G41A-01 08:00h
InSAR Observations of low Slip Rates on the Major Faults of Western Tibet
Two contrasting views of the active deformation of Asia dominate the debate about how continents deform. In the first, deformation is primarily localized on a few major faults separating plate-like crustal blocks. In the second, deformation is distributed throughout the continental lithosphere. In the first model, western Tibet is being extruded eastwards between the major faults bounding the region. Surface displacement measurements across the western Tibetan plateau using satellite radar interferometry (InSAR) indicate that slip rates on the Karakoram and Altyn Tagh faults are lower than would be expected for the extrusion model and suggest a significant amount of internal deformation in Tibet. To determine slip rates we used 5 independent, 500-km-long strips of radar data acquired by the ERS-1 and ERS-2 satellites between 1992 and 1999. The data span the entire width of the western plateau, from the Tarim basin in the north to just north of the Himalayas. Atmospheric disturbances were observed to be low on the plateau. The interferograms cover time intervals of between 1.9 and 3.6 years, and were stacked to reduce the impact of atmospheric and topographic errors. We solved for slip rates assuming surface displacements are equivalent to those caused by an infinitely long screw dislocation in an elastic half space, and use a locking depth of 10~km. To account for data uncertainties, we used the Monte Carlo simulation of correlated noise. The right-lateral slip rate of the Karakoram Fault was found to be in the range 1$\pm$3~mm/yr (upper bound of 7~mm/yr at 95% confidence level); the left-lateral slip rate of the Altyn Tagh Fault was found to be 5$\pm$5~mm/yr. These estimates disagree with some geologic measures of slip rates but agree with results from other geodetic data.
G41A-02 08:15h
Microplate Tectonics in East Asia: Evidence from GPS Velocities and Block Modeling
The possibility of Okhotsk plate motion independent of the North American plate and Amurian plate motion independent of Eurasia were tested rigorously using velocities from 106 GPS sites (58 from within the proposed Okhotsk and Amurian plate boundaries) in order to constrain the plate kinematics of East Asia. A block modeling approach was used to incorporate both rigid block rotation and near-boundary elastic strain accumulation effects in a formal inversion of the GPS velocities. Considered models include scenarios with and without independent Okhotsk and Amurian plate motion. Our modeling favors a scenario with independent Okhotsk plate motion but does not require it, based on the application of F-test statistics. Euler poles, calculated from our optimized inversion, suggest an independent Okhotsk plate rotating 0.206 deg/Myr clockwise with respect to Eurasia about a pole located north of Sakhalin. The plate-motion parameters of the Okhotsk plate are consistent with right-lateral motion in northern Sakhalin and contraction in southern Sakhalin, inferred from focal mechanism solutions. However, subtle changes in block and segment geometry can cause significant changes in the estimated pole of rotation of the Okhotsk plate. This is due, in large part, to the close proximity of essentially all the GPS stations in the Okhotsk region to plate boundaries, such as the Kamchatka-Kurile subduction zone and Sakhalin Island contraction and strike-slip shear zone. GPS velocities on the Kamchatka peninsula capture a complex pattern associated with the locked subduction zone. This locked subduction zone may require a more complex model than a simple elastic dislocation for the rotational signal to be resolvable. We also considered the possible influence of an independent Amurian plate that may also affect the determination of Okhotsk plate existence and motion. Independent Amurian plate motion is more difficult to formally document than Okhotsk plate motion due to the diffuse and uncertain nature of the continental plate boundaries and the small number of GPS sites in the region. Nevertheless, the best-fit plate motion model of an independent Amurian plate is consistent with the overall kinematics of earthquakes and structures in proposed plate boundary zones along the Baikal rift and the Stanavoy mountains range.
G41A-03 08:30h
Kinematic Behaviour and Active Tectonics of the SE Asia Triple Junction Inferred by GPS Measurements in Sulawesi (Eastern Indonesia)
Sulawesi Island, eastern Indonesia, is located at the triple junction of the Australian, Philippine, and Sunda Plates generating a complicated configuration of active plates' boundaries. We quantify the kinematics and the active tectonics of Sulawesi and surrounding region based on 10 years of GPS measurements on a network which was expanded from 8 sites in 1994 up to more than 40 sites nowadays. In the Sunda reference frame, the northern part of Sulawesi Island, the Sula Block, moves quickly toward the NNW and rotates clockwise at ~3°/Myr. The southern part, the South Sulawesi block, rotates in the opposite direction at ~2°/Myr. The tectonic boundaries between these two blocks are materialized by the Matano Fault to the South, followed by Palu-Koro Fault to the West and ending northward into the Minahassa Trench limiting Sula and Sunda blocks. The South Sulawesi block can be considered as rigid in first approximation while the Sula block shows an important internal deformation. The Gorontalo fault has been evidenced by geology within the Sula domain. Although few seismicity has been reported, GPS measurements across the fault trace indicate about 8 mm/yr of right-lateral movement across Gorontalo fault with accumulation of interseismic elastic deformation. Moreover, extension on the Tomini Bay, identified by both GPS and seismicity, may indicate back-arc stretching behind Minahassa subduction. Most of the relative motion (about 30 mm/yr) between the two blocks is accommodated on the Palu-Koro and Matano left-lateral strike-slip faults. At the first order, the velocities measured across the Palu-Koro Fault can fit a dislocation model of a fault interseismically locked at 5-10 km. However, both recent GPS measurements and geomorphology analysis in the region of Palu City indicate that the fault presents there a pull-apart complex geometry distributing the motion on two parallel active locked faults. This fault continues northward and connects to the Minahassa Trench that accounts for the motion between Sula and Sunda Blocks. This trench is the location of many earthquakes and GPS time series show non-linear displacements on the subduction interface.
G41A-04 08:45h
On the comparisons of seismic moment rates determined from historic earthquakes, GPS and Quaternary fault slip rates
The long-term horizontal strain rate field in the India-Eurasia plate boundary zone is determined from Quaternary fault slip rates and recent GPS data. Strain rates estimated from recent GPS data in Southeastern and Northeastern China are less than 1x10-9 /yr, insignificantly different from zero, which is consistent with the rigid block or plate assumptions in recent tectonic models. We estimate the tectonic moment rates in various seismic zones in China and surrounding regions using the horizontal strain rates from the joint-inversion of geological and GPS data. The total earthquake moment rate inferred from historic earthquakes over different time periods is about 70-80 % of the total tectonic moment rate within the seismogenic volume. They are consistent with each other within one sigma confidence level. However the moment rates inferred from historic seismicity in the seismic zones containing the Kunlun fault, the Altyn Tagh fault, and Yushu-Mani fault zone in northern Tibet, the eastern Tian Shan, and Shanxi Graben system are significantly smaller than the tectonic moment rates. Our preliminary results also indicate that the extension rates across the Shanxi graben, the Weihe Basin, and Hetao basin are no more than 2 mm/yr. Nor do the left-lateral or right-lateral slip rates across these active graben systems exceed 2 mm/yr. The total horizontal slip rates across these graben structures seem to be lower than the slip rate estimates from previous studies. The discrepancies between the long-term moment rates and moment rates inferred from seismicity or geological information could have significant implications in seismic hazard calculations in China. Although the uncertainties in the current GPS data are still too large to confidently reject or modify any previous results, the addition of GPS data will no doubt help to improve uncertainties or reduce biases in seismic hazard models that are based either on historic seismicity alone or seismicity data combined with geological data.
G41A-05 09:00h
Crustal block kinematics of the Kanto and the Izu regions, central Japan estimated from GPS and leveling data
We develop a kinematic block-fault model (Matsu'ura et al., 1986) in the Kanto and the eastern Tokai and the Izu regions based on geodetic data including GPS and leveling. Tectonics in those regions are characterized by the subduction and the collision of the Izu-Ogasawara(Bonin) arc at the Sagami and the Suruga-Nankai Troughs and the northern Izu Peninsula, respectively. First, we estimated horizontal and vertical velocity at 173 GPS sites from April 1996 to May 2000 and vertical velocity in 605 leveling sections between adjacent benchmarks in late 90's. Then, the observed data are inverted to infer the poles of rotation of the crustal blocks, the degree of elastic strain accumulation on faults, and volumetric change of inflation sources beneath volcanoes in the regions, simultaneously. The data are explained by four distinct crustal blocks, namely, the Izu micro-plate, the Central Japan block, the Pacific, and the Philippine Sea plates. The Izu micro-plate rotates clockwise rapidly (9.4 deg/Myr) with the rotation pole just north of its northern boundary. The zone of the strong-coupled faults extends from the rupture zone of the 1923 M7.9 Kanto earthquake to its southeast extension in the subduction zone along the Sagami trough. The boundary between the Izu micro-plate and the Philippine Sea plate has a left-lateral motion with the rate of ~30 mm/yr at east and south of the Izu Peninsula. The most part of this boundary is completely locked and has a large potential of future earthquakes. Though the subduction boundary along the Suruga-Nankai Trough between the Central Japan block and the Izu micro-plate is completely locked, convergence rate (10-20 mm/yr) is slower than that of the other part of the Nankai Trough ($>$45 mm/yr). It suggests the Tokai gap hypothesis to anticipate an M~8 earthquake along the Suruga-Nankai Trough should be reconsidered because the hypothesis implicitly assume convergence rate does not differ along the Nankai Trough significantly.
G41A-06 09:15h
Seafloor Geodetic Observation Along the Major Trenches Around Japan - Focusing on Results at Off-Miyagi Area -
We have been developing a system for precise seafloor geodetic observation with the GPS/Acoustic combination technique and deploying about fifteen seafloor reference points on the landward slope of the major trenches around Japan, such as Japan Trench and Nankai Trough. The primary purpose of our observation is to detect and monitor the crustal deformation caused by the subduction of the oceanic plate near the plate boundary where huge earthquakes repeatedly occur. The Off-Miyagi reference point is located about 100km landward from the axis of the Japan Trench, NE Japan, where huge earthquakes have repeated with a relatively regular interval of about 30 to 40 years. Four acoustic transponders are installed as a set on the seafloor with the water depth of about 1700m just above the possible rupture area of the future earthquake on the plate boundary. At this point, we have been carrying out intensive campaign observations with our system since 2001. The data analysis consists of (1) kinematic GPS analysis, (2) acoustic wave analysis to obtain round travel time between the transducer on board and seafloor transponder, and (3) a combination of results from (1) and (2) to get the precise seafloor station position. The last analysis (3) is performed by a linear inversion method based on the least squares estimation. Time series of estimated coordinates of the Off-Miyagi seafloor reference point from six campaign epochs during the period 2002-2004 show a linear trend in time in the horizontal components with the repeatability of several centimeters except for one epoch. The linear reduction of the time series gives a crustal movement of about 7-8 +- 1-2 cm/year WNW relative to the stable part of the Eurasian continent, which is reasonable in view of the expected intraplate deformation rate in this area.
G41A-07 09:30h
GPS Measurement of Neotectonic Motions in the Antarctic Interior
Campaign GPS measurements made between 1996 and 2004 are used to describe the crustal velocity field in southern Victoria Land, Antarctica. The GPS network stretches along the Transantarctic Mountains from the region of the David Glacier near $75\deg$S to the Byrd Glacier region at $81.5\deg$S and encompasses volcanic islands within the adjacent West Antarctic Rift. It is composed of 32 sites, six of which have been upgraded to run continuously. The continuous stations have been installed to separate long term secular trends in vertical motion from short period seasonal signals. The array was designed to record intraplate neotectonic motions associated with Glacial Isostatic Adjustment (GIA) and possible neotectonic motions between East and West Antarctica. There is conflicting evidence on the degree of contemporary neotectonic activity in the study area. Seismic reflection profiles document faults cutting through the sedimentary column and reaching the seafloor in the north-south trending Terror Rift in the western Ross Sea, yet seismicity has yet to be detected in this region. GIA models predict that uplift in the region will have a secular motion of between 0 and 4mm per year. This uplift will be associated with a small amount of horizontal motion. GPS records an average horizontal motion of the westernmost sites in our network of 14.4mm/yr towards $149\deg$ east. When this average motion is removed from the motion of all the network sites, to approximate an East Antarctic cratonic reference frame, a significant residual motion of coastal and offshore sites is observed. This residual motion is between 1 and 3 mm/yr towards the east-northeast, perpendicular to the trend of the Transantarctic Mountains and the Terror Rift. This motion may be accommodated by faults that occur onshore in the frontal zone of the Transantarctic Mountains and/or by faults within the Terror Rift. Reasonable models of crustal response to GIA with a laterally homogeneous Earth and the D91 ice sheet history produce horizontal motion orientations similar to the GPS measurements but are unable to reproduce the magnitude of motion. The same models predict vertical motions greater than those observed. These results suggest that GIA drives part of the observed horizontal motion but that there is also an active tectonic component.
http://www.geology.ohio-state.edu/TAMDEF
G41A-08 09:45h
Building on Current Space-Based Geodesy to Infer Long-Term Tectonic Reduction in Continental Area
GPS network data gathered globally over the past decade have been analyzed to calculate plate velocities and a world strain rate model {\it (Kreemer et. al 2003; http://gsrm.unavco.org)}. Use of this global model has allowed us to quantify the current changes in the areal extent of the continental crust and oceanic basins. After applying a correction for the recoverable effects of elastic strain in subduction collision zones, we find that tectonic processes alone produce a reduction in continental area equal to approximately 0.18$\pm$0.02 km$^{2}$yr$^{-1}$ and an oceanic crust creation rate of 2.7$\pm$0.02 km$^{2}$yr$^{-1}$ (primarily at mid-ocean ridges, but also through extension at shallow back-arc basins). This implies that 6% of generated oceanic area is not destroyed through subduction but is taken up through continental deformation. While these estimates represent the current rate of change calculated from GPS, we can infer longer-term rates of similar order through comparisons with geologic data. The global velocities from the global strain rate model used for this study correlate well with velocities determined from 3 Myr averaged plate motion models from sea-floor spreading rates and fault azimuth data {\it (DeMets et. al 1990, 1994)}, implying that the areal change rates estimated here are a reflection of the long-term, 3 Myr average rate. Further, using sediment volume delivered to oceans as a proxy for surface uplift volume produced by specific collision events, we can estimate the collision zone areal reduction rate for the formation of older orogens for the past 450 Myr. Similar estimates can be made for areal creation produced by major rifting events. Without a counterbalance to this tectonic areal reduction, our current rate implies an approximate 10% reduction in total continental area every 100 Myr. The reduction is due to the disparate temporal and geographic prevalence of continental compression versus extension. That is, continental collision and subduction occur throughout the Wilson cycle, while major rifting events are generally singular and relatively short-lived for a particular cycle. If constant or slowly increasing continental area is to be maintained (as suggested in several continental growth models), 7.65$\pm$0.90 km$^{3}$yr$^{-1}$ of new or recycled volume must be added to the continental landmass through one or more of the following methods: sediment accumulation at continental margins producing additional area; introduction of new continental volume through volcanism; incorporation of oceanic crust underlying margin sediment into the continental mass during collision; or some additional processes. Any of these processes will introduce chemical changes to the continental crust and may help explain the structural, compositional, and tomographic disparity between ancient and younger continental regions. We argue that continental areal reduction is a long-term consequence of plate tectonics and likely had a profound influence on the character and evolution of continental crust at least through the Phanerozoic.