinterpretation of the offset distances. Optically stimulated luminescence dating of river silts from the terrace tread provides more accurate age constraints than the weathering rind dating technique used previously. The calculated slip rate of 18-20 mm/yr lies at the lower end of the range of published estimates, in line with the relative plate motion and geological slip rate estimates on the adjacent faults.
T41C-1219 0800h
Tear in the subducting slab beneath the southern Mariana Arc: evidence from P-wave tomography
Seismic tomography has developed into one of the most effective and significant sources of information in modern geophysical research, specifically in understanding subduction zone dynamics. The Western Pacific convergent margin is a well studied area, but as more data is collected and technology improves further detailed observations and theories are being developed. New tomographic images of the Mariana Arc region have enhanced resolution of gradients and strong variations in wave speeds through the use a three-dimensional ray-tracing inversion algorithm. Although the images obtained from the Mariana arc show relatively low amplitudes of heterogeneity due to the limited number of seismic stations in the area, details of the geometry and morphology of the Pacific Plate subducting beneath the Philippine Sea Plate are visualized in three dimensions with unprecedented detail. The slab has an arcuate shape in depth, similar to the surface expression of the trench axis, and plunges steeply into the lower mantle in the northern and central portion. A markedly different geometry exists south of 14°N where the slab continues to have a steep dip, but is only 200 km in length. Between the two distinct morphologies is a seismic velocity anomaly that we interpret as a tear. This near vertical tear in the slab strikes E-W, is 75-200 km in depth, and has an eastern extent of 145.5°E and continues westward to approximately 143.5°E. We propose three models to explain the formation of the tear: rapid rollback of the Challenger Deep slab segment, the junction of the Pacific and Caroline plates, or the subduction of the Caroline Islands Ridge.
T41C-1220 0800h
Bathymetry of Mariana Trench-Arc System and Formation of the Challenger Deep as a Consequence of Weak Plate Coupling
The Challenger Deep in the southernmost Mariana Trench (western Pacific Ocean) is the deepest point on the earth's surface (10,920 m below sea level). Its location within a subduction trench, where one plate bends and descends below another, is not surprising. However, why is it located in the southernmost Mariana Trench and not at its central part, where the rate of subduction is higher, where the lithosphere is the oldest (and densest) on Earth, and where the subducted lithosphere pulling down is the longest in Earth (~1000 km or more according to seismic tomography)? We suggest that although subduction rate and slab age generally control trench depth, here, the width of the Plate Coupling Zone is more important and counteracts trench deepening. Beneath the central Marianas the subducted slab is attached to the upper plate along a 150-km-wide surface that holds the shallow portion of the subducted plate nearly horizontal, in spite of its great load and, thus, counters trench deepening. In contrast, along the south Mariana Trench the subducted length of the lithosphere is much shorter, but its attachment to the upper plate is only along a relatively narrow, 50-km-wide, surface. In addition, a tear in the slab beneath this region helps it to sink rapidly through the mantle and this combination of circumstances allows the slab to roll back, steepen, and form the deepest trench on Earth. In a wider perspective, the interrelations shown here between trench deepening, ridge shallowing, slab steepening, and forearc narrowing may shed light on other subduction zones located near edges of rapidly retreating slabs.
T41C-1221 0800h
Along Strike GPS and Seismic Constraints on Extension in the Eastern Central Range, Taiwan
Synorogenic extension is well documented in many active mountain belts around the world and has gained recent attention in the formation of the Taiwan arc-continent collision (e.g. Crespi et al., 1996; Yu et al., 1997; Kao and Chen, 2000; Wang and Hung, 2002; Bos et al., 2003). Available GPS, seismic and geologic data sets and new GPS data from the Hengchun Peninsula, Taiwan illuminate a pattern of extension east of the Taiwan mountain crest. Kinematic data extracted from the BATS data set suggest a change in the extensional mechanism from gravitational collapse in the north to lateral extrusion in the south. We surveyed GPS monuments in southern Taiwan between January 2000 and November 2002 that can be divided into two data sets: five sites west of the mountain crest and three sites east of the crest. West of the crest, the five monuments have an average velocity of 51 mm/yr westward, whereas east of the crest the average velocity of three monuments is 40 mm/yr. This decrease in velocity from west to east indicates a zone of extension just east the mountain crest. The eastern monuments also have higher horizontal north component velocities than the western monuments. Extension in the Hengchun Peninsula correlates with similar patterns of extension along strike and together define a zone of extension east of the crest that extends nearly the entire length of the orogenic belt. The geometry and kinematics of faults in the eastern Central Range extensional zone, however, show a change from north to south. In northern Taiwan (in the vicinity 23.5\deg N) the geometries of major fault structures are interpreted from patterns of seismicity prior to and after the 1999 Chi-Chi earthquake, using, in part, the collapsing method of Bossu (2000). These data suggest two north-striking and near vertical fault zones: a post-Chi-Chi, west-side down normal fault zone just east of the mountain crest and extending from about 5 to 15km and a pre-Chi-Chi, west-side up reverse fault east of the normal fault and near the Longitudinal Valley that extends from 15 to 30 km. Focal mechanisms for the reverse faults (from the BATS database) show P-axes plunging gently NW and T-axes plunging near-vertical. Focal mechanisms for the extensional faults show P- and T-axes that are inverted relative to the solutions for the reverse faults. This inversion to extension following the Chi-Chi earthquake is interpreted to represent a period of west-side-down gravitational collapse along the fault east of the topographic crest. To the south, in the vicinity of the Southern Cross Island Highway, the fault geometries are less organized and pre- and post-Chi-Chi patterns have not been recognized. The available kinematic data are also only partially consistent with the data to the north. Although extensional earthquakes dominate the region east of the crest the associated T-axes plunge gently SW. This change in the T-axes orientations during extensional events from steeply plunging in the north to gently southwest plunging in south is interpreted to reflect the lateral flow of material as the oblique collision propagates towards the South China Sea. In this southern zone of extrusion extension appears to be accommodated by a variety of structures, including oblique strike-slip and normal faults.
T41C-1222 0800h
Tectonic Wedging Along the Rear of the Offshore Taiwan Accretionary Prism
The structural geometry, kinematics and density structure along the rear of the offshore Taiwan accretionary prism were studied using seismic reflection profiling and gravity modeling. Deformation between the offshore prism and forearc basin at the point of incipient collision, and southward into the region of subduction, has been interpreted as a tectonic wedge, similar to those observed along the front of mountain ranges. This tectonic wedge is bounded by an east-dipping roof thrust and a blind, west-dipping floor thrust. An east-dipping sequence of forearc-basin strata in the hanging wall of the roof thrust reaches a thickness in excess of 4 km near the tip of the interpreted tectonic wedge. Section restoration of the roof sequence yields an estimate of 4 km of shortening, which is small compared with that inferred in the collision area to the north, based on the variation in distance between the apex of the prism and the island arc. Previous studies propose that either high-angle normal faulting or backfolding has exhumed the metamorphic rocks along the eastern flank of the Central Range in the collision zone on land. To better constrain the initial crustal configuration, we tested 350 crustal models to fit the free-air gravity anomaly data in the offshore region to study the density structure along the rear of the accretionary prism in the subduction and initial collision zones before the structures become more complex in the collision zone on land. The gravity anomaly, observed in the region of subduction (20.2oN), can be modeled with the arc basement forming a trenchward-dipping backstop that is overlain by materials with densities in the range of sedimentary rocks. Near the point of incipient collision (20.9oN), however, the free-air gravity anomaly over the rear of the prism is approximately 40 mgals higher, compared with the region of subduction, and requires a significant component of high density crustal rocks within the tectonic wedge. These results suggest that the forearc basement may be deformed along the rear of the prism, associated with the onset of collision, but not in the subduction region further to the south.
http://www.gps.caltech.edu/~wchi
T41C-1223 0800h
Deformation Pattern of an Accretionary Wedge in the Transition Zone From Subduction to Collision Offshore Southwestern Taiwan
Swath bathymetry data and seismic reflection profiles have been used to investigate details of the deformation pattern in the area offshore southwestern Taiwan where the Luzon subduction complex first encroaches on the passive Chinese continental margin. Distinctive fold-and-thrust structures of the convergent zone and horst-and-graben structures of the passive margin are separated by a deformation front that extends NNW-ward from the eastern edge of the Manila Trench to the foot of the continental slope. This deformation front gradually turns into a NNE-SSW trending direction across the continental slope and the Kaoping Shelf, and connects to the frontal thrusts of the mountain belt on land Taiwan. However, the complex Penghu submarine canyon system blurs the exact location of the deformation front and nature of many morphotectonic features offshore SW Taiwan. Based on the analyses of seismic reflection profiles and interpretation of various morphologic features revealed from high-resolution bathymetry maps, it appears that the deformation front offshore SW Taiwan does not appear as a simple structural line, but is characterized by a series of N-S trending folds and thrusts that terminate sequentially in an en-echelon pattern along the passive Chinese continental slope. A number of NE-SW trending lineaments cut across the fold-and-thrust structures of the accretionary wedge and exhibit prominent dextral displacement indicative of the lateral expulsion of SW Taiwan. One of the prominent lineaments, named the Yung-An lineament, forms the southeastern boundary of the upper part of the Penghu submarine canyon, and has conspicuous influence over the drainage pattern of the canyon.
T41C-1224 0800h
Changes in Plate Boundary Conditions Along the NW Pacific Rim as Recorded in the Tertiary Mineoka Ophiolite (Central Japan)
The Tertiary Mineoka ophiolite, which occurs in a fault zone at the intersection of the Honshu and Izu forearcs in central Japan, displays structural and geochemical evidence for its multi-stage evolution in various tectonic settings that attests to rapid changes in plate boundary conditions along the NW Pacific Rim throughout the late Cenozoic. The Mineoka ophiolite first developed as a mid-ocean ridge crust (49-45 Ma) on the Mineoka plate (counterpart of the New Guinea plate). A change in the regional stress regime around 43 Ma resulted in the initiation of a west-dipping subduction zone, which produced island arc tholeiitic magmatism and subsequently backarc basin rift magmatism superimposed on the earlier-formed _eMineoka_f oceanic crust. Eruption of Ocean Island Basalt (OIB)-like alkali-basalts on the Mineoka plate occurred near the paleo-Japan continental arc, around 20-19 Ma, after the cessation of backarc rifting. These alkali basalts and the _eMineoka_f oceanic crust were covered with Paleogene to Miocene pelagic sediments on the way to the oblique convergent margin of Eurasia (paleo-Japan arc). Much of the _eMineoka_f ocean floor was subducted beneath the paleo-Japan continental arc (Eurasian plate) by 14 Ma. With southward migration of the unstable TTT-triple junction between the Pacific and Mineoka plates, the northern tip of the Izu volcanic arc collided with the Honshu Arc (continental arc) and fragments of the _eMineoka_f oceanic crust were incorporated as an ophiolite into the upper plate of the subduction-accretion system of this continental arc. After the consumption of the Mineoka plate and the emplacement of the Mineoka ophiolite, the Boso triple junction was established with the E-W-trending transpressional Sagami Trough adjoining the Izu-Bonin and Japan Trenches. Around 2 Ma, a change in the relative motion of the Philippine-Eurasian plates resulted in the development of a dominantly dextral strike-slip stress regime within the Sagami Trough, part of which has occupied today_fs Boso Peninsula and the Mineoka fault zone. This stress regime initiated a broad zone of dextral strike-slip deformation between the Boso triple junction and the recently established Fuji triple junction further to the west that is still active today.
T41C-1225 0800h
Knocker Problem Revisited: Plutonic and Metamorphic Blocks in Mineoka Ophiolite of Oblique Subduction Boundary, Japan
In Boso Peninsula on Sagami trough, Japan, ophiolitic rocks called the Mineoka Ophiolite form an east-west fault zone along a forearc sliver fault parallel to the present oblique subduction boundary. The fault zone is composed of serpentinite, picrite, alkali basalts (WPB), tholeiitic basalts (MORB, BABB or IAT), chert, and limestone, most are of Eocene age, and are associated with Miocene volcaniclastic or serpentine-bearing sedimentary rocks. Plutonic rocks such as diorite and gabbro (island arc chemistry), and metamorphic rocks (mafic with siliceous and psammitic) such as hornblende schist and amphibolite are included. We observed some critical contacts between these rocks to know the processes and mechanisms of emplacement, and further discussed the tectonic implication for the knocker problem in the forearc ophiolite fault zone. These knockers are a few m to 100 m in scale, involved as isolating exotic or in situ blocks, being cut by faults in various directions with strong deformation into breccia. A diorite knocker at Yamada (Yamada Knocker, 30 _~ 15 _~ 15 m) is in fault contact with serpentinite on the north which is thermally metamorphosed and tectonically sheared by diorite intrusion and dislocation. Two system shear zones of cataclasite to mylonite are remarkable. The first stage shear penetrates the rocks, but the second is only around the boundary. These planes are of normal-fault sense with strike-slip component, and are plotted on a small circle of a vertical pole. We interpreted that the first stage brecciation occurred during obducting stage of the ophiolite, and the second faulting during the rearrangement of the fault zone at the present setting as a forearc sliver fault. Such two stage deformation is common in all the blocks, including metamorphic, plutonic and basaltic rocks, suggesting the role of the knocker deformation is strongly control the oblique plate boundary tectonics.
T41C-1226 0800h
Structural Analysis of Broken Formation in the Shimanto Accretionary Prism Implies Oblique Subduction Tectonics
In order to understand strain distribution inside the strongly folded and deformed broken formation of the Cretaceous-Paleogene Shimanto accretionary prism Yakushima Island, Kyushu, Japan, we carry out detailed mapping of outcrop features based on high-resolution aero-photographs and AMS measurement on oriented core specimens. The so-called melanges, composed of folded and faulted turbidites, are classified into zone 1 and 2 in this area. The former contains blocks that preserve original bedding and development of layer-sub-parallel high strain shear zones in otherwise chaotic sequence, whereas the latter develops rootless folds and sheath folds. In zone 1, maximum principal axis of AMS fabric (Kmax) is disposed vertically inside vertically dipping bedding planes and Kmin normal to the bedding. Kmax orientations are sub-parallel to axes of vertical meso-scale folds inside zone 1. Zone 2 is highly deformed, but Kmax direction is also vertical as zone 1. Axes of rootless folds were plotted on a great circle on a stereonet projection suggesting development of sheath folds. Possibilities of these folds developed due to submarine landslide were ruled out because quartz veins normal to flattening surface of sandstone lenses and fan-shaped in the hinge zones were developed and in some cases silicified totally replacing original depositional texture of SS lenses. We attribute this to deformation under moderate T-P conditions and under influence of slica-saturated liquid. High-resolution mapping elucidates relationship between both zones; Zone 1 is distributed in upper part of thrust anticline and zone 2 beneath it. Since rootless folds in zone 2 rotated during high-strain deformation, we used relationship between vertical axes of cylindrical folds and AMS orientation of its hinge zones to estimate shape of strain developed during melange formation. In our analysis, Kmax was sub-parallel to the vertical fold axes and this implies maximum stretching direction was subparallel to axes of thrust anticline. We attribute this stretching parallel to fold axes to transpressional deformation during oblique subduction.
T41C-1227 0800h
Oblique Subduction of the Pacific Plate at the Kuril Trench and Lateral Movement of the Forearc
Northeast-southwest trending Kuril forearc in the northwest Pacific is formed by oblique subduction of the Pacific plate (PA) at the Kuril Trench. Its northwestern boundary corresponds to a volcanic front and southwestern front collides with the northeast Japan arc in the central Hokkaido. We estimate lateral movement of the forearc in two ways. At first we combine relative plate motion vectors with slip vectors of major interplate earthquakes at the southern Kuril Trench. Plate motion vector is evaluated at each epicenter using the most recent global model REVEL-2000 (Sella et al., 2002), and decomposed into two components: one parallel to the earthquake slip vector that is related to interplate earthquake cycle and margin-parallel residual that may be taken up by the forearc lateral movement. The residuals show that the southern Kuril forearc moves southwestward at a mean rate of 12mm/yr. In more detail, northeastern part of the forearc moves at a faster rate of about 20mm/yr, and the rate decreases to 10mm/yr at the southwestern front. This implies a deceleration due to the collision with the northeast Japan arc. Next we estimate forearc movement using crustal velocities from nationwide continuous GPS array though data are limited only in Hokkaido region. In horizontal velocity field the most dominant is a crustal shortening of the forearc in the direction of plate convergence, which can be modeled as an elastic deformation caused by back slip vectors acted on plate interface. We determine plate interface geometry referring to high-precision unified hypocenters by Japan Meteorological Agency and apply plate motion vectors from REVEL-2000 on the plate interface. Full plate coupling is assumed to a depth of 50km. After subtracting calculated elastic deformations from the observed velocities, residuals are nearly margin-parallel in the forearc and show drastic change at around the northwestern boundary. Mean of the residuals in the forearc is about 10mm/yr. The result is consistent with that of the first estimation.
T41C-1228 0800h
Variability in Coastal Neotectonics Along the Kamchatka Subduction Zone
The eastern coast of Kamchatka can be divided into several morphotectonic zones, which appear to correspond primarily to variations in the subducting crust, rather than to characteristics of the subduction zone such as rate of subduction or subduction angle. Pleistocene marine terraces and Holocene coastal stratigraphy along Kamchatka show variability in uplift (and subsidence) rate both at the degree-latitude scale, and at a scale of kilometers to 10s of km. The southern coast (51-53$^{o}$N, to Petropavlovsk) is primarily rocky headlands with narrow embayments, some filled with volcaniclastics. North of Petropavlovsk, the coast is subdivided into a series of broad embayments separated by four peninsulas--Zhupanovskiy, Kronotskiy, Kamchatskiy, Ozernoi, from south to north. These peninsulas are landward, respectively, of the Kruzenstern fracture zone, the Meiji seamounts, the Aleutian-Komandorskiy island chain, and the Beta Rise. Our field studies of Quaternary coastal evolution and seismotectonic regime over the last decade have included: 1) tephra chronology for dating and correlation 2) study of paleotsunami deposits to estimate recurrence rates of tsunamigenic earthquakes; and 3) analysis of the modern and paleo-topography of marine terraces and beach ridges in order to determine the direction and intensity of tectonic deformation over different spatial and temporal scales. The shorter the time interval we are able to specify (decades to hundreds of years), the higher the rate of vertical movements we tend to obtain. Estimates of net deformation for longer periods (10s to 100s of thousands of years) are commonly an order of magnitude slower, because seismic cycles are averaged out. The net Quaternary deformation of eastern Kamchatka is uplift, with peninsulas exhibiting the highest rates, up to 2 mm/yr for the last 0.5 my on Kamchatskiy Peninsula. The large embayments between peninsulas typically exhibit evidence for subsidence or stasis on a Holocene time scale. On Kamchatskiy Peninsula (56-56.7$^{o}$N), we have enough field sites to show that there is a similar (fractal?), smaller-scale segmentation of deformation. But why the nature of south Kamchatka, and the transition to the Kuril Islands?
T41C-1229 0800h
Observations of Seafloor Outcrops in the Oblique Subduction Setting of Adak Canyon: Implications for Understanding the Early History of the Aleutian Island Arc
Submarine canyons in the western Aleutians (west of 177°W) are formed by oblique subduction, which has broken crustal blocks away from the arc massif and rotated them in clockwise sense, resulting in the formation of triangular-shaped summit basins and deep, structurally controlled submarine canyons (Geist et al., Tectonics v7, p327, 1988). A series of dives with the ROV Jason II on July 28-30, 2004 on Adak Canyon has provided the first-ever view of seafloor outcrops in an Aleutian canyon formed by this process. Two dives on the canyon's steep eastern wall revealed extensive exposures of blocky outcrops of volcanic rock at depths of 2900-1500 m. Samples of these units collected by the Jason II are a mixture of dark, pyroxene and plagioclase-phyric lavas and volcaniclastics. Degree of weathering/alteration is highly variable but some samples appear fresh. We anticipate that these rocks are offshore-equivalents of the Finger Bay Volcanics, which represent the earliest phase of Aleutian volcanism exposed on nearby Adak Island (e.g., Coats, 1956, USGS Bull. 1028-C). Exposures of granitic rock in Adak Canyon form low ledges of exfoliating outcrop interspersed with spheroidally weathered, bouldery sub-crop, in the depth range of 1800-1600 meters. Obtaining in-situ samples from these massive and subrounded exposures was not possible with the Jason II, but recovery of large, sub-angular slabs that litter the surface included samples of fresh diorite, fine-grained felsic intrusives and hydrothermally altered volcanic country rock. The stratigraphically highest exposures observed in Adak Canyon are gently dipping, poorly lithified `Middle Series' sedimentary rocks of probable Miocene-Oligocene age. All outcrop surfaces in Adak Canyon are covered with a uniformly dark brown, opaque coating of Mn oxide less than 1mm thick. Well-rounded cobbles and boulders interpreted to be glacial drift are largely free of Mn oxide coatings. Thick pavements of Mn-oxide were not observed. These observations indicate that combined seismic and dredging-geochemical studies will be a successful approach to unraveling the complete magmatic and tectonic history of the Aleutian arc crust, which due to the absence of significant arc-parallel rifting and backarc spreading remains largely intact and available for study. Dating and geochemical analysis of samples recovered on the July '04 Jason II expedition will provide a much-improved view of early Aleutian history.
T41C-1230 0800h
Submarine Volcanic Cones in the Central Aleutian Arc: Relationship to Arc Rifting and Oblique Plate Convergence
Plate convergence along the 2200km Aleutian Arc varies from orthogonal at the Alaskan Peninsula to fully strike-slip on the west end of the arc. Deformation response of the upper plate to oblique convergence appears to accelerate markedly between Adak (177W) and Amchitka Pass (180W). On a regional scale, this deformation appears to be concentrated at the boundaries of crustal blocks, with clockwise rotation and westward translation [Geist et al., Tectonics 7, 327-341, 1988]. In the block rotation model, extensional rift structures develop between the blocks in arc-normal orientation. Summit basins develop at the northern, trailing edge of the blocks in arc-parallel orientation. These summit basins are located near or within the volcanic front. Thus structures in the upper plate driven by oblique convergence are predicted to interact with arc volcanism. We report on multibeam mapping in 2003-2004 and ROV Jason II dives in 2004. The data reveal locations and patterns of fault structures, volcanic cones, and lithologies in several locations critical to understanding the arc's response to oblique convergence. A large submarine volcano, named Amchixtam Chaxsxii in the Unangan language, was mapped next to Semisopochnoi Island. Additional small cones are identified on the flank of Tanaga Volcano, and near Bobrof Volcano on possible fault structures. The largest extensional `block boundary' is located at Amchitka Pass; in this area the seafloor is offset by a network of faults. Small volcanic cones are clustered at these faults. Some show signs of erosion and mass wasting; others, especially deeper ones, are intact. Surfaces are dominated by `a`a flows and spatter, and have light sediment cover and moderately fresh lavas. Our mapping focused on specific sites that were chosen to be representative, and suggests that (1) small, probably monogenetic cones are common; (2) the cones occur preferentially in areas of extensional faulting in the volcanic front; (3) these cones are present largely because of oblique convergence and arc deformation. Geochemical analyses will test their relationship to nearby subaerial arc volcanoes.
T41C-1231 INVITED 0800h
Crustal structure of transpressional margins based on along strike variations in kinematics and exposed level in strike-slip systems of southern Alaskan
Strike-slip systems dominate the Cenozoic tectonics of the northern North American Cordillera, and variable levels of exposure together with lateral kinematic variations make this region a premier area for studies of strike-slip processes. Neogene interactions between the Yakutat microplate and North America provide one of the most extreme examples on earth of an ongoing oblique-collision where the obliquity of convergence varies rapidly along strike. Here the St. Elias orogen occupies the transition from the Queen Charlottes transform to the Aleutian subduction zone and the microplate caught up in the transition shows clear spatial variations related to this variation in kinematics, but that process has clearly varied in time as the microplate has been carried into the strike-slip/subduction transition. Three important results from recent work on this orogenic system include: 1) evidence that slip partitioning recognized in the southern, highly oblique, segment of the orogen carries through to the core of the orogen where convergence angles are high; 2) strike-slip is transferred into a an oroclinal bend with complex structural overprinting including superimposed fold systems in unmetamorphosed sedimentary rock; and 3) development of a predominantly "one sided" orogen in which contraction is focused toward the outboard (oceanic) side of the orogen. The last appears to have been a long-term effect because exhumation rates from thermochronology and distribution of metamorphic assemblages suggest exhumation was focused on the rocks outboard of a strike-slip "backstop". This exhumation pattern is almost certainly the result of intense erosion on the oceanic side of the orogen driven by variable orographic precipitation, much of which is converted to glacial ice that profoundly influences erosion within the orogen. The St. Elias orogen also exposes older, deeply exhumed strike-slip systems that are surprising similar to the neotectonic system and provide insights into deep crustal processes. The Eocene Hanagita fault system, which represents a dextral reactivation of the Mesozoic arc-forearc boundary, shows mid to upper crustal level exposures of structures very similar to the St. Elias orogen-a dominantly one-sided, slip-partitioned "positive flower structure" along the eastern limb of the Alaskan orocline that transfers into a zone of complex overprinting in the "core" of the orocline. Finally, constraints on deep-crustal processes are provided in a spectacular down-plunge exposure of an Eocene forearc strike-slip shear zone that shows a low-angle mid-crustal detachment between uniformly deformed high-grade gneiss at depth and overlying amphibolite facies schist deformed by distinct ductile shear zones.
T41C-1232 0800h
Crustal Behavior in a Strike-Slip-Related, Mid-Crustal Attachment Zone: Evidence From a Down-Plunge Section in the Chugach Metamorphic Complex, Southern Alaska
The western termination of the Chugach Metamorphic Complex in southern Alaska provides an excellent example of a down-plunge crustal section that we are utilizing to examine strike-slip/transpressional plate boundary behavior. This exhumed section was developed in monolithologic forearc accretionary sediments (Valdez group), which eliminates rheologic complications associated with heterogeneous crust. Following an "attachment zone" model of Teyssier et al. (2003), we are developing models linking deformational history between rheologic layers of the crust, with particular focus on the development of mid-crustal decoupling. Previous field studies documented mid-crustal decoupling at the lithologic transition from schist to gneiss with homogeneous deformation below and localized deformation in shear zones above. Our recent detailed structural mapping has revealed high strain bands and a synformal foliation "cusp" that has first order similarities to the attachment zone model. The foliation "cusp" is formed by a 3-5km wide high-strain core of steep foliations with abundant dextral shear sense indicators that can be traced along strike for more than 60km, penetrating more than 4km of the crustal section. These steep zones pass outward, on both sides, through a 10-15km wide roll-over to nearly flat-lying foliation that parallels the underlying decoupling surface. This geometry is broadly consistent with foliation trajectories predicted for areas mid-way between two strike-slip shear zones in the attachment zone, but paradoxically, it appears that the steep-foliation is THE major dextral shear zone that penetrates the exposed crustal section; the reverse of the geometry predicted by the attachment model. Our continued work is emphasizing reevaluation of the attachment model (through redefining assumptions) using observed foliations, lineations, finite strain kinematic axes (to constrain the model), considering variable offset amongst the high-strain zones and variable thicknesses of the attachment zone, and variable crustal level geometry. Ultimately more field work may be needed to assess the continuity of the shear zone through the upper part of the crustal section where the discrepancy between observed and theoretical structures is most pronounced.
T41C-1233 0800h
Effects of Boundary Conditions and Convergence Geometry on Deformation Style at Oblique Convergent Margins
The accommodation at convergent margins of the partitioned components of oblique plate motion can result in spatially complex patterns of deformation. In thin-skinned fold-belts this can result in deformation that has distinct zones of nearly pure dip-slip thrusting at the deformation front and strike-slip within the fold-belt or in the accretionary wedge. Most convergent margins have along-strike variations in obliquity and corresponding variations in deformation style. We measure the horizontal strain field in a variety of frictional analogue models of oblique convergence in which we examine the kinematic and structural effect on deformation of buttressing due to changes in along-strike margin geometry. We find that even at oblique convergent margins that have a constant along-strike obliquity to plate motion considerable along-strike variations in deformation can result from boundary conditions associated with the finite length of the margin. Such boundary conditions affect the manner in which strain is accommodated over the entire length of the margin, not just locally. We find that significant along-strike variations in strain accommodation can also result from curved margin geometries commonly observed in convergent settings. At margins with such a geometry the gradual along-strike variations in plate motion obliquity will yield accommodation structures in the analogue models that transition from one structural style to another (e.g., margin normal compression to along-strike extension). While the local style of deformation will reflect the local margin obliquity the along-strike variations in deformation style may not be obvious in the gross morphology of the model fold-belts. In many cases the change in deformation style is only apparent in the strain rate or velocity fields.
T41C-1234 0800h
Oblique Subduction as a Controlling Factor for Stresses and Earthquakes in Subducting Slabs at the Cascadia and Nankai Subduction Zones
Obliquely subducting slabs experience viscous mantle resistance in both updip and margin-parallel directions. The updip component is locally balanced by downdip slab-pull, but there is no body force to balance the margin-parallel component locally. Margin-parallel slab resistance thus has a first-order impact on in-slab stresses. In-slab stresses and thermally controlled dehydration embrittlement are jointly responsible for intermediate-depth earthquakes. At both the Cascadia and Nankai subduction zones, young and warm slabs subduct at 4 - 5 cm/yr with a right-lateral obliquity, but oblique subduction affects in-slab stresses differently in the two places because of differences in age distribution and kinematics of the subducting plates. We determine stresses in these slabs by inverting earthquake fault mechanisms and investigate the different effects of oblique subduction. Along the E-W striking Nankai subduction zone, margin-parallel slab resistance pulls the slab eastward. The oldest part of the slab is subducting at the Kyushu margin to the west, and the slab pull force stretches the slab westward. The balance of these two forces results in a margin-parallel tension in the slab, consistent with the state of stress determined from focal mechanisms. Along the N-S striking Cascadia subduction zone, the slab resistance is in the southerly direction and, together with other boundary forces, gives rise to a torque that tends to rotate the slab clockwise. Shear force along the Nootka shear zone to the north generates a torque that tends to rotate the slab in the opposite direction. The torque balance results in an E-W stretch in the slab that decreases from north to south. The observed E-W tension in the northernmost part of the slab and a southward decrease in slab seismicity are consistent with the torque balance mechanism.
T41C-1235 0800h
Structural vergence variation and clockwise block rotation in the Cascadia accretionary wedge, offshore central Oregon
Along the Cascadia margin offshore Oregon, the structural vergence at the toe of the accretionary wedge varies from landward vergent offshore northern Oregon to seaward vergent across the southern Oregon margin. A transition zone between these vergence domains occurs along the central Oregon portion of the wedge, centered on the Hydrate Ridge region. We examine the past variability in structural vergence across the Hydrate Ridge region through detailed structural mapping using multichannel seismic reflection data and gridded bathymetry. These data are coupled to biostratigraphic age constraints obtained from ODP drilling to constrain the timing of accretionary wedge growth since the early Pleistocene (<1.7 Ma). Our results indicate that the wedge in the Hydrate Ridge region was accreted in three structural phases: an early Pleistocene seaward vergent phase (~1.7-1.2 Ma), an early to middle Pleistocene (~1.2-0.3 Ma) landward vergent phase, and a late Pleistocene-Holocene (~0.3-0.25 Ma to present) seaward vergent phase. Age constraints on the timing of landward vergent deformation suggest coincidence with the timing of the deposition of the Astoria fan. High pore fluid pressures due to rapid fan deposition have been suggested as the likely cause of landward vergence for the northern Oregon and Washington margins. The large bathymetric expression of northern Hydrate Ridge is likely due to its history of continued seaward vergence, which permitted some sediment subduction, likely underplating and observed thrust duplexing, all resulting in an increase in the thickness of the accretionary wedge (more uplift) beneath this region. Superimposed on the accretionary wedge growth in the Hydrate Ridge region, two basement involved transverse strike-slip faults have affected the wedge development. Evidence of clockwise block rotation of the Hydrate Ridge tectonic block between the two transverse strike-slip faults appears most pronounced in the older portion of the wedge, and decreases toward the west. Constraints on the timing of propagation of the basement strike-slip faults into the abyssal plain section near the deformation front indicate that the early-middle Pleistocene landward vergent phase (~1.2-0.3 Ma) may have been terminated by this faulting. We speculate that the propagation of the strike-slip faults into the upper plate may have reduced pore fluid pressures and increased coupling along the decollement, triggering the change from landward to seaward vergence at the deformation front.
T41C-1236 0800h
Sequential Kinematic Restoration as a Tool for Deciphering Evolving Plate Boundaries: the Western North America-Pacific Plate Boundary System
Orogens are often characterized by their most well known segment or most well described section. This tendency may inadvertently suggest that changes along strike may be anomalous, unique or wrong. Combining both cross section and plane view sequential reconstructions across several portions of an orogen allows us to track the evolution of a larger region in time and space. This four-dimensional reconstruction links together along strike changes and connects, in kinematically feasible ways, diverse portions of an orogen that may initially appear incompatible. This approach was tested on the Basin and Range province of western North America. The Basin and Range has been proposed to be the diffuse eastern edge of the Pacific-North American plate boundary with up to 20% of Pacific North America motion being accommodated east of the Sierra Nevada Mountains. However, how this percentage of plate motion is transferred from the Gulf of California through the southern Basin and Range/Mojave region to the eastern California shear zone is still not well understood. We obtained a permissible, kinematic history of the Basin and Range by compiling kinematic data (amount, timing and direction of displacement) along three transects through the northern (40\deg N) central (37\deg N) and southern (34\deg N) portions of the province. Extension and strike-slip deformation in all areas was sequentially restored using the kinematic data in an ArcGIS program over 2 m.y. to 6 m.y. time intervals. The process of sequential restoration highlighted misalignments, overlaps or large gaps in each incremental step, particularly in the areas between data transects. In areas where no information is available we use regions where the kinematics are known to constrain adjacent areas where the kinematics are not defined. The new sequential reconstructions show that compatible slip along the entire N-S extent of the inland shear zone from Baja to the northern Walker Lane is possible and supported by available data and that this inland shear zone had migrated westward with time. The reconstructions also highlight new problems particularly with regard to strain compatible extension east and west of the Sierra Nevada/Great Valley block.
T41C-1237 0800h
Range Front Faulting and Ancestral Cascades Arc Magmatism in the Central Sierra Nevada at 10 Ma: Onset of Basin and Range Extension or Sierran Root Delamination?
The volcanic rocks of Sonora Pass, Central Sierra Nevada, California, occupy a transition between Cascadian subduction and Basin and Range extension; these rocks thus provide a key to understanding the evolution of these provinces. Our preliminary work includes detailed mapping, 40Ar/39Ar geochronology on 15 mineral separates, and a geochemical study of 110 samples. Volcanic units are from the Mehrten and Stanislaus Formations. Within our study area, small shifts in alkalinity are apparent, and occur within a larger temporal pattern of increasing alkalinity with time. A shift from subalkalic to mildly alkalic, then back to subalkalic volcanism matches similar relative trends (of larger magnitude) in the southern Sierra, which have been interpreted as representing lithosphere delamination. On a regional scale, Sonora Pass volcanics are also intermediate in alkalinity compared to Pliocene-age volcanics from the southern Sierra Nevada and volcanic rocks from Lassen; on the basis of K2O vs. SiO2, three very distinct magmatic suites are evident: a relatively low K2O Lassen series, an ultrapotassic Pliocene series, and an intermediate series represented by Sonora Pass. But, the origin of these K2O variations is unclear. At Lassen, Sr/P ratios are a positive index of slab-derived fluids (Borg et al., 2002), and such fluids have been called upon to explain elevated K2O contents of Pliocene Sierra volcanics (Farmer et al., 2002). But our data show an inverse correlation between K2O and Sr/P. Also, a granite xenolith from the Little Walker Caldera indicates that the crust might not provide an alternative reservoir to explain the K2O-Sr/P systematics. Isotope and trace element analyses are being undertaken to improve our understanding of the major element geochemistry. Miocene strata of the Sonora Pass area record the onset of faulting both during and after eruption of the andesitic Early (?)-Late Miocene Mehrten Fm, with continued faulting after eruption of the latitic Late Miocene Stanislaus Fm (Busby et al., 2003; Rood et al., 2004). At Sonora Peak, an angular unconformity exists between the Mehrten Fm and the overlying Table Mountain Latite flows of the Stanislaus Fm. Exposures of subvertical fault planes restricted to the Mehrten Fm show shallowly-plunging lineations and Riedel shears that suggest dextral deformation. Such dextral strike-slip motion on faults as old as 10 Ma may record the birth of the Walker Lane fault zone in the region. Our new field, 40Ar/39Ar, and geochemical data indicate a rapid change in the structural and geochemical setting of the Sonora Pass area at 10.25+/-0.06 Ma coeval with eruption of the Table Mountain Latite flows. This was followed by large-volume eruptions of high-K quartz latite (Eureka Valley Tuff) between 9.30+/-0.03 Ma and 9.16+/-0.03 Ma. Latitic volcanism was both preceded by andesitic arc volcanism, as young as 10.10+/-0.06 Ma, and followed by andesitic arc volcanism at 7.12+/-0.06 Ma. Latitic volcanism at 10.25 Ma cannot be attributed to a slab window because the triple junction was located well to the south of Sonora Pass at that time. We speculate that the latites erupted during a phase of rapid extension in an arc otherwise dominated by andesitic volcanism, perhaps indicating the initiation of Basin and Range faulting or the delamination of the Sierra Nevada batholithic root.
http://www.geol.ucsb.edu/faculty/busby/VSSAC-FR.html
T41C-1238 0800h
Geodynamics Of Mid-Tertiary Extensional Phase In Southwest Texas And Relationship With The Rio Bravo Fault
Three independent sets of data lead us to conclude that gravity collapse alone cannot account for the Cenozoic evolution of the Texas margin of the GOM and that there has been since Paleocene a significant reactivation of the extension there.1) We have examined an extensive set of thermal data from wells including 2000 Reservoir temperatures offshore Texas and Louisiana. Solving for 1-D thermal and stratigraphic evolution of 166 representative wells we obtain the basal heat flow that defines the existence of a well defined anomaly centred on the Corsair Fault. The basal heat flow increases over less than 100 km from 35 mW/m2 northwest of the Corsair Fault to 55 mW/m2 on the fault. This increase is best explained by a crustal extensional episode during Upper Cenozoic as demonstrated by a simple modelization. The thermal structure results in very high temperatures at depth. The deepest wellls at 6000 m depth give a temperature larger than 200C. Below the Corsair rift, the extrapolated temperature is more than 300C at 10 000 m (6 s twt), 375C at 12.5 km (7 s twt) and close to 500C at 18.5 km (9 s twt). 2) Velocity/depth data from refraction (Ebeniro et al., 1988), from Moho inversion based on gravity and from 11s TWT seismic reflection depth sections show that the deep decollement layer is indeed very deep and very hot and that there is little if any igneous crust below the Corsair Rift. The velocity structure lead us to conclude that the major decollement that is generally identified with the Middle-Cretaceous Unconformity (MCU) and that lies at a depth of 7 s twt increasing to 9 s below the Corsair Rift plunges from 12-13 km northwest of the Corsair Fault to 18-19 km below the Corsair Rift. There the Moho is localized at 21-22km. The 3 km thick material between these two depths could potentially include the Cretaceous, Jurassic and Triassic as well as the whole igneous crust. In any case at this depth the present extrapolated temperature is about 500°C. The brittle-ductile transition at the present time is then expected to be situated within the sedimentary section and the brittle - ductile transition is probably situated between 6 and 8 s twt. The material, whatever its composition, below the main decollement in the area of the Corsair Rift must be metarmorphized and ductile.3) Field data led to the identification a major left-lateral shear zone most active during the late Eocene-Oligocene that we have called the Rio Bravo Fault zone. The fault zone had been previously described in the literature as the N120° Texas lineament assumed to have been inherited from the Jurassic opening of the Gulf of Mexico and located at the boundary between Texas and Mexico, approximately coinciding with the Rio Bravo (or Rio Grande).We demonstrate that this zone of shear was active during Oligocene from about 31°N to about 25°N. We conclude that an approximately 1000 km long left-lateral shear zone was active during mid-Tertiary with a total offset of 40-60 km. Its activity affected the Tertiary depocenters in Texas and within the Burgos Basin. It could account for the Paleocene to Oligocene extension in Southwest Texas.
T41C-1239 0800h
Basement Control of Fault Orientation and Style of Deformation Along the Active Strike-Slip Boundary of the Inner Southern California Borderland, Gulf of Santa Catalina
Although the active faults of the inner southern California borderland south of Los Angeles appear to comprise a relatively simple system of northwest-trending strike-slip faults, closer inspection of these faults reveals a more complicated fault pattern than might be expected from models of simple shear. The major fault systems of the inner borderland include the Palos Verdes, Coronado Bank, Newport-Inglewood-Rose Canyon, and San Diego Trough fault zones. Seismic reflection profiles show that these fault zones extend over distances of 60-150 km and generally trend about N$30\deg$W, about $10\deg$ more northerly than predicted by relative Pacific-North America plate motions. The fault zones vary in structural style from the relatively simple, continuous San Diego Trough fault zone in the west, to more complicated fault zones to the east that are typically comprised of short, often multi-stranded, fault segments. These faults show significant variations in deformational style along strike, involving right and left stepovers, changes in fault orientation, and variations from transpressive to transtensive faulting. Structures adjacent to these faults also appear to show complicated histories. A comparison between the location of active fault traces with respect to prominent, partly subsurface, basement ridges and knolls composed of middle Miocene and older rocks show that the pre-existing basement ridges partly control the deformational style of modern fault-related structures. The ridges themselves are probably relict features of early to middle Miocene extension and related volcanism.
T41C-1240 0800h
Subsidence and Strike-slip Tectonism of the Upper Continental Slope off Manzanillo, Mexico
The direction of convergence between the Rivera and North American plates becomes progressively more oblique (in a counter-clockwise sense as measured relative to the trench-normal direction) northwestward along the Jalisco subduction zone. By analogy to other subduction zones, the forces resulting from this distribution of convergence directions are expected to produce a NW moving, fore-arc sliver and a NW-SE stretching of the fore-arc area. Also, a series of roughly arc parallel strike slip faults may form in the fore-arc area, both onshore and offshore, as is observed in the Aleutian arc. In the Jalisco subduction zone, the Jalisco block has been proposed to represent such a fore-arc sliver. However, this proposal has encountered one major problem. Namely, right-lateral strike slip faulting within the fore-arc sliver, and between the fore-arc sliver and the North American plate, should be observed. However, evidence for the expected right-lateral strike slip faulting is sparse. Some evidence for right-lateral strike-slip faulting along the Jalisco block-North American plate boundary (the Tepic-Zacoalco rift system) has been reported, although some disagreement exists. Right-lateral strike-slip faulting has also been reported within the interior of the Jalisco block and in the southern Colima rift, which forms the SE boundary of the Jalisco block. Three-fold, multi-channel seismic reflection data were collected in the offshore area of the Jalisco subduction zone off Manzanillo in April 2002 during the FAMEX campaign of the N/O L'Atalante. These data provide additional evidence for recent strike-slip motion within the fore-arc region of the Jalisco subduction zone. This faulting offsets right-laterally a prominent horst block within the southern Colima rift, from which we conclude that the sense of motion along the faulting is dextral. These data also provide additional evidence for recent subsidence within the area offshore of Manzanillo, as has been proposed.
T41C-1241 0800h
Varying Rates and Modes of Subduction Erosion Along the Peruvian Margin
At least half of the world's active margin length now is attributed to the erosive type with regard to their mass transfer modes. However, the mechanisms, loci and rates of subduction erosion still are not fully understood. Among factors contributing to subduction erosion, subduction of asperities, roughness of the downgoing plate, and rheological properties of the overriding plate are thought to be of major importance in determining its styles and amounts. The Peruvian margin, the southern portion of which has experienced collision of the Nazca Ridge, is an exceptionally suitable location to study subduction erosion. Here it is possible to compare portions of a margin that have either been affected or not affected, respectively, by the subduction of major asperities. Swath bathymetry data acquired during RV Sonne cruise SO146 reveals high, but regionally different roughness of the almost sediment free Nazca plate. RMS roughness, local dip, curvedness, and void volume, i.e. the volume between peaks and valleys potentially filled with eroded material, decrease from north to south. Recently acquired wide angle data now together with information on long-term subsidence from ODP Leg 112 Site 683 enable to estimate the rates of subduction erosion across the Peruvian margin at 9\deg S, a region which was not affected by ridge subduction. Rates are 24 to 30 km$^{3}$ km$^{-1}$ myr$^{-1}$ since the middle Miocene, 15 km$^{3}$ km$^{1}$ myr$^{-1}$ since about 40 Ma and 6.5 km$^{3}$ km$^{1}$ myr$^{-1}$ for the interval 40 to 13 Ma. These rates are considerably smaller than published long-term rates (since 40 Ma) estimated further to the south (12\deg S) or short-term (since the Pliocene) erosion rates estimated further north at about 7\deg S. However, at 9\deg S, the void volume is not much less than the eroded volume, whereas at 12\deg S, the eroded volume is about 3 times the void volume. This comparison reveals that different mechanisms of subduction erosion, which are suggested to be attributed to differences in the rheology and strength of the overriding South American plate, take place along the Peruvian margin.
T41C-1242 0800h
Thermochronologic and Sedimentologic Evidence for Variations in Exhumation of the Cordillera Blanca Detachment Fault System, Peru
Thermochronologic and sedimentologic data for the 6-km-high Cordillera Blanca, an active detachment fault system in Peru, indicate rapid along- and across-strike variations in late Cenozoic exhumation. The Cordillera Blanca is composed of an $\sim$8 Ma ($\sim$3 kbar) granodiorite batholith in the Andean hinterland above the subducted flat Nazca slab. The range forms the footwall of a detachment fault striking $330\deg$, oblique to Neogene $082\deg$ Nazca-South America convergence. Previous studies indicate mainly dip slip along the 20-45$\deg$ west-dipping fault and west-down shear sense within a footwall mylonite. Total fault offset is poorly constrained but may exceed 10 km of normal slip. $^{40}$Ar/$^{39}$Ar analyses of granitic footwall rocks collected along several transects reveal considerable variability across and along strike. Across strike, samples 5-10 km from the fault show cooling shortly after $\sim$8 Ma batholith emplacement, 1-2 Myr prior to cooling of samples collected $<$2 km from the fault. These results are consistent with footwall tilting during faulting. Along strike, samples from the central fault segment show rapid cooling from $>$300$\deg$C at 5-6 Ma to below $\sim$150$\deg$C by 2-3 Ma, about 1-2 Myr prior to samples from the south. We tentatively interpret maximum slip along the central fault segment with diminished slip to the north and south, but additional data are required. Investigation of the hanging wall (Callejon de Huaylas) also reveals along-strike variability. New $^{40}$Ar/$^{39}$Ar ages for the $\sim$5.4 Ma tuff at the base of the $\sim$1300 Lloclla Formation (southern segment) within a supradetachment basin are younger than new ages for the $\sim$7.5 Ma Yungay tuff (central segment) that unconformably overlies Jurassic-Cretaceous strata. Although previous workers correlate these two tuffs, the new ages reveal discrepant histories for the central and southern hanging wall. These relationships suggest extensional basin development limited to localized regions along the detachment, implying no direct correlation between slip magnitude and sediment accumulation.
T41C-1243 0800h
Along-strike Variations of Subduction Parameters at the Chilean Plate Boundary
Newly compiled data of the geometric, kinematic and mechanic properties and their variations along-strike the oblique Chilean subduction margin between $20\deg$S and $46\deg$S are used to weigh their competing influence on forearc deformation. Special emphasis lies on the formation of margin-parallel strike-slip systems. Among the parameters considered are the convergence rate and obliquity, the ocean floor age, the dip of the down-going and the slope of the overriding plate, the geodetic and seismic coupling depth, the interplate seismicity, the depth of the trench-fill and the mass transfer mode at the subduction front. Commonly discussed control factors for forearc deformation can be attributed to three major elements of a subduction system, namely (1) the plate kinematic boundary conditions, (2) the plate coupling properties that govern the effectiveness of force transmission from the subducting plate to the overriding plate, and (3) the upper plate heterogeneities affecting its rheology (e.g. elasticity, shear strength) or resistance to block motion (buttressing). An example is given for each of these elements: (1) Oblique convergence is a pre-requisite for the activation of margin-parallel strike-slip systems, but apparently not a sufficient condition. For example, strike-slip motion can presently be observed along the Liqui\~{n}e-Ofqui Fault Zone in southern Chile, while neither the Atacama Fault Zone nor the Precordilleran Fault System in northern Chile accommodate significant amounts of margin-parallel slip since the Pliocene. This difference can not be explained by variations of convergence rate or obliquity as the plate kinematic framework is almost constant along the Chilean trench. (2) The plate coupling force is a function of the frictionally coupled area on the plate interface and of the shear friction that needs to be overcome. Along the Chilean margin various factors affect coupling in opposing manner: The slab-dip is shallower in southern Chile compared to northern Chile, resulting in a greater plate contact area. On the other hand, subduction of younger and hotter oceanic plate in the south could limit the frictionally coupled area (counter acted by increased buoyancy forces?). Subduction of wet sediments in the accretive margin of southern Chile compared to the erosive margin in the north may additionally weaken the interface. (3) The trenchward concave-shaped margin in North-Chile likely hampers margin-parallel motion of a forearc sliver, while strike-slip faulting may be supported in southern Chile due to the lateral proximity of the downdip end of coupling on the plate interface and the rheologically weakened zone of the active volcanic arc. Establishing the current state of plate coupling in southern Chile compared to northern Chile thus remains ambiguous. Margin-parallel strike-slip activity in southern Chile, however, may be facilitated by superposition of two conditions: a shallow-dipping slab that transfers stresses at the base of the overriding plate further arcward and an exceptionally close position of the arc to the trench.
T41C-1244 0800h
West Antarctica record of oblique convergence along the Cretaceous proto-Pacific margin of Gondwana
In Marie Byrd Land, West Antarctica, mafic alkalic dike arrays and A-type granitoids of 115 to 95 Ma age are typically viewed as a record of continental extension and as a direct precursor to orthogonal breakup between New Zealand and Marie Byrd Land at circa 70 Ma. New kinematic data for a mafic dike array in the Ford Ranges, an approximately 1000 km2 region dominated by plutonic and metamorphic bedrock, provide a mean dike trend of N16W. This corresponds to a maximum finite strain axis oriented N74W, highly oblique to the Marie Byrd Land margin and to major fault trends in the Ford Ranges. The stretching direction is sub-parallel to the N65W strain axis determined from structural analysis of folds and fabrics in a migmatite dome in the northern Ford Ranges. 40Ar/39Ar dolerite dike emplacement ages are 146-97 Ma, broadly coeval with emplacement of the dome and with development of the eastern Ross Sea rift. The oblique orientation of maximum finite strain with respect to large faults, geophysical lineaments and the rifted margin of western Marie Byrd Land suggests that transcurrent tectonics were in effect along this segment of the Gondwana margin in the middle Cretaceous, prior to breakup. The Ford Ranges result contrasts with that from a contemporaneous dolerite dike array in central Marie Byrd Land, 300 km to the east, which records stretching orthogonal to the rifted Marie Byrd Land margin, but is compatible with kinematic results from the Antarctic Peninsula. Consequently West Antarctica fits the pattern of along-strike variations in kinematics and tectonics related to oblique convergence along the Cretaceous circum-Pacific margin.