T41A-1156 0800h
Deformation of the Manihiki Plateau, Western Equatorial Pacific Ocean
The submarine Manihiki Plateau, an oceanic large igneous province (LIP), encompasses $\sim$800,000 km$^{2}$ of seafloor in the western equatorial Pacific Ocean. Of Early Cretaceous ($\sim$120 Ma) age, the plateau stands several kilometers above the surrounding seafloor, and comprises three major structural highs. The `High' plateau to the east contains several islands, including the eponymous Manihiki atoll. To the west, the `Western' plateaus lie in slightly deeper water, and are separated from the High plateau by a long, northeast-trending, bathymetric low known as the Danger Islands troughs. North of the Western plateaus is the small, nearly separate `North' plateau that is separated from the High plateau by a deep ocean basin. In August/September 2003, we acquired $\sim$25,000 km$^{2}$ of multibeam bathymetry and reflectivity, together with magnetics and gravity, data along the Danger Islands troughs and northwestern margin of the High plateau aboard R/V {\it Hakuho Maru}. On the basis of these data, we propose a tectonic model for post-emplacement (Cretaceous?) deformation of the Manihiki Plateau involving both normal and transform faulting. The deep basin separating the High and North plateaus probably formed by seafloor spreading; the northwestern and southeastern margins of the respective plateaus appear to be conjugate rift margins, whose physical separation approximately equals the right-lateral offset between the southern flanks of High and Western plateaus. The curvilinear southwest edge of the deep basin abuts the steep northeast flank of the Western plateaus, and in our interpretation represents a fracture zone juxtaposing LIP and normal oceanic crust. The right-lateral fracture zone continues uninterrupted to the south of the deep ocean basin as the Danger Islands troughs, which in the study area comprise a series of major en echelon, right-lateral faults that step to the right, producing extensional relay zones and pull-apart basins between fault tips where displacement was transferred. Geochronology of igneous rock samples dredged from the Danger Islands troughs during the {\it Hakuho Maru} expedition may yield information on the timing of deformation, which would in turn contribute to a better understanding of regional Cretaceous plate kinematics.
T41A-1157 0800h
Tectonic Setting of the Danger Islands Troughs in the Manihiki Plateau
The Manihiki Plateau, which lies in the western equatorial Pacific Ocean, is considered to be one of the Cretaceous Large Igneous Provinces, which formed at or near a triple junction (Winterer et al., 1974). Three major geomorphic plateaus, High, North, and Western plateaus, are discernible within the plateau itself. The depth of the Western Plateau, about 4000 m, is deeper than those of the other plateaus, about 3000 m. The High Plateau is separated by linear deep narrow depressions named the Danger Islands Troughs (Mammerickx et al., 1974). The Danger Islands Troughs is thought to be a trace of the plate boundaries (Winterer et al., 1974). Geophysical and geological investigations were conducted by R/V Hakuho-maru, Ocean Research Institute, the University of Tokyo, in August 2003. The bathymetric survey by the SEA BEAM 2120 multi-narrow beam echo sounders was carried out along the Danger Islands Troughs (DIT) between 9__deg$40'S and 6__deg$30'S. The interval of the track lines is about 5 km. Geomagnetic and gravity fields were measured during the bathymetric survey. Volcanic rocks were dredged at the four sites of the slopes of the troughs. Our bathymetric survey exposes the detailed topographic expression of the DIT. The topographic expression of the DIT is a trough bordered by ridges with a height of 2500 m above the floor of the troughs. The depth of the northern part, north of 7__deg$30_fS, of the DIT is 5900 m. That of the southern part of the DIT is 4800 m. There is a seamount which high is 2500 m between the northern and southern parts of the DIT. O investigations indicates that the DIT is not an extinct normal spreading system, but a fracture zone which is a trace of an oblique spreading system or a leaky transform fault.
T41A-1158 0800h
Geology and Structure of the Southern End of the South Fiji Basin, SW Pacific Ocean
The opening of the South Fiji Basin (SFB) in relation to the Norfolk Basin, and the eruptive activity and migration of the Three Kings Ridge (TKR) are long-standing Cenozoic SW Pacific tectonic problems. Seven research cruises in the past 9 years by New Zealand, France and Australia have collected multichannel seismic, multibeam bathymetry, and dredge samples in the Norfolk to southern SFB region, providing a wealth of new data, most of which has yet to reach the mainstream literature. Results presented here focus on the data gathered close to New Zealand. The Northland Plateau, at the southern end of the SFB, is a large, NW trending terrace that forms the largest feature of the northern New Zealand continental slope. It appears to contain tectonic elements of the Three Kings Ridge and Norfolk Basin as well as the SFB. It is divided longitudinally into an inner sedimentary basin and an outer volcanic rise, generally matched by inner negative magnetic and gravity anomalies, and outer positive ones. On the southeastern side of the TKR, we have found NNW trending normal fault terraces, from which we have dredged Oligocene (32 Ma) basaltic andesite; evidence of a rift margin with the SFB. On the west side of the TKR, N-S trending terraces are also extensional; peridotite, and a spectrum of subduction related rocks including BABB, boninite and shoshonite has been reported from them. Terraces are absent on the Northland Plateau but linear gravity trends that correspond with the terraces flanking the TKR appear to merge with NW gravity trends on the Northland Plateau. Some NW gravity trends are due to the Vening Meinesz transform, which cuts the S end of the TKR and appears to continue southeastwards under the continental slope. Within the extended region, ultramafics are suspected in an (undredged) horst trending towards the northern tip of NZ, where such peridotite crops out, possibly linking the onshore ophiolite to the western TKR suite. Schist and granite, outcropping further southeast along the plateau, have been attributed to a continental margin metamorphic core complex with 10 km of exhumation between 23-21 Ma. The presence of a transform in this region introduces the possibility of strike slip exhumation. Paleodepths from fossils indicate 500 m to more than 1500 m of Early Miocene subsidence in the extended Northland Plateau terrains, and 2500 m of subsidence on the western Three Kings terraces facing the Norfolk Basin. Dredged lavas, exclusively Early Miocene calc-alkaline rocks, including both low-med-K basaltic andesite and shoshonite, link the eruptive history of the outer half of the Northland Plateau to that of the Northland Arc and to the Early Miocene phase of the Three Kings arc. Strong positive magnetic anomaly trends suggest that the Northland Plateau volcanics might be continuous only with the TKR volcanics, but also suggest that SFB magma may have contributed to the plateau. The large negative magnetic anomaly of the inner Northland Plateau is radically different from those of the adjacent Norfolk Basin. Its lateral extent coincides with that of the Northland Allochthon and may represent the parautochthon, which disappears under and is also possibly faulted against the Early Miocene outer plateau volcanics.
T41A-1159 0800h
The Wishbone Ridge
The Wishbone Ridge gravity lineation appears on global gravity maps as the left-most branch of a series of forked gravity anomaly lineations which extend c. 1300 km from the Osbourn Trough at 25 degrees S to the Chatham Rise at 43 degrees S. The northeast trending gravity lineation forms a c. -40 mgal gravity anomaly southwest of the lineation intersection with Louisville Ridge at 39 degrees S. It also coincides with a bathymetric ridge rising up to 2 km above the neighbouring seafloor and extending from the Chatham Rise as far north as 40.3 degrees S. Swath bathymetry surveying indicates the bathymetric ridge is comprised of a series of southeast facing, tilted, en-echelon fault-blocks. Seismic reflection profiling across the south-eastern extension of the lineation reveal this tilted block structure extending south-westwards beneath the north-eastern crest of the Chatham Rise. The extent and continuity of the gravity lineation suggests the lineation marks a major crustal boundary. The tilted en-echelon blocks of the bathymetric ridge are suggestive of an extensional-strike slip fault. Seismic reflection profiling across the lineation north of its bathymetric exposure illustrates a major contrast in character and depth of the oceanic crust either side of the lineation consistent with transverse displacement along the lineation boundary. The northeast trend of the Wishbone lineation is mirrored in parallel volcanic ridges on the adjacent Hikurangi Plateau LIP and on the oceanic crust further north as well as in offsets of the Louisville Seamount trend. Offsets in the eastern Chatham Rise bathymetry also mirror this widespread trend suggesting motion along the Wishbone lineation may have originally broken through the Cretaceous North Chatham Rise Gondwana convergent margin forming the location for subsequent New Zealand-Antarctic rifting.
T41A-1160 0800h
Enriched MORB in the Northeastern Pacific, Petrological and Geochemical Features of igneous Basement at Site 1224, ODP Leg200
During Ocean Drilling Program (ODP) Leg 200, 45-Ma igneous basement was cored in the northeastern Pacific at Site 1224. The basement surface was assumed to be 28 m below seafloor (mbsf). Basement lithology down to 170 mbsf is divided into three units: Unit 1 massive flow, Unit 2 pillow breccia, and Unit 3 massive flow. The shallowest Unit 1 shows massive structure with some altered layers and vein deposits, and core recovery was 30%-50%. This unit is divided into two thick lava flows based on grain size stratigraphy and alteration layers. Intermediate-depth Unit 2 shows different characteristics than Unit 1. Core recovery was so low (<10%), and many of the small pieces retain almost circumferential alteration. Some calcite-cemented hyaloclastite layers were found. The difference in induration between the two units was so great that the drill bit broke at the Unit 1/Unit 2 boundary. The deepest Unit 3 shows features similar to Unit 1. Portions of at least two cooling units were recovered, and possibly more. Recovery in Unit 3 was lower than that in Unit 1 but higher than that in Unit 2 (20%~30%). Bulk compositions of Site 1224 rocks show interesting characteristics, the most important of which is high high-field-strength element (HFSE) content compared to typical normal and enriched mid-ocean-ridge basalts (MORB) (Sun and McDonough, 1989). Chemical stratigraphy (chemostratigraphy) differences among the three units at this site are clear. Unit 3 rocks are fractionated, and Unit 2 rocks have relatively primitive FeO/MgO ratios. Large-ion-lithophile element (LILE) content is higher Unit 2 than in Units 1 and 3. The high LILE content of Unit 2 is thought to be caused by hydrothermal alteration rather than a petrographic feature. HFSE content patterns are similar to those of FeO/MgO ratios. Unit 1 is separated into upper and lower subunits, which correspond to flow units, on the basis of HFSE content. Unit 2 has lower HFSE content and higher Y/Zr ratios, thought to result from a magma generation environment that was different from other units. The parent mantle of Unit 2 basalt was probably of a different, depleted composition compared to Units 1 and 3. Perhaps the most interesting result from this site is isotope characterization. The Sr and Nd isotope ratios are more enriched than typical Pacific MORB (Hickey-Vargas et al., 1995). These characteristics are attributed to mantle enrichment, and this enriched component is thought to still exist in the Pacific mantle. Many drill sites in the eastern Pacific show compositions similar to N-MORB (Sun and McDonough, 1989) (i.e. ODP Leg148; Brewer et al., 1996). Brewer et al. (1996) studied chemostratigraphy of the basement at from Holes 896A and 504B. They concluded that variations in chemostratigraohy were mainly caused by differentiation. When the Leg148 sites and Site 1224 results, are compared, Site 1224 basement shows more than twice the HFSE content than Sites 896 and 504, and the compositional changes is noted at Sites 896 and 504 differ from those of Site 1224. It is thought that the mid-ocean-ridge volcanism at Site 1224 was produced from more enriched mantle than that of the recent Eastern Pacific Rise and that these activities were associated with different mantle sources simultaneously. Chemostratigraphic and lithologic differences between the basement units correlate to differences in physical properties between the three units. Each lithologic unit displays different P-wave velocity, bulk density, and other physical properties. Physical properties are thought to be associated with petrological features.
T41A-1161 0800h
Plate Kinematic Model and Resulting Evolution of Oceanic Transform Fault Tectonics Adjacent to the Macquarie Triple Junction, SW Pacific
This research presents reconstructions illustrating the configuration of the Australian-Pacific-Antarctic plate system and the Macquarie Triple Junction (MTJ) at 7 times from 33.3 Ma to the present (A13o, A8o, A6o, A5o, A3Ay, A2Ay, and 0). Plate kinematic models are used to characterize the style of plate interaction between the three plates as the dominantly Ridge-Transform-Transform triple junction evolved. Models of MTJ evolution presented here focus on the complex crustal interactions that occurred at the southern Australian-Pacific plate boundary during this time, and constrain the tectonic evolution of the obliquely-convergent transform boundary that actively constructs the southern Macquarie Ridge Complex. Surprising results include the apparent shortening of the easternmost Southeast Indian Ridge (SEIR) by ~200 km during the lengthening of the transform due to southern migration of the MTJ. Such behavior explains the curved shape of the transform that evolved into the Hjort Trench and may be one process by which the Hjort region has accommodated convergence during its transition from transform fault to obliquely convergent incipient subduction zone. Differences between the amount of observed subduction at the Puysegur and Hjort Trenches may be a result of the partial accommodation of convergence by shortening of the easternmost SEIR spreading segment adjacent to the Hjort Trench during transform fault evolution. Reconstructions indicate that the Antarctic-Pacific boundary was a transform fault connecting the Antarctic-Pacific spreading ridge system with the SEIR for times between 33.3 and 5.9 Ma. Between 26.6 and 10.9 Ma the Antarctic-Pacific boundary was likely a leaky transform, accommodating some divergence between the two plates. At 5.9 Ma, a change in Antarctic-Pacific relative plate motion caused the transform fault to evolve into a spreading ridge. The spreading since 5.9 Ma served to more efficiently link the SEIR and Antarctic-Pacific spreading ridge systems, as the relative motion of the Antarctic-Pacific and Antarctic-Macquarie plates became similar in direction and magnitude.
T41A-1162 0800h
Seismic and Acoustic Evidence of a Large Seafloor Gas Seep on the Miyako Section of the Okinawa Trough
A marine geophysical survey was carried out, by the R/V Science 1 of the Institute of Oceanography, Chinese Academy of Sciences (IOCAS), in the year 2000, on the Miyako Section of the Okinawa Trough. Here we present seismic and acoustic evidence of a gas seep on the seafloor on the western part of the Okinawa Trough, near the lower slope of the East China Sea Slope, and discuss the possibility of related formation of gas hydrate. A gas column reflection was observed in echo-sounder data above a section where the seafloor reflector was missing, in both the echo-sounder and the seismic data for line H14. The seismic data also show an acoustic curtain reflection and a turbidity reflection at this section. These anomalies are evidence of the existence of a gas seep, which occupies an area 2.2km in diameter. Based on the acoustic curtain on line H14, we believe that the amount of gas contained in the sediments below the gas seep is larger than 1% by volume of sediment. Tectonically, the gas seep developed in a small basin controlled by basement uplift in the north, south and in the east. The thickness of the sediment layer can be greater than 3.5km. A mud diapir structure was found in layer D beneath the gas seep. Over-pressure may occur due to the large sediment thickness and also the tectonic basement uplift in the north, south, and east. The mud diapir could be the preferential pathway for methane-rich fluids. The acoustic curtain may indicate that free gas related to the gas seep can be formed on the seafloor. Judged from the temperature and the pressure, and the estimated heat flow data, a gas hydrate layer exists above the acoustic curtain seems reasonable.
T41A-1163 0800h
Bathymetry and Seafloor Reflectivity of the Intra-transform Spreading Center Along the Rivera Transform at 108°15'W
The first proposal of an intra-transform spreading center along the Rivera transform at 108°15'W was based on microseismicity data. During projects BART and FAMEX, conducted during 2002 aboard the N/O L'Atalante, 100% of the Rivera transform was covered by multibeam bathymetric and seafloor reflectivity data. These data show that the Rivera transform is marked by a broad, deep basin between 107°30'W and 109°12'W and can be subdivided into two smaller basins based on morphologic characteristics. The previous proposed intra-transform spreading center is located within the eastern basin which lies between 107°30'W and 108°35'W. This basin is approximately 110 km long, 18 to 20 km wide and exhibits a rhomboidal shape elongated in a NW-SE direction. Broad sediment filled depressions lie at the basin's eastern and western sides. The surface sediments in these depressions are undisrupted. Narrow bathymetric troughs lying at the base of prominent escarpments form the basin's northern and southern sides. The reflectivity data suggests that along the northern trough recent faulting is occurring only west of the proposed intra-transform spreading center, whereas along the southern trough recent, discontinuous faulting is occurring only east of the spreading center. The seafloor within the central part of the basin is rough and domed shaped. Seismic reflection and seafloor reflectivity data indicate a lack of sediments over the dome. This dome is broken by a deep sediment free trough at the location of the intra-transform spreading center indicated by the microseismicity data (108°15'W). Several small discontinuous faults trending oblique to the trend of the northern and southern boundary troughs are observed within the basin. The above characteristics suggest that the intra-transform spreading center does indeed exist. The magnetic lineations associated with this spreading center are difficult to model. However, assuming a spreading rate of 7 cm/yr, as is predicted from several proposed Rivera-Pacific relative motion models, the magnetic anomaly pattern is consistent with an initiation of spreading sometime between Chron 2 and the Jaramillo event, and that spreading has been asymmetric (2 cm/yr to the west and 5 cm/yr to the east). The discontinuous faulting along the southern trough and the presence of small discontinuous oblique faults in the center of the basin seems indicate a recent reorganization of the transform boundary. This reorganization may involve an extinction of the intra-transform spreading and the formation of a single transform running through the center of the basin. In support of this proposal, the Rivera transform east of the basin is very well fit by a single small circle and that the projection of this small circle runs through the center of the basin, oblique to the trend of the northern and southern boundary troughs.
T41A-1164 0800h
Characteristics and Possible Triggering Relationship of Earthquakes at the Pacific Transform Faults
We utilize hydroacoustic data collected between May 1996 and November 2001 by NOAA/PMEL's Equatorial Pacific autonomous hydrophone array to investigate seismicity patterns at six transform faults near the equatorial Pacific Ocean: the Clipperton, Siqueiros, Quebrada, Discovery, Gofar, and Yaquina. The calibrated seismicity rate of a transform fault, when normalized by the transform fault length, changes from one transform to the other. The Clipperton transform fault, the only transform in our study with a single fault segment, has undergone tectonic compression during the last 0.5 Ma and has the lowest hydroacoustically recorded seismicity rate of the six transforms. Based on earthquakes recorded in the Harvard CMT catalog, however, Clipperton has had the highest amount of cumulative energy released of the 6 transforms from 1976 to 2004. The last Mw=6.6 occurred in May 1995, immediately preceding the deployment of the NOAA hydrophone array. One possibility for the current low seismicity rate is that Clipperton transform is currently in a seismic quiet period. In contrast, the Siqueiros, Quebrada, Discovery, and Gofar transforms all consist of multiple transform segments separated by intra-transform spreading centers. These transforms have undergone tectonic extension during the last 0.5 Ma and they each exhibit relatively high seismicity rates during the hydrophone monitoring period. Siqueiros, Discovery and Gofar show moderate seismic moment release for the last 28 years. On the other hand, Yaquina transform, thought to be a hybrid transform/non-transform offset (Lonsdale, 1989), has a low hydroacoustically recorded seismcity rate and a low moment release for the last 28 years. Through correlating hydrophone data to teleseismically recorded events, we have located several moderate size earthquakes on the Siqueiros, Discovery, and Gofar transform faults. Eight of the nineteen moderate size events ($>$ 5.3 Mw) recorded in the Harvard CMT catalog between 1996-2001 occurred closely in space and time, suggesting the possibility of earthquake interaction. Stress calculations were carried out for several pairs of the moderate size events. The relatively close correlation of the calculated Coulomb stress changes with the observed spatial and temporal variations in the main events and with the aftershock activities provides strong evidence for both static and dynamic stress interaction of earthquakes along the transform faults. Furthermore, aftershock sequences associated with the moderate size earthquakes are relatively short, lasting less than 2 days and have fewer than 120 events, thus showing considerable differences from typical aftershock sequences of continental earthquakes, which last significantly longer and have several times more aftershock events.
T41A-1165 0800h
Magmatism in the Siqueiros Transform: Major and Trace Element Evidence for Mixing and Multiple Sources.
The Siqueiros transform is a fast slipping transform fault (63 km Ma-1 half rate) that offsets the Northern East Pacific Rise (NEPR) to the west by 138 km. Alvin submersible dive observations made in 1992 and Sea MARC II sonar data collected in 1987 identified three intra-transform spreading centers that exhibit organized spreading and recent volcanism. Deep portions (3000-4000 m) of some of the active strike-slip offsets within the transform domain were found to contain evidence of recent volcanic eruptions in the form of small constructs of pillow basalts. The intra-transform spreading centers (Fornari et al., 1987) are believed to result from plate motion changes over the last few million years that has caused extension within the transform (Pockalny et al., 1997). The spreading centers may have begun as leaky transforms evolved into organized spreading centers with continued extension. Compared to the adjacent EPR, Siqueiros transform basalts are typically more primitive and porphyritic. The transform domain contains a diverse group of MORB including enriched (E), normal (N), and depleted (D) varieties. Samples from similar morphotectonic locations within the transform domain (e.g. transform splays versus spreading centers) generally have similar geochemical characteristics. D-MORB from the strike-slip faults, especially the A-B fault, differ from the spreading center basalts in that they are more primitive and depleted in incompatible elements. Major and trace element models indicate that the majority of the N-MORB samples from the spreading centers can be explained by low-pressure fractional crystallization of parents similar to primitive samples found within the deep transform offsets within Siqueiros. However, the compositional variability among the spreading centers requires more than one parental composition to explain the entire transform suite. Mixing of magmas is evident in major element variations, disequlibrium phenocryst compositions, over enrichments in some incompatible elements, and the range in REE patterns. Much of the major and trace element variability can be explained by mixing of relatively primitive melt bodies with highly fractionated magmas. Such mixing is believed to occur in melt lenses or sills where highly evolved interstitial melt exists. Significant differences in REE patterns indicate that the D-MORB and N-MORB cannot be produced from similar sources. REE abundances typical of N-MORBs can be produced by mixing D-MORB with approximately 4-6% of E-MORB. Variability in Na8.0 and Fe8.0, coupled with trace element and isotopic data, suggest mixing of melts from variable depths and extents of melting and from sources with different compositions. The more fractionated nature of samples from the intra-transform spreading centers and the eruption of only N-MORB at those sites suggest that more crystal fractionation and mixing of source compositions occur beneath the spreading centers. The more primitive and variable REE patterns of basalts from the strike-slip faults suggest that magmas erupted there without extensive low-pressure fractionation. The overall distribution of samples suggests that melt lenses capable of mixing different parental compositions exist beneath the more well developed spreading centers.
T41A-1166 0800h
Did the Cocos-Nazca Spreading Center Form at a Transform Fault During Farallon Plate Break-Up?
The break-up of the Farallon plate was the most important event during the Early Miocene plate tectonic reorganization of the East Pacific. The opening of a new oceanic spreading center perpendicular to the existing Pacific spreading is unique and probably had far-reaching consequences for the active continental margins of Central- and South America. Most of the original fissure where the Farallon plate split into the Cocos plate and the Nazca plate has already been subducted beneath Central- and South America together with much of the oceanic crust that was formed during the early phase of Cocos-Nazca spreading. This makes a full reconstruction of plate kinematics in the area back to the time of the opening poorly constrained and leaves many questions open about the mechanisms involved and the subduction zones affected by this event. The remains of the fissure in the Cocos- and Nazca plates have been identified offshore Costa Rica and Peru, resp., where they have been investigated with geophysical methods in recent years. The 600 km long still existing segment of the fissure in the Nazca plate has recently been almost completely mapped with swath bathymetry. If the origin of the Farallon break-up was a transform fault then this structure must also be present on the Pacific plate in the area conjugate to the Farallon remains offshore Costa Rica and Ecuador. However, very few ship data were available from this area in the Central Pacific so far and at the scale of satellite altimetry no such structures were detectable. During a R/V Sonne cruise the area in question was investigated with magnetics, gravity, and swath bathymetry to fill a major gap in the framework of East Pacific plate tectonic development and to test the transform fault hypothesis. The magnetic measurements were carried out with a gradiometer and a three-component vectormagnetometer to identify the almost north-south striking seafloor spreading anomalies at low geomagnetic latitude.
T41A-1167 0800h
An Earthquake Swarm on the Galapagos Transform Fault: Implications for Earthquake Triggering
Transform faults on the East Pacific Rise spreading system often have large amounts of seismicity in short periods of time. The magnitude vs. time distribution of these sequences ranges from what would be classified as similar to a typical continental earthquake but with an elevated foreshock to aftershock ratio to sequences that would traditionally be classified as an "earthquake swarm" closer to those seen in volcanic regions (e.g. Forsyth et al., 2003). Analysis of declustered earthquake catalogs (i.e. ignoring the swarms) suggest that the anomalous foreshock to aftershock ratio on EPR transforms requires different triggering processes in the oceanic regime from those that explain continental seismicity. Here we investigate the best recorded earthquake swarm on an EPR transform to evaluate whether it has similar implications. Due to the lack of oceanic earthquake catalogs with low detection thresholds and reliable magnitude estimates, the unique characteristics of transform sequences have been difficult to quantify. This study uses the earthquake catalog derived from NOAA's hydroacoustic array in the equatorial Pacific Ocean and data from a land-based seismometer array on the Galapagos Islands (from Toomey et al.) to examine the temporal and size distribution of an earthquake swarm which occurred on the Galapagos Transform in 2000. Seismic moment estimates are determined with a high degree of accuracy using an Empirical Green's Function based, cross-correlation method. Characteristics of the Galapagos swarm can be compared to that normally observed in continental strike-slip faults by applying the Epidemic Type Aftershock Sequence (ETAS) Model [e.g. Helmstetter and Sorenette, 2002], which is a common earthquake triggering model used to explain continental seismicity. This model assumes aftershocks are triggered from all large earthquakes with a triggering rate that decays in time following the Omori Law and increases with triggering event magnitude. The utility of the ETAS model to characterize the Galapagos Transform swarm is examined using both common continental time decay and triggering exponent parameters and parameters appropriate for the abundant foreshocks and few aftershocks often observed in oceanic settings. ETAS can explain the Galapagos swarm only with relatively low values of the triggering exponent, $\alpha \sim 0.3$ and relatively high values of the Omori exponent, p$\sim$1.8. We will discuss the feasibility of simultaneously explaining both the swarm and ordinary EPR events with the ETAS model as well as explore explanations for the low value of $\alpha$ in terms of the Coulomb stress model.
T41A-1168 0800h
Initiation of Ridges and Transform Faults
No clear consensus has emerged to explain initiation of the strikingly regular pattern of ocean ridges and transform faults. The question is important on the continents also, because a less regular pattern of step-overs on faults such as the San Andreas influences the sources of earthquakes. We explore the question by finite element modeling and a study of observational data on ridges and transforms. We focus on the simplest case, where ridges and transforms seem to self-organize at new plate boundaries as soon as new oceanic (magmatic) crust forms. The South Atlantic supplies a clear example. Continental South America and Africa separated along an irregular break, whose general shape is still preserved in the Mid-Atlantic Ridge. In detail, however, the sea floor magnetic anomalies and satellite gravity show that traces of the ridges and transforms extend to the base of the continental slope, i.e. they formed quickly in the new oceanic crust. The Gulf of California provides another clear example and is notable because of its northward transition into the continental San Andreas fault system. In continental crust, dike segments connected by transform faults provide the clearest analogues of oceanic ridges and transforms. Remarkably, the ridge-transform pattern has been simulated by pulling the crust on molten wax [Oldenburg and Brune, {\it JGR}, {\bf 80}, 1975] and also observed in the crust of a molten lava lake [Duffield, {\it JGR}, {\bf 77}, 1972]. In neither of these models, however, do the spatial and temporal scales permit investigation of the dikes whose repeated emplacement and inflation builds layer 3 of the ocean crust. It is well established that, under a buoyant head of magma, dikes tend to fracture and intrude the crust in planes perpendicular to the least horizontal stress, and they relieve the stress difference as they inflate [e.g. Parsons and Thompson, {\it Science}, {\bf 253}, 1991]. Dikes are commonly used as stress-direction indicators analogous to artificial hydraulic fractures. In a simple 2-D finite-element modeling setup we simulate the initiation of rifting by the dike intrusion and opening of several mis-aligned cracks in a pre-extended elastic basalt-like crust. The design of the model parameterization and the dimensions of the cracks, their mis-alignments and mutual distances resemble beginning rift systems observed in the Gulf of California and the Red Sea. Two major processes are assumed to control the expansion and interaction of the cracks and the subsequent development of transform faults. Tectonic extension and dike inflation widen the cracks and stress concentration causes the tips to propagate. Intruding magma aids the opening by exerting stresses on the crack boundaries. Then stress changes induced by the interacting cracks cause the region between the cracks to break. In our modeling the Coulomb failure criterion controls the development of faults around the cracks. Under plane stress assumptions we study the evolution of the stress regime with time, while varying the spreading rate of the lithosphere, the melt pressure of the intruding magma, the degree of mis-alignment and the distance between the cracks. We propose a process of dike intrusion to explain the orientation of ridges; mis-alignment of dikes propagating from different magma supply centers leads to formation of transforms. The hypothesis is supported by the discovery of magma-poor, ultra-slow spreading ridges that are spreading obliquely and generally lack transforms [Dick, Lin and Schouten, {\it Nature}, {\bf 426}, 2003].
T41A-1169 0800h
Relative Plate Motion Changes Observed in an Analog Freezing-Wax Model
We investigate the tectonic processes related to changes in seafloor spreading velocity using an Oldenburg and Brune (1972; Nature) freezing-wax analog model, which simulates plate tectonic processes at a shorter time scale (minutes instead of millions of years). Using videotape, we run several experiments at 5, 10, 20, 30 and 40 degrees of relative plate motion direction change. Observing the resulting seafloor spreading geometry after a designed change in relative motion, provides insight for understanding the evolution of corresponding examples in the ocean basins, such as the fracture zone bends near magnetic anomaly 34 along the Kane fracture zone in the Atlantic, and near anomaly 20 along the Mendocino fracture zone in the Pacific and the La Boussole fracture zone in the Indian Ocean. In addition, we document the origin, evolution, and eventual extinction of wax microplates forming within the center of a transform fault strand, much in the same manner as proposed by Bird and Naar (1994; Geology) for the presently active Easter and Juan Fernandez microplates along the southern East Pacific Rise. Although the wax analog model has been shown to have geometric and kinematic scaling properties, it has been difficult to show dynamic scaling. Recent work at Cornell by Bowdenschatz and Katz (URL below) with a different kind of wax and analog model has succeeded replicating the ridge and rift morphology associated with fast and slow seafloor spreading. These recent advances suggest that the model is a viable tool in obtaining insight of the rigid and non-rigid processes that occur during reorganizations of divergent and transform plate boundaries.
http://milou.msc.cornell.edu/waxtectonics.html
T41A-1170 0800h
Atlantis Bank as a Key to Understanding the Nature of the Moho and Crust-Mantle Boundary
Atlantis Bank on the Southwest Indian Ridge is a key site for the study of the oceanic lower crust and mantle. It exposes an eroded oceanic core complex lying along the western flank of the Atlantis II transform. The complex exposes of huge 400+ km2 gabbroic massif that is at least 1500-m thick at ODP Site 735. Mantle peridotites are exposed on the lower slopes of the transform wall that flanks the core complex, where they locally appear to underlie the gabbro body. During the SHINKAI 6500 cruises in 1998 and 2002, seafloor gravimetry was carried out at 16 stations on the Bank to detect its precise density structure. The average density of the basement rocks on the western slope of the bank was estimated as 3.0 g/cc (lower part), 3.8 g/cc (middle part) and 3.1 g/cc (upper part), corresponding to mantle peridotite, oxide Fe-Ti gabbro and peridotite/oxide olivine gabbro collected at these sites respectively. Average density of basement rock on the eastern slope was estimated as 3.1 g/cc (lower part) and 2.6 g/cc (middle part). The southwestern slope of the bank is characterised by exposure of unaltered layered gabbro underlain by serpentinized peridotite according to observations and sampling during the 2002 survey. Average bedrock density at this site is 2.5g/cc, extremely low considering the exposure of lower crust and upper mantle rocks there. According to the seismic survey crossing the bank (Muller, et al., 1997, 2000), a discontinuity at 4 km below seafloor at the foot of the bank and at 6 km bsf at the summit was identified. P wave velocity contrast indicates that the boundary is equivalent to Moho. Their result also shows that the southwestern slope corresponds to the Layer-2 (Layer-1 is missing), and the estimated P wave velocity of 3.6 km/s is in good agreement with the basement rock density derived from seafloor gravimetry. The Moho is not necessarily identical to the petrological crust-mantle boundary and it is possible that it is the boundary between serpentinized and non-serpentinized peridotite as shown in 'Hess model', and as also suggested by Muller, et al. (1997). This hypothesis will be tested in the future IODP riser drilling.
T41A-1171 0800h
Ocean Floor Orogenesis: Polyphase Metamorphism, Mylonitisation And Deformation Mechanisms Of Metabasites Along The Romanche Transform
We studied mylonites samples from several dredge-haul profiles accross the Romanche Transform Fault (RTF),in the Equatorial Atlantic. From 7000 to 2500 m water-depths mylonites and ultramylonites grade from high amphibolite (7000 - 6000m), to upper greenschist (6000 - 4000 m), to lower greenschist facies conditions. Deformation temperature during the main mylonitic and synschistose isoclinal folding event systematicaly increases with crustal depth. At each crustal depth mylonitisation degree (grain size reduction down < 1 micrometre) is correlated with increase of metamorphic temperatures determined by hornblend-plagioclase thermometry. The deformation mechanisms range from fracturing and dislocation creep of amphiboles and plagioclase accompanied with chemically induced grain boundary migration in protomylonites and mylonites. Randomizing of CPO and mixing of mineral phases indicate that the grain size sensitive flow operates on all mineral phases in ultrafinegrained ultramylonites of any grade. Mineral compositional zonations and presence of microveins in all crustal depths are consistent with channelised fluid infiltration mechanisms. Heat source for this prograde metamorphic event is ascribed to dissection of gabbroid magma chamber by the RTF. The main metamorphic event was followed by lower greenschist facies metamorphism at all crustal levels. This second event is associated with chevron type refolding of main fabric and dissolution creep deformation mechanism during transpressional regime generating the uplift of the RTF North wall. Lastly, low tempearture metamorphic minerals precipitated in open fractures marking the end of the uplift. The regional extent of prograde metamorphism and polyphase character of deformation and metamorphic recrystallization are consistent with definition of orogensis in the continental lithosphere. Therefore, we propose that the RTF represent an oceanic equivalent of such an orogenesis. The major difference with continental mylonitisation is the abondance of seawater-derived fluids which control all of the metamorphic crystallizations and deformations of the RFZ mylonites.
T41A-1172 0800h
Tectonic Details of the Tjornes Fracture Zone, an Onshore-Offshore Ridge-Transform in N-Iceland.
The Tjornes Fracture Zone (TFZ) links the northern rift zone (NVZ) in Iceland with the Kolbeinsey Ridge north of Iceland. The TFZ was initiated during the Miocene (about 7 Ma), following an eastward jump of the spreading axis in northern Iceland. A roughly 150 km long (EW) and 50 km wide (NS) deformation zone has since developed incorporating both right-lateral movement along WNW-trending strike-slip faults and oblique extension (105°) within three major N-S trending grabens (from west to east the Eyjafjardar ll, Skj ifandi and Oxarfjordur basins). Recently collected EM300 and RESON8101 multibeam bathymetric data, and CHIRP subbottom data combined with onshore mapping have enhanced our understanding of the rift-transform interactions within the TFZ. The transform motion is incorporated within two seismically active WNW trending zones, the Grimsey Seismic Zone (GSZ) and the Husavik-Flatey fault (HFF), spaced ~40 km apart along the margins of the extensional basins. Being the propagating continuation of the NVZ offshore the GSZ has both the characteristics of an oblique rift zone and a transform whereas the HFF is more akin to oceanic transform faults. Four left-stepping, en-echelon, NS-striking rift segments (volcanic systems) exist along the GSZ. Large GSZ earthquakes, however, seem to be mainly associated with lateral strike-slip faulting along WNW-striking fault planes. Fissure swarms transecting the offshore volcanic systems also indicate right-lateral strike-slip motion parallel to the spreading direction. The HFF has an overall strike of N65°W and can be traced continuously onshore and offshore along its 75-80 km length, between the NVZ, across Skj ifandi and into Eyjafjardar ll. Four pull-apart basins occur along the fault, the largest at the intersection with Eyjafjardar ll, the southward but magma-starved, continuation of the KR. Tertiary dikes, parallel to the HFF indicate it has been a leaky transtensional feature. The southwestern margin of the fault is characterized by NE-striking lavas which dip steeply (30-50°) towards Eyjafjardar ll. The lavas are dissected by en echelon arrays of conjugate strike-slip faults intersecting the HFF fault at angles of N20°-30°W and N20°E. Some can be traced onto land where they exhibit complicated flower patterns. Destructive earthquakes occurred on the HFF in 1755, 1867 and 1872. The 1867 events were most likely associated with rift-transform interaction within the M n reyjar volcanic system, similar to the 1975-1989 Krafla rifting episode, when a lateral intrusion event triggered a M6.5 strike-slip earthquake at the junction of the Krafla fissure swarm and the GSZ. Although transform motion within the TFZ is currently taken up by two parallel fault systems the Tjornes microplate will merge with the North American plate as continued northward propagation of the divergent plate boundary gradually deactivates the extensional basins and HFF.