T43A-1076
A Passive Seismic Array Across the Dead Sea Basin (DSB)
Beginning in September 2006, a temporary network of 30 broadband and 45 short-period seismic stations has been set up on both sides of the Dead Sea Basin (DSB). The aperture of the network is 70 km E-W and approximately 70 km N-S. It is anticipated that the network be kept in operation till March/April 2008. Data are continuously recorded at 100 Hz and 200 Hz sample frequency for the broadband stations and short-period seismic stations, respectively. The aims of the project are (1) to investigate the crust and upper mantle discontinuities with the receiver function method, (2) to investigate anisotropy in the crust and upper mantle from shear wave splitting, and (3) to study local seismicity in the area. First preliminary receiver function analysis reveals a crustal thickness of about 30-35 km in the eastern part of the DSB and possibly an upper mantle low- velocity layer. It also shows a basin which is mainly filled with salt about 10 km thick beneath the Lisan peninsula.
T43A-1077
Seismic Wide-Angle Reflection / Refraction Profiling from the DESIRE Project Reveals the Deep Structure Across the Southern Dead Sea Basin
As part of the DESIRE project a 240 km long seismic wide-angle reflection / refraction (WRR) profile was completed in spring 2006 across the Dead Sea Transform (DST) in the region of the southern Dead Sea basin. The DST with a total of about 105 km multi-stage left-lateral shear since about 18 Ma ago, accommodates the movement between the Arabian and African plates. It connects the spreading centre in the Red Sea with the Taurus collision zone in Turkey over a length of about 1100 km. With a sedimentary infill of about 10 km in places, the southern Dead Sea basin is the largest pull-apart basin along the DST and one of the largest pull-apart basins on Earth. The WRR measurements comprised 11 shots recorded by 200 three-component and 400 one- component instruments spaced 300 m to 1.2 km apart along the whole length of the E-W trending profile. Models of the P-wave velocity structure derived from the WRR data show that the sedimentary infill associated with the formation of the southern Dead Sea basin is about 8.5 km thick beneath the profile. With around an additional 2 km of older sediments, the depth to the seismic basement beneath the southern Dead Sea basin is about 11 km below sea level beneath the profile. In contrast, the interfaces below about 20 km depth, including the top of the lower crust and the Moho, show less than 3 km variation in depth beneath the profile as it crosses the southern Dead Sea basin. Thus the Dead Sea pull-apart basin is essentially an upper crustal feature with N-S upper crustal extension associated with the left-lateral motion along the DST. The boundary between the upper and lower crust at about 20 km depth must act as a decoupling zone. Thermo-mechanical modelling of the Dead Sea basin supports such a scenario. http://www.gfz-potsdam.de/pb2/pb22/projects/desire/desire-home.html
T43A-1078
Continuous Seismic Monitoring by Permanent and Portable Seismic Networks of the Dead Sea Basin in the Last 25 years
Seismic monitoring in the Eastern Mediterranean started at the end of the 19th century. Since then several strong events were instrumentally recorded in the Dead Sea rift in general and in the Dead Sea basin (DSB) in particular. The events in the DSB were mostly harmless except the fatal earthquake of July 11, 1927. The recent main shock in the northern Dead Sea basin earthquake, MW 5.1, on February 11, 2004, just reminded us of the potential of the occurrence of strong earthquakes in the DSB. The mainshock was followed by a relatively short sequence of aftershocks, which lasted for about a few months, and most of the aftershocks occurred in the first 6 weeks having small magnitudes. Two permanent seismic networks (JSO + ISN) and 3 temporary seismic networks (PSN + DESERT2000 + DESIRE) have monitored the seismicity in the DSB in the last 25 years. Small and moderate earthquakes in the Dead Sea basin, part of the Dead Sea rift, were recorded and analyzed. The earthquakes, of 0.5≤Md≤5.2, follow a normal rate of seismicity for the region with b values around 0.9. The recorded seismicity is mainly confined to the basin and its boundaries. The activity is along the north- south left lateral faults on the east and west boundaries of the basin, however, there is some clear activity along the transverse faults crossing the basin. In the Dead Sea basin we found a tendency to clustering, which is clearly demonstrated in very similar seismograms of several events recorded at the same station. Focal plane solutions of the strong earthquakes suggest a strike slip faulting. However, the focal mechanisms of the small events and aftershocks present a mixture of strike slip motion, normal and reverse faulting. Scaling relationships of moment- magnitude and other source parameters (stress drop, corner frequency) reveal similar results to those of California and other regions.
T43A-1079
A Three-Dimensional Seismic Model of the Dead Sea Plate Boundary From Active Source Data
The Dead Sea fault system is a north-south striking left-lateral shear zone separating the African and Arabian tectonic plates. The southern part of the plate boundary is located within the Dead Sea valley. The valley, much of it below sea level, is surrounded by highlands on both sides, and contains subsurface sedimentary basins, including the large (~150 km long) a deep (6-8 km) Dead Sea basin. A wide-angle seismic reflection and refraction experiment was carried out in the Dead Sea Region in October 2004 to study the deep structure of the plate boundary. The experiment consisted of two perpendicular profiles a 280-km long profile along the valley and the international border between Jordan, Israel and the Palestinian Territories, and a 250 km long profile from Gaza strip to eastern Jordan across the Dead Sea basin. Modeling of the West-East line shows a low velocity zone extending to a depth of 18 km below the basin, which includes >6 km of "syn-rift" sediments (ten Brink et al., GRL, 2006). The lower crust and Moho are not perturbed. The uplift surrounding the Dead Sea Transform also appears to be an upper crustal phenomenon. The shear deformation, associated with the transform plate boundary motion appears, on the other hand, to cut throughout the entire crust (Ibid.). Two-dimensional modeling of the South-North line is more complex due to the fact that sedimentary basins do not occupy the entire width of the valley hence some sources and some receivers are located within the basins whereas others are located outside. This heterogeneous near-surface structure explains why a simple 2-D velocity model does not fit the observed travel times from all shots. Therefore, we are using 3-D travel-time tomography to model the heterogeneous near-surface and deeper structure of the Dead Sea. Preliminary models indicate that some ray-paths from sources near the basin use the edges of the basin as a wave-guide and generate earlier than expected arrivals at receivers near the shot. We find seismic confirmation of sub-basins along the Dead Sea Transform that have been identified using gravity and aeromagnetic surveying (ten Brink, et al., G-cubed, 2007, ten Brink, et al., Geology, 1999), and will present a preliminary model for variations in the crustal structure.
T43A-1080
The Diurnal, Multiday and Seasonal Evolution of Rn-222 Within Rocks Along the Dead Sea Transform in Relevance to Earthquakes Related Phenomena.
The basic assumption underlying monitoring of Radon along the Dead Sea Transform is the capability of Radon to emanate to the porous media and be available for further transport by diffusion via the pore content (geo-gas) or to a convective migration, to a limited range (half life = 3.8 days). To this end, the association between the temporal variations of Radon at daily, multi day and seasonal scale is investigated versus other geophysical variables: pressure, temperature, precipitation, humidity, groundwater level, gravity, and geodynamic changes. Monitoring Radon in the desert-like region enables to set apart those variables that are not active in the area, like precipitation and keeping part of them stable as by installing the Radon sensors at the Amram site near Eilat. It is a deep dead-end, confined horizontal tunnel (170 meter span and 75-200 meter below surface) with well defined environmental conditions as: stable indoor humidity and temperature (28±0.3°C), low CO2 content and very dry porous media, located high above the ground water table. Monitoring Radon by total gamma and alpha counting together with environmental parameters, since 2002, at sampling rate of several minutes, provided the following results: No correlation occurs between Radon and atmospheric pressure variations at the Amram site. Since the subsurface and external pressure is the same it does not create a local pressure gradient to produce an advection that can account for the radon signals (anomalies). In addition, there is no connection between these temporal variations of Radon and the stable subsurface temperature (28±0.3°C). Nevertheless, an apparent correlation in the time domain occurs between the Radon variation at depth and the temporal variation of the external temperature. The Radon follows sequentially the daily temperature variation with a delay of about 10 hours, consistently. It indicates that the diurnal variation of the Radon concentration at Amram tunnel is caused by the interplay of the ambient temperature gradient at the outside rock/air interface. The delay is a result of the Radon migration time towards the tunnel and back, within the fracture rock porous media. An additional external detector (at 7 meter borehole), present a clear daily correlation between the Radon and the temporal variation of the outdoor temperature, but with a delay of only 2-4 hours. Furthermore, the multi-day and the seasonal variation of the radon emanation also delineate the same interconnection with the ambient temperature. The non-linear relation between the intensity of the radon anomalies and the ambience temperature may explain the reduction of the intensities by factor of 2-5 during the winters (November-March). Similar relations occur at the shallow sites of NW Dead Sea shore, between part of the radon anomalies and the temporal variation of the environmental temperature. The rest of the radon signals could be related to other geophysical processes such as earthquakes (Geology, v. 31, 505-508). Similar phenomena have been observed in worldwide corporate research in Canada and India.
T43A-1081
Evaluation of geodynamic activity of the Dead Sea transform Fault by radon gas concentrations
Twelve radon lines of dosimeters (detectors) were placed across the Jordan Valley active fault, which is a segment of the active Dead Sea transform fault system. Each line of the dosimeters shows one or more peaks of radon anomaly concentrations. Some of these peaks prove the intersection of the fault trace with these lines in areas where the fault plane is inferred. In other lines, the peaks correspond to the arrangement of faults in areas of pull-apart basins (sag ponds) or pressure ridges, formed due to the left or right step of the fault. Sag ponds usually show low radon emanations, because they are the place for the accumulation of very fine sediments, which decreases their porosity and hence the upward migration of the radon gas. The northern part of the Jordan valley relatively shows high radon emnation, which could be attributed to the presence of a seismic gap in the upper Jordan valley
T43A-1082
An oxygen isotope excursion in planktonic foraminifera from the Red Sea as an indicator of coseismic hydrothermal activity
Analysis of the active tectonics of the Dead Sea Transform is hampered by a lack of paleoseismic data from the Red Sea. Recent field observations in Saudi Arabia accord with the palaeoshore record of other parts of the central and southern Red Sea to the effect that the coasts have undergone negligible uplift during the last 120 ka, a finding which is consistent with strike-slip as the dominant style of axial deformation. As this is a marginal sea with limited exchange with the open ocean, tectonic episodes may be recorded by the products of enhanced hydrothermal inflow. Accordingly, geochemical analysis was undertaken of biogenic-planktonic carbonates from cores taken from the sea floor. Tests of Globigerinoides ruber from Core CH10003-3 PC taken at a water depth of 1374 m in the central rift of the Red Sea revealed a major positive (>2.5‰) δ 18O excursion older than 40,000 yr. The δ 18O excursion is not associated with hardgrounds, fluctuations in δ 13C or any other indicator of a climatic or hydrological phase, and is therefore provisionally ascribed to a major coseismic event during which there was an intense exchange of sea water and fresh basaltic rocks, or the sudden release of fluids that had interacted with basaltic rocks for a long time, or both. The lack of any corresponding signal in 87Sr/86Sr ratios could result from the relatively low concentration of Sr in the hydrothermal fluid.
T43A-1083
Seismic history of the Southeastern Sea of Galilee margins, Dead Sea Rift, Israel
We studied the seismic-history of the SE margins of the Sea of Galilee, situated along the Dead Sea Rift. The study area hosts the N tip of the seismically-active Jordan valley fault-segment. During the past two millennia a few strong earthquakes are known to have occurred along this segment and current seismicity is recorded. Morphotectonic mapping using aerial-photos and field observations revealed in the studied area numerous 10 to 100 m long N-S striking lineaments as well as some old landslides. Resolving the interaction between these morphological elements can be a key factor in understanding the level of local seismic hazard and the relative timing of the events. We excavated trenches across 4 lineaments showing distinct morphological vertical steps. Each trench hosts one or more westwards dipping faults with normal displacement of the soil profiles. Total displacement is up to a meter and in places more than one displacement event can be traced on a single fault plane. The bed rock is highly fractured near the faults. OSL dating of the soil profiles indicates Holocene ages in the westernmost trenches and upper Pleistocene in the easternmost ones. The largest landslide observed in the field, 500m wide and 1500m long, was selected for detailed analysis. A trench was excavated on the landslide, crossing a lineament clearly detected on both the N and S boundaries of the slide. The 3m deep trench exposed 3 soil profiles developed in the colluvium composing the landslide. The development of the three soil profiles implies cyclic down-slope movement of material and quiescent periods between movement episodes, allowing soil formation. Sedimentary and pedogenic disturbances detected in the trench, may be associated with the slide-crossing lineament. OSL dating of the soil profiles reveals upper Pleistocene ages in the lowermost soil profile and Holocene in the uppermost one. The temporal relationship between the slide and the lineaments are yet unclear, although the overall time frame for sliding events is similar to displacement events. Results of slope-stability analyses for the slide indicate that the slope is stable under static gravitational loading. Pseudo-static analyses results indicate that a horizontal PGA value of about 0.3g would be required for sliding if residual strength values of the rocks are assumed. According to the observed displaced soil-profiles in the trenches and the high PGA needed for slope instability we conclude that the study area experienced an intense seismic history with strong earthquakes (Mw 7) and high PGA (0.3g). These findings need to be taken into considerations when seismic hazard evaluations are done.
T43A-1084
THE APAME PROJECT: ARCHEO-PALEOSEISMOLOGY FOR THE PROTECTION OF ARCHEOLOGICAL SITES AND CULTURAL HERITAGE IN THE MIDDLE EAST
The APAME EC project (Contract ICA3-CT-2002-10024) consists of multidisciplinary studies of seismically active regions along the Dead Sea Fault. Several medieval cities, citadels, fortresses and known archeological sites of the Middle East suffered of severe earthquake damage in the past. A repetition of similar seismic events nowadays will cause tremendous damage and destruction mainly due to the last century increase in population and modern constructions. In the APAME project, we combine historical, archeo-paleoseismological and earthquake hazard studies and provide an integrated analysis of their consequent seismic hazard. The results are: 1) An inventory with maps of archeological and cultural sites affected by past earthquakes and detailed archeoseismological and paleoseismological studies of some specific sites using precise dating techniques. 2) An analysis of damage distribution based on the historical archives, new documents and field investigations and preparation of updated sesimicity catalogue of past earthquakes. 3) Investigations of man-made constructions and building practices for each site with a comparison between heavily damaged and undamaged buildings. Striking results are ~ 40 m fault offset of a Tell site in the Amik Basin (Turkey), the severe earthquake damage in the remaining "lost villages" and ~ 13.6 m offset of a Roman Aqueduct in northern Syria, the trace of the 1837 and 1759 earthquake fault ruptures in southern Lebanon, and the constraint of fault slip rate using offset streams and paleoclimatic reconstructions along the Jordan Valley Fault. A fault segmentation is obtained from the compiled instrumental and historical earthquake catalogue and related damage distribution. The rich historical seismicity catalogue and fault segmentation combined with the tectonic and geodetic strain pattern along the DSF contribute to the characterization of rupture zones with a potential for a future large earthquake. http://apamea.u-strasbg.fr/
T43A-1085
Coseismic growth of sedimentary basins along the Yammouneh strike-slip fault (Lebanon)
The left-lateral Yammouneh fault is the main active branch of the Dead Sea Fault System (DSFS) within the Lebanese restraining bend. Despite the overall transpressional setting, a series of sedimentary basins have developed along the trace of the Yammouneh fault. Consequently, paleoseismic studies within these basins provide an opportunity to study the processes of co-seismic growth of the basins, as well as elucidate earthquake behaviour of the fault, in general. The active fault trace along a 51-km length of the Yammouneh fault was delineated as relatively young surface ruptures and fault scarps that affect Holocene deposits, with intermittent offset geomorphic markers. A detailed field investigation focused on the rhombohedral Yammouneh basin, an actively evolving example of pull-apart basins that is obliquely cut by the active fault. Three-dimensional correlation and analysis of palaeoseismic investigations exposed a composite shear zone with a total subsidence that exceeds 2 m over the past approximately 7300 years. Stratigraphic and geochronological constraints suggest the occurrences of at least four large seismic events during that period, with the last event being coeval with 240-441 AD. Our results suggest a subsidence rate due to faulting of about 0.15 mm/yr within the Yammouneh pull-apart basin and a total throw across the Yammouneh fault of about 200 m across the basin. Comparing the present-day slip rate and the size of the Yammouneh basin suggests an apparently young age for of 1.0 – 1.6 Myr. A palinspastic reconstruction suggests that the pull-apart opening of the Yammouneh basin may have been followed by a possible 10 degree -clockwise rotation that could have had a major role on the basin's subsequent evolution in terms of localized extension, confined compression, change in the stress field, and fault migration. These results permit revisiting competing conceptual models for development of strike-slip pull-apart basins.
T43A-1086
Active tectonics and fault interactions in the Ghab Valley pull-apart basin (Dead Sea fault system) in northwestern Syria
Along the northern Dead Sea fault system (DSFS) in northwestern Syria, the left-lateral transform splays into two distinct faults bounding the 70 km long Ghab Valley. Plate tectonic models predict 4 – 7 mm/yr of slip along this part of the transform, although recent GPS results from Syria suggest slower rates of faulting. Historical records document the occurrence of large earthquakes (magnitudes greater than 7) within the vicinity of the Ghab Valley and suggest that both faults bounding the valley may be active, seismogenic structures. This study integrates new neotectonic observations with other available information on the basin structure, fault slip, and earthquake recurrence in an effort to address how slip is transferred and partitioned between the faults bounding the Ghab Valley. We also investigate how this structural framework may relate to seismogenic fault segmentation. Structural and topographic relief demonstrate that uplift is asymmetrically distributed about the Ghab Valley, with greater structural relief along the western margin of the valley. Along the both the western and eastern bounding faults, late Quaternary faulting truncates alluvial fans, Quaternary lava flows, beheaded drainages, and hanging valleys. For example, in the northeastern Ghab Valley, a displaced lava flows suggests a minimum late Quaternary slip rate of approximately 1 mm/yr. Also along the eastern fault, recent faulting, including paleoseismic indicators, is also expressed in recent sediments along the fault zone. Radiocarbon dating constrains the last coseismic event within the past 1,500 years, possibly corresponding with the historically documented earthquake of 1157. These observations provide a basis for investigating fault interactions and slip partitioning using boundary element modeling. The modeling demosntrates that transfer of fault slip produces local transtension even within an overall transpressive setting. Furthermore, the obliquity of plate motion relative to the transform influences the localization of depocenters within the pull-apart basin.
T43A-1087
Tectonic Activity and Processes Preceding the Formation of the Dead Sea Fault Zone
Analysis of geological-geophysical data indicates that at the end of the Proterozoic, blocks of the Arabian Shield (AS) were thrust to the north-west onto the crust of the proto-Mediterranean (PM). This was caused by the pushing of oceanic crust from the south-east forming the Najd faults system (NF). This thrusting took place between 630 and 590 Ma, and is confirmed by the offsets between the Yanbu suture of the AS and Allaqi-Sol Hamid suture of the Nubian Shield (NS), the Bi'r Umq suture of AS and Nakasib suture of NS, and parts of the Yanbu and Nabitah sutures of AS. This caused the separation of AS from NS, and AS from the continental crust to north-east of it with its north-western displacement, resulting in opening of the Persian Gulf. This caused the start of deposition of huge amounts of Vendian-Cambrian evaporites in Saudi Arabia, Oman, Persian Gulf, Zagros, central Iran and other regions. The fact of the formation and preservation of the evaporites, and the common similarities in Vendian-Triassic sedimentary cover of Central Iran, Zagros, Taurus, and Arabian Plate (AP) and common Late Proterozoic–Early Paleozioc magmatic activity, show that these regions did not change their position significantly since then. Results of the DESERT project show that the lowermost part of the crust is present east of the Dead Sea Fault Zone (DSFZ), but it is absent west of it. This could be explained by detachment of the bottom part of the crust west of DSFZ during AP thrusting onto the crust of PM. The lithospheric slice discovered by seismic data between Moho and depth of about 55 km in S. Israel could be a remnant of that crust. During the thrusting, the AP overrode the detached slice. The slice was later remelted during formation of the postorogenic magmatic rocks of 590-530 Ma widespread in Jordan. The formation of three dyke swarms in S. Israel (600-540 Ma), widespread dykes in Sinai (590-530 Ma) and AP (590-530 Ma), as well as high-T-low-P metamorphism between 600 and 530 Ma, point to a huge release of heat likely caused by friction during the thrusting. Presence of giant quantities of K-rich granites also points to reworking of continental crust in the region. Small amount of magmatic formations younger than Cambrian age west of DSFZ and significant amount of magmatic formations of this period east of DSFZ also indicate to presence of the plate beneath Israel. Offset of suture zones within AS shows that displacement was maximal for the northern blocks of AS. This is in agreement with known separation of Israel's crust into three blocks: Negev, Judea-Samaria and Galilee- Lebanon. Numerous markers of high to ultra-high pressure conditions signify to collision between the AP and the PM. These markers are: iron rich aegerine-augite and olivine-rich phenocrysts in S. Israel; peridotite xenoliths in S. Israel equilibrated at depth ~33-34 km; discovery of diamonds, micro-diamonds and indicator minerals (Cr- diopside, orange garnet, pyrope, coesite, picroilmenite, moissanite, carbonado, corundum, olivine, perovskite, aegerine, Ti-augite) in S. Israel; garnet clinopyroxenites, garnet granulites, indicator minerals, and eclogite-like rocks in Mt. Carmel area of N. Israel; iron-rich garnets in Sinai; eclogites, diamonds and indicator minerals in non-kimberlite environments in Syria; ophiolites exposed in Syria at northern extension of the DSFZ; ophiolite-like rocks in Sinai. These markers are located along or in close proximity to the DSFZ. Different isostatic conditions east of the edge of- and above- the underlying plate, along with the eventual activation of the plate caused the later formation of the DSFZ.
T43A-1088
Formation of deep basins along strike-slip fault systems: The Dead Sea fault
Sedimentary basins are often formed along strike-slip fault systems. The Dead Sea fault is associated with some large and unusually deep basins. The width/depth ratio of these basins is often less than 1. In areas where the deep basins occur, two strands of the Dead Sea fault overlap in an en echelon pattern. This situation is quite rare along the Dead Sea fault; normally most basins are bordered only on one side by a strand of the Dead Sea fault, leading to their asymmetry. Geophysical data suggest that the deep basins are probably bordered on all sides by vertical faults that cut deep into basement. It has been previously proposed that in the deepest basin, at the southern Dead Sea, an isolated block of crust and lithosphere has dropped into the mantle. In this presentation we investigate the mechanism of formation of this and other deep basins along the Dead Sea fault and propose that dropping down of pre-existing heavy magmatic bodies into the mantle took place in these regions. Density heterogeneities formed in the crust or upper mantle during a previous stage of regional magmatism, drop into the upper mantle when strike-slip faults are created that detach them from the surrounding lithosphere. The suggested mechanism of deep basin formation is supported by the results of three- dimensional numerical simulations of spatial-temporal evolution of seismicity patterns and faults in a regional lithospheric model of crustal deformation and self-organization of regional earthquakes and faults. The simulations indicate that the resulting basin is rhomb-shaped and that with time it grows by the addition of distinct segments to its edges. The proposed mechanism could account for the formation and evolution of large sedimentary basins along other strike-slip fault systems, such as the San Andreas fault and other continental transform faults.
T43A-1089
Cessation of Slip on the Pilarcitos Fault and Initiation of the San Francisco Peninsula Segment of the (Modern) San Andreas Fault, California
The Pilarcitos Fault (PF) is widely regarded to be the long-term (3-19 Ma) trace of the San Andreas Fault on the San Francisco Peninsula. The active modern San Andreas Fault (MSAF) in this area, however, records less than 3 m.y. of displacement. Middle-Miocene and younger displacement is partitioned to a single San Andreas Fault where the PF and MSAF merge in the Santa Cruz Mountains south of Palo Alto and Monte Bello Ridge. Aside from such general constraints, evidence for the timing of initiation of the MSAF and for cessation of slip on the PF are poorly constrained. As a consequence, there is a lack of consensus on the long term slip rates for the PF as well as the MSAF. Two independent datasets indicate significantly different offsets along the MSAF since 3 Ma. In one dataset distinctive 3.3 Ma (Blancan) fluvial gravels of the Santa Clara Formation are offset 28-31 km from their clast source east of the MSAF near Loma Prieta, yielding a slip rate of 9-10 mm/yr since 3 Ma. In the other dataset, a linear magnetic anomaly associated with Mesozoic serpentinite and gabbro of the Coast Range Ophiolite is offset 22 km across the MSAF, yielding a slip rate since 3.3 Ma of 7 mm/yr. Both datasets assume cessation of slip on the PF by 3 Ma, but imply different amounts of offset for the MSAF. Both offset estimations also yield lower long term slip rates than the present geodetic rate of 13-19 mm/yr. A closer review of these data reveals stratigraphic relations that indicate long term slip rates of 14-21 mm/yr for the PF and MSAF, which are more in line with geodetic rates. These long-term rates result from recognition that the distinctive Santa Clara Formation strata west of the MSAF, that were deposited in a pull-apart wedge between the PF and MSAF, are also underlain by the 5.4-7.0 Ma marine Purisima Formation and by middle Miocene and older marine strata. Along the east side of the MSAF the northwestern most exposures of the Purisima Formation underlain by Miocene marine strata are 74 km southeast of this pull-apart wedge, in the vicinity of the Sargent oil field northwest of Hollister. This relation requires about 43 km of pre-MSAF slip to have been taken up by a segment of the PF that we propose bounds Miocene strata in the fault wedge between the main PF and MSAF. The slip rate implied for the PF between 5.4 and 3.3 Ma is 20-21 mm/yr; the rate implied for the combined PF + MSAF since 5.4 Ma, is 13.7 mm/yr. These stratigraphic relations require a complex displacement history of incremental slip partitioning from the PF to the MSAF from 5.4 Ma until as late as 1.0-1.6 Ma, based on the long term rates for the PF and for the PF + MSAF. The discrepancy between the apparent offset of the Santa Clara Formation across the MSAF and offset of the magnetic anomaly by the MSAF are attributable to 9 km of Santa Clara Fm displacement that was taken up after 3 Ma by the PF south of the offset magnetic anomaly.
T43A-1090
Large Alluvial Fans in the Araba Valley (Jordan) as a Record of Tectonic Deformation of the Southern Dead Sea Fault and Regional Paleoclimates
The Dead Sea fault is the 1000 km-long strike-slip fault that accommodates northward motion of Arabia relative to Sinai at a rate of about 5 mm/yr. This study focuses on the southern segment of this fault, the Wadi Araba fault. From the Dead Sea basin to the Gulf of Aqaba, the fault runs along an axial valley, about 20 km wide and 150 km long, bounded to the east by the Jordanian Plateau, reaching 1500 m in elevation and to the west by the Negev Plateau, lower in elevation (500-700 m). The Araba valley is floored with Plio-Quaternary deposits and in particular with large alluvial fans that are cut by the fault and offset relative to their feeding channel. We mapped the valley floor in details and dated some of these fans first to assess their lateral offset and further constrain the slip rate on the fault, and secondly to try to correlate alluvial fan aggradation periods in this arid/semi-arid environment to paleoclimatic variations at the regional and global scales. To identify the possible sources of the large alluvial fans, we analyzed the drainage network and the catchment basins on the valley rims based on SRTM3 topography. Previous study suggested that these fans were Pliocene and underwent offsets of 15 to 30 km. Ages as young as 50 to 350 kyr derived from 10Be exposure dating of 33 samples from the surface of some of these fans indicate that such offsets are very unlikely. Actual offsets have to be smaller. Preliminary reconstructions suggest offsets of 700-1300 m and 1600-2000 m for fans dated at 160 +/- 20 kyr and 330 +/- 22 kyr respectively, and offsets of 2.5-3.5 and 4.4-5.4 km for older fans that have not been dated yet. Consistent ages at different sites suggest simultaneous fan emplacement controlled by some external controlling factor such as climate variations. The correlation between these aggradation episodes and paleoclimatic variations would allow us to draw hypotheses on the age of these older surfaces. These preliminary results provide new constraints on the slip rate of the fault of 4-9 mm/yr since the Middle to Late Pleistocene.
T43A-1091
Mid-Miocene to Present Growth and Uplift of Mt-Lebanon and Tectonic Evolution of the Dead Sea Transformê's Restraining Bend.
The Lebanese Restraining Bend (LRB), principal irregularity along the left-lateral Dead Sea Transform (DST), is a ~25? clockwise inflexion of the Yammouneh Fault (Y.F.), through-going branch of the DST between Galilee and Syria. This is a region of high and broad relief, with the highest mountain range of the Levant (Mt-Lebanon, 3100m asl), whose east flank is truncated by the fault. Transpressional shortening is mostly taken up by a large, partly submarine thrust system ? the Mount Lebanon Thrust (MLT) ? that dips beneath the west flank of the range and is responsible for its growth and uplift. This flank is marked by the steeply west-dipping Lebanese Flexure, which warps a thick sequence of Mesozoic limestones, separating the uplifted, deeply incised, Mt Lebanon mega-anticline core from the flat narrow, coastal stretch. Such west-vergent, asymmetric folding is the surface expression of the crustal-scale MLT ramp, which extends for ~ 120 km from south to north. A well-developed foreland fold and thrust belt exists in northern Lebanon, and smaller-scale thrusting is observed along the southeastern flank of the range. Fieldwork helps constrain the onset and evolution of shortening and uplift. There is quantitative evidence that shortening started around 15-20My ago in the south, but not much before 5Ma in the north. Geological observations in central Lebanon suggest a Mid- to Late- Miocene age for the Flexure, with much of the mountain building postdating 13Ma. The bulk of uplift and largest amount of folding postdate the Late-Miocene (10-7Ma). At Kousba, the Flexure is of late-Miocene-Pliocene age, with much of the growth between the Messinian and Lower Pleistocene (7-2Ma). Folding, uplift and faulting along the northern foreland belt started after the end of the Miocene (5Ma), and continues today. The Qalhat, Tourbol, Bebnine-Miniara and Aabdeh ramp-anticlines thus provide a clear example of progressive stepping and migration of thrust ramps into a foreland basin. The timing of mountain building is in keeping with the inference that the Dead Sea Transform propagated northwards across the LRB, and implies that its main "Syrian" branch did not accumulate much offset prior to 5Ma.
T43A-1092
Co-seismic and paleo-seismic slip along the Kokoxili Mw 7.8 earthquake rupture
Co-seismic slip values along a strike-slip rupture are found to be very irregular with variations up to one order of magnitude. Data usually scattered and sparse, are now becoming more dense and continuous with slip functions derived from InSAR or image correlations. Whether the fast variations in slip along strike reveals long- lived structures of the fault plane at depth, only incomplete slip at the surface or inelastic accommodation of slip remains debated. In addition, how these slip disparities are accommodated with time is unclear. Here we investigate the co-seismic and paleo-seismic slip of the Kunlun Fault, along the Kokoxili 14 November 2001, Mw=7.9 earthquake rupture. At one site, a set of about 10 inset terraces is used to infer the slip history of the fault. The particularly low slip value at the site, about 2.4 ± 0.6 m, lower than the value derived from Spot image correlation or from Insar, follows a maximum slip value of about 8 m only a few kilometers to the west. Site interpretation from field data and high resolution Ikonos images allow to reconstruct the different riser offsets across the damaged rupture zone. Be10 cosmogenic exposure ages of cobbles from the surface of the terraces allow to constrain the age of one of the terraces and to infer an average slip-rate of 10-15 mm/yr in good agreement with previous results along the fault. When a characteristic slip model is considered, and depending of the measurement method or location of coseismic slip, the recurrence time for events comparable to the Mw 7.8 Kokoxili event may vary from 200 to 800 years.
T43A-1093
New insight into Bottom Morphology and active faults of the Northern Gulf of Eilat/Aqaba mapped from multi-beam sonar data
A high-resolution marine geophysical survey in the northern gulf of Eilat/Aqaba region was conducted during October and November 2006 on board the R/V "Etziona". This survey was carried out by an international research group (Israel, Jordan and USA) funded by MERC. The overall aim of the research is to provide the municipalities of Aqaba and Eilat with a base map of active faults for seismic hazard analyses and earthquake preparedness planning through the collection and interpretation of high-resolution marine geophysical imaging of the seafloor and subsurface strata. The morphology of the northern margin of the Gulf of Eilat/Aqaba is characterized by a shelf-slope structure in the north; steep slopes with almost no shelf in the east and terrace structure with a moderate slope to the west. Processing of the Kongsberg-Simrad EM 1002 multi-beam sonar data, together with slope analyses, reveals very interesting features on the seafloor (e.g. canyons, slumps, reefs, etc.) that indicate recent faulting activity in this area. Based on the multi-beam data we have classified the morphology of the northern gulf into seven domains: (i) modern reef and terrace; (ii) shelf with buried reef; (iii) submerged cliff and fan delta; (iv) Yutim and Arava submarine canyons; (v) collapsed margin; (vi) uneroded structural highs, and (vii) basin floor. These domains clearly delineate the areas within the Dead Sea transform valley that are tectonically active from more stable regions. Faults identified from seismic reflection data can be traced to near the seafloor indicating recent seismic activity. The fault pattern crosses obliquely from the eastern side to the west thus transferring strain across the zone in the offshore. This fault system is much more complex than previously assumed.
T43A-1094
Damaged Cave Deposits in Haifa: Evidence for Destructive Earthquakes from the Carmel Fault
A study conducted in a cave located in the Mount Carmel region (Denya Cave), shows that broken speleothems can provide evidence for damage related to ancient earthquakes on the Carmel fault. Following the work of Kagan et al. (2005) in the Judean Hills, in which accurate 234U/ 230Th ages of collapsed speleothems were consistent with independent evidence for strong earthquakes, we present here initial results from the Denya Cave. Mt. Carmel is a manifestation of tectonic movements in the north of Israel. It is a continental uplift of more than 800m above sea level. It is defined by the northwestern to north-northwestern trending Carmel fault, a branch of the Dead Sea transform system, and continues into the Mediterranean continental shelf (Hofstetter et al., 1989). The fault has been active since the Miocene and activity during the Pleistocene is evident, and yet very little is known about the extent of its seismic activity during the Quaternary (Gluck, 2002). The region of Mt. Carmel features abundant karst , which can be used in a comprehensive paleo-seismological study of the Carmel fault . The Denya cave is situated on the southern side of the mountain on a spur sloping to the west from the summit of Mount Carmel, toward the town of Tirat Ha-Carmel. Work in Denya Cave, up to this stage includes detailed mapping and initial collection of speleothem samples suspected of being seismites (broken or deformed speleothems that could result from seismic activity), and their dating using the 234U/ 230Th method. Measurements for dating are conducted on the Multiple Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS) in the Geological Survey of Israel. The age of one seismic event, as evident from three damaged stalagmite samples, is ca. 11ka. This age was determined for post- and pre-collapse laminae. Five ages of ca. 10ka were obtained from other seismites and it is not clear yet whether they represent the same event or a subsequent one. A cluster of older events between 14-18ka was identified in five samples and further study is needed in order to verify whether this range of ages represent clustering of several events. No indications in the cave were yet identified for seismic activity younger than the ages of 10-11ka obtained. Excluding archaeological indications for damage from Tel Megiddo (Marco et al., 2006), these ages are the youngest known for paleoseismicity on the Carmel fault to date.
T43A-1095
Episodic behavior of the Dead Sea fault from slip rate variations and integrated seismicity
The 110-km-long fault rupture of the Jordan Valley section of the Dead Sea Fault exhibits a significant late Quaternary active deformation and is capable of producing large and destructive earthquakes (Mw > 7.2). Here, we address its behavior over the last 48 kyr by means of historical seismicity, archeoseismicity, paleoseismicity and geomorphology to evidence variations in seismic activity and suggest a generally episodic behavior. First, we analyze 20 offset drainages evenly distributed along the southern section of the JVF. From the history of past lake-level fluctuations and intense rainfall episodes, we identify six climatic events likely to have triggered the onset of gully incisions in the Lisan and Damya formations and propose a well-constrained chronology using related absolute dating. Hence, measured lateral offsets from aerial photographs and field control points provide: i) a long-term average value of 4.9 mm/yr in good agreement with GPS velocities and ii) strong variations over short time spans with a threefold increase from 3.5 mm/yr to 11 mm/yr during a 2000-yr-long period. In addition, an integrated seismicity catalogue is produced from existing historical data, archeological results and original paleoseismic investigations at two trench sites. Archeological results from 10 sites, half of which located in the close vicinity of the fault, provide evidence for 7 destructive events since ~ 2900 B.C.. At Ghor Kabed, paleoseismic trenches dug across the bounding faults of a pull-apart basin show that at least two fault movements have occurred between A.D. 560 and A.D. 1800 (2sigma-calibration) and can be related with the A.D. 749 and A.D. 1033 large earthquakes (M>7) that struck the Jordan Valley. Further north, at Tell Es-Saydiyeh, paleoseismic trenches evidence up to 8 surface-rupturing events during the last 14 ka, of which the most recent may be correlated to the historical A.D. 1033 earthquake. The integrated catalogue reveals an episodic fault behavior with clusters of seismicity and quiescence periods as well as a 600- to 1000-yr-long recurrence interval for large earthquakes in the last 14 ka. Considering the last large earthquake in the Jordan Valley occurred in A.D. 1033, the fault may have accumulated 3.5 m to 5 m of slip. This pleads for a Mw 7.2+ earthquake yet to be released along the Jordan Valley fault segment.
T43A-1096
Synchronous Late Quaternary slip rate variability on two strands of the San Jacinto fault, California
We present new results that show slip rates varied synchronously by a factor of two over the past 35 kyr along two parallel strands of the San Jacinto fault. Our results combine high-resolution LiDAR digital topography, field mapping and 10Be exposure-age dating from two of the most active strands of the southern San Jacinto fault: the Clark fault (CLF) and Coyote Creek fault (CCF). These faults form numerous NW-striking scarps that offset three generations of Quaternary alluvial fan surfaces, Q2b, Q3a, and Q3b. We dated alluvial fans along both the CCF and CLF using 10Be sampling methods adapted for available material and degradation of the surface. For younger surfaces with well-preserved bar and swale morphology, we used a new sampling method where 12 to 20 chips from quartz-bearing boulders lodged within a bar were amalgamated into a single sample. For older surfaces we either sampled individual meter-sized boulders or collected samples from a 2 m-deep depth profile. Surface ages are consistent between CCF and CLF sites: 40 ± 12 ka and 31 ± 6 ka for Q2b, 7.1 ± 1.6 ka and 4.6 ± 1.6 ka for Q3a, respectively. Samples from Q3b near the CLF yielded ages of 1.0 ± 0.2 ka and 2.1 ± 0.3 ka. CCF samples have not yet been corrected 10Be inheritance, thus we use the CLF ages to calculate preliminary slip rates. Late Pleistocene to present rates are CLF: 2.2 ± 0.5 mm/yr, CCF: 3.4 ± 0.9 mm/yr, and 5.6 ± 1.4 mm/yr combined. Mid-Holocene to present rates are CLF: 4.1 ± 1.5 mm/yr, CCF: 6.7 ± 2.8 mm/yr, and 10.9 ± 4.3 mm/yr combined. Latest Holocene CLF slip rate exceeds 3 mm/yr. The combined Late Pleistocene to present slip rate for the southern San Jacinto fault is less than one third the rate deduced from the onset of faulting ca. 1 Ma. Mid-Holocene to present slip rates for both the CLF and CCF are about double their ca. 35 kyr rates, but are less than the 16 - 20 mm/yr geodetic loading rates and the >16 mm/yr slip rate since 1 ka at Hog Lake. We conclude that (1) overall slip rate of the San Jacinto fault has probably decreased significantly since fault inception, (2) slip rates vary over 1 to 5 kyr time scales and may be presently elevated due to a cluster of activity, and (3) other structures such as normal faults and folds could be accommodating up to one half of the deformation across the fault zone.
T43A-1097
Paleoseismic and Holocene slip rate investigations along the San Andreas Fault, at Parkfield, California
Prior to the 2004 Parkfield M6 earthquake, we excavated two paleoseismic trenches across the main San Andreas Fault (SAF) ~1 km south of Parkfield, CA. These excavations showed evidence of deformation from both aseismic creep and ground rupturing earthquakes such as the 2004 event. Despite this effort, it remains unclear if the Parkfield segment of the SAF experiences ground rupture from earthquakes >M6. This is an important question for understanding earthquake hazard in central and southern California, especially considering the central California foreshocks that were felt just prior to the 1857 Fort Tejón M 7.9 earthquake. Additionally, numerous slip budget calculations predict a slip deficit of ~5 m extending into the Parkfield segment and suggest that coseismic slip along the Parkfield segment could propagate further southeast. However, these slip budgets do not include a locally-determined geologic slip rate for the Parkfield segment. It is plausible that the slip rate at Parkfield is lower than along the Carrizo segment to the southeast. A lower slip rate would imply a lower seismic hazard from the SAF and indicate that slip is distributed along adjacent structures at Parkfield. To more thoroughly understand the Parkfield segment of the SAF, the Southwest Fracture Zone (SWFZ; which also ruptured in 2004) must be considered in both paleoseismic and slip rate investigations. One site along the SWFZ, Miller's Vineyard site, is located ~1 km west of Carr Hill along the Ranchita Canyon Rd. Here the SWFZ strikes ~315 degrees (parallel to the SAF) and ruptured in the 2004 earthquake. The Miller's Vineyard site is located along an ephemeral tributary of the Little Cholame Creek that appears to be offset right-laterally. Aerial photography suggests that the SWFZ is delineated by a 30 m wide zone of enhanced vegetation that extends more than 100 m. We expect that excavating this site will reveal deformed Holocene layers of fluvial sand, silt, and channel gravels that may be suitable for geologic slip rate investigation. Using the B4 LiDAR data we have identified two other paleoseismic sites of interest along the SWFZ and six sites along the main SAF. Improving the length of earthquake recurrence data and establishing a Holocene slip rate for Parkfield are essential for studies of the temporal aspects of fault strain release histories, fault mechanics, and for characterization of earthquake hazard for south-central California. http://activetectonics.la.asu.edu/Parkfield/
T43A-1098
Evolution of an Intermontane Basin Along the Northern San Andreas System: Evidence from Basin Structure of Little Lake Valley (Willits), Northern California Inferred from Gravity and Geologic Data
Associated with the northern strands of the San Andreas fault system in California is a series of small intermontane basins. While it is tempting to ascribe their formation to simple pull-apart tectonics along the dominantly strike-slip fault strands, direct evidence for basin genesis is lacking. In this study, a detailed gravity survey throughout the Little Lake Valley region (Willits, California) provides constraints on mechanisms of basin formation along this young segment of the San Andreas fault system. Interpretation of isostatic gravity anomaly data provides insight into fault geometry, basin structure, and thickness of Quaternary fill in Little Lake Valley, California. Although the active strike-slip Maacama fault zone diagonally trends through the southwest part of the valley, gravity and geologic interpretations indicate the valley conceals an earlier basin and faulting history. The isostatic gravity anomaly of the basin is negative (up to 13 mGals) and rhombic in shape. Modeling indicates two splays, less than a km apart, of an up-to-the-east East Valley fault; the basinward fault is buried by fill and the more easterly fault defines the eastern margin of the basin. Cumulative up-to-the-east vertical fault displacement along the East Valley fault increases southward up to 610 m in the southern portion of the valley. Gravity gradients also suggest approximately east-west trending faults bound the northern and southern sides of the valley and offset Quaternary fill. From gravity and geologic data combined, the basin floor dips approximately 7 degrees to the south in the north part of the valley and both the Quaternary sediment and basin floor dip approximately 13 degrees to the north in the south part of the valley, implying an approximately east-west axis of dip reversal of the basin floor at the northern stretch of East Hill Road (latitude 39.39 degrees N). Faults and basin fill structure are not consistent with any one simple structural model of basin evolution. Although gross fault geometry is consistent with a pull-apart basin origin, a simple pull-apart basin model does not account for the timing of basin initiation, dip of Quaternary basin deposits, and relation of basin bounding faults to the main strand of the Maacama fault.
T43A-1099
Paleomagnetic and Seismologic Evidence for Oblique-Slip Partitioning to the Coalinga Anticline From the San Andreas Fault
The Coalinga Anticline is a one of a series of fault-related folds in the central Coast Ranges, California, oriented subparallel to the San Andreas Fault (SAF). The development of the Central Coast Range anticlines can be related to the relative strength of the SAF. If positing a weak SAF, fault-normal slip is partitioned to these subparallel compressional folds. If the SAF is strong, these folds rotated to their current orientation during wrenching. Another possibility is that the Coast Range anticlines are accommodating oblique-slip partitioned from the SAF. The 1983 Coalinga earthquake does not have a purely thrusting focal mechanism (rake =100°), reflecting the likelihood that oblique slip is being partitioned to this anticline, even though surface expression of fold-axis-parallel slip has not been identified. Paleomagnetic vertical-axis rotations and focal mechanism strain inversions were used to quantify oblique-slip deformation within the Coalinga Anticline. Clockwise rotations of 10° to 16° are inferred from paleomagnetic sites located in late Miocene to Pliocene beds on the steeply dipping forelimb and backlimb of the fold. Significant vertical-axis rotations are not identified in the paleomagnetic sites within the nose of the anticline. The varying vertical axis rotations conflict with wrench tectonics (strong SAF) as the mechanism of fold development. We use focal mechanisms inversions of earthquakes that occurred between 1983 to 2006 to constrain the seismogenic strain within the fold that presumably help to build it over time. In the upper 7 km, the principal shortening axis is oriented N37E to N40E, statistically indistinguishable from normal to the fold (N45E). The right-lateral shear in the folded strata above the fault tip, evident from the paleomagnetically determined clockwise vertical-axis rotations, is being accommodated aseismically or interseismically. In the region between 7 and 11 km, where the mainshock occurred, the shortening direction ranges from oblique to normal to the fold trend. Our results show that right-lateral slip is resolved along the main fault plane and not distributed to the smaller aftershocks at depths of 7-11 km. The principal strain axes and clockwise paleomagnetic rotations indicate that the Coalinga Anticline is accommodating minor right-lateral shearing and thus shares some of the strike-slip motion of the San Andreas system.
T43A-1100
3D Fault Geometry and Basin Evolution in the Northern Continental Borderland Offshore Southern California
Grids of recently released high-quality industry multichannel seismic (MCS) reflection data, combined with bathymetry and offshore well data are used to map digital 3D fault surfaces and stratigraphic reference horizons in the northern Continental Borderland offshore of southern California. This area experienced large-scale oblique crustal extension and translation associated with the initiation and development of the Pacific-North American plate boundary. The 3D surfaces of structure and stratigraphy can thus be used to better understand and evaluate regional patterns of uplift, subsidence, fault interaction and other aspects of plate boundary deformation. Our mapping in Santa Cruz basin and on Santa Rosa and Santa Cruz-Catalina Ridge reveals an unusual pattern of faulting, folding and basin subsidence. This subsidence is significant (up to 3-4 km since early-Miocene time) and is responsible for the development of several major Borderland basins. Vertical motions can be estimated from an early-Miocene unconformity that likely represents a paleo-horizontal, near-paleo-sea-level erosional surface. As such, it can be used to reconstruct Borderland forearc geometry prior to rifting, subsidence and subsequent basin inversion. Major findings to date include: (a) a better characterization of the complex 3D geometry and pinch-out of the eastern edge of the northern forearc Nicolas terrane and its implications for Borderland basin development, plate reconstructions, and vertical motions associated with oblique rifting; (b) recognition that the East Santa Cruz Basin fault, previously thought to be a predominantly high-angle, large- displacement right-slip fault representing the eastern edge of the Nicolas terrane, is in fact a series of reactivated right-stepping, NE-dipping reverse-separation faults; (c) discovery that NW-striking faults associated with Santa Cruz-Catalina Ridge bend west into a horse-tail structure to interact with and contribute to the southern frontal fault system of the Northern Channel Islands anticlinorium; and (d) recognition that both Santa Cruz-Catalina Ridge and the even larger Santa Rosa Ridge represent complex inverted basins resulting from post-Miocene compressional folding.
T43A-1101
Strike-slip Tectonics in the Schouten Basin: Western Branch of the Bismarck Sea Seismic Lineation
The Schouten Basin is located offshore the north-western coast of Papua New Guinea, approximately between longitudes 144° and 145°. The major tectonic feature in the area is the Bismarck Sea Seismic Lineation (BSSL), a sinistral strike-slip fault that bounds the north side of the basin and separates the North and South Bismarck Sea Plates. We collected bathymetry and backscatter data in the Schouten Basin and elsewhere in the Bismarck volcanic arc in 2004 aboard the research vessel Kilo Moana. In the area of the Schouten Islands, the BSSL changes its orientation from WNW east of Wei Island (144°21.5) to ENE west of Wei. The predominant structural geometry is a pattern of in-line structures, where several faults are parallel to the strike- slip zone. This geometry could be a result of strain partitioning to accommodate oblique shortening. The fault zone crosses less than 2 km off Wei's south coast and has probably affected the island itself. Our data reveals a major contrast offshore north and south of Wei, with a well developed insular slope and apron on the north side, eroded by a radial system of submarine canyons, and an extremely steep and uncommon insular slope on the south side, that also lacks the presence of an insular apron. We suggest that this south part of the island has been cut off and displaced left-laterally by the BSSL a distance of 45 km. In addition to the main structural direction, approximately E-W, the other predominant direction is given by a set of NE-SW faults. The latter are controlling the orientation of a set of submarine canyons off-shore from the Sepik and the Ramu rivers. These faults may also control local volcanism through the alignment of seamounts.