S52B-01 10:20h
Studying global seismicity and other phenomena with the Global Seismographic Network
The Global Seismographic Network (GSN) is the primary source of data for studies of global earthquake activity. The growth of the network to more than 100 well-distributed stations has dramatically increased our ability to detect, locate, and investigate earthquakes anywhere on Earth. The Harvard centroid-moment-tensor (CMT) project was started more than 20 years ago to study global and regional seismicity using data from the global network. The CMT catalog now contains more than 21,000 moment-tensor solutions for the period 1976 to the present, providing a nearly complete record of earthquakes with magnitudes greater than 5.5. Recent progress in the mapping of crust and upper-mantle structure has translated into an improved ability to model shorter-period surface waves. Starting with earthquakes in 2004 we now use these surface waves in the CMT analysis, and are able to analyze significantly smaller earthquakes routinely. Improved station coverage, real-time data access, and new data-processing techniques now allow us to perform much of the CMT analysis automatically and in near-real time following even moderately sized earthquakes. To complement the earthquake reporting of the NEIC, we have recently developed and implemented a technique to use the GSN as a global, steerable array to search for localized sources of intermediate-period (40--150 s) Rayleigh waves. The method detects nearly all earthquakes with magnitudes greater than 5, including a substantial number of events that go unreported in standard catalogs (PDE, ISC, REB). Many of the previously undetected events are sufficiently large that we can perform a CMT analysis to determine their focal mechanisms, and most are found to be standard tectonic earthquakes. We have also identified many seismic events that appear to be non-tectonic, in particular more than one hundred M$\sim$5 earthquakes on Greenland that we attribute to sudden glacial sliding. Other unassociated events await analysis and interpretation. Results from our CMT project and real-time search for seismic events can be found on our web site, www.seismology.harvard.edu.
S52B-02 10:35h
The Lithosphere of the NW-Atlantic and the Iceland Plume Track
The dynamics of the North Atlantic is dominated by the Iceland plume and the mid Atlantic spreading center separating the North American and Eurasian lithospheric plates. An important parameter for understanding the plume-lithosphere interaction is the thickness of the lithosphere which we determine with very high accuracy. The new seismic technique of S receiver functions (using S to P converted teleseismic waves, Li et al.2004) provides images of the lithosphere-asthenosphere boundary (LAB) with as yet unprecedented detail. The lithosphere under most of the Iceland and large parts of central Greenland is about 80km thick. This is an unexpectedly large value for Iceland since it is located on a mid ocean ridge. Our interpretation is that the Icelandic lithosphere may not be oceanic in origin, but continental. In the region of flood basalts in eastern Greenland, where the Iceland plume track is proposed (Lawver and Müller 1994), the lithosphere is only 70km thick, about 10 km less than in Iceland directly above the plume. We interpret this as rejuvenation of the lithosphere along the plume track. At the coast of western Greenland, the lithosphere thickens to 100-120km and no indication of the Iceland plume track is found there.
S52B-03 10:50h
Data From Massive Seismic Experiments in Central Europe Produce Lithospheric Transects From East European Craton Through the TESZ, Bohemian Massif and Carpathians to the Pannonian Basin and Eastern Alps
A series of very large seismic refraction experiments (POLONAISE'97, CELEBRATION 2000, ALP 2002, SUDETES 2003) began in 1997 in Central Europe. As a result of these efforts, a large region has been covered by an unprecedented network of profiles and with 3-D coverage. The total length of all the profiles is about 19,000 km and over 300 explosive sources were recorded. This network of profiles extends from the East European Craton (EEC) across the bounding Trans-European suture zone TESZ) to traverse the adjacent Bohemian massif, Carpathian Mountains, Pannonian basin and Eastern Alps. Analysis and modeling of the two longest profiles (P4 from POLONAISE'97 $\sim$1000 km long and CEL05 from CELEBRATION 2000 $\sim$1400 km long) produced synoptic lithospheric velocity models. The variations in lithospheric structure across the region are large and complex. For example, the crustal thickness changes from 22-25 km in the Pannonian basin to $\sim$55 km in the Trans-European suture zone in SE Poland to $\sim$50 km in the EEC. Regionally, the most complicated structure observed is in the transition from the Pannonian basin to the EEC. The sedimentary cover with velocities Vp$<$5.5 km/s reaches a maximum depth $\sim$20 km in this region. In the Pieniny Klippen belt (boundary between the Inner and Outer Carpathians), a low velocity sedimentary slab dips towards the south at 35-40o to a depth of 10 km. In the TESZ area along profile CEL05, all of the crustal layers are upwarped producing high velocities ($\sim$6.4 km/s) at a depth of only $\sim$10 km. Since good quality record sections were obtained to offsets of $\sim$600 km from the shot points, clear reflected/refracted arrivals from the lithospheric mantle were observed. Two-dimensional interpretation of the reversed system of travel-times constrains a series of reflectors in the depth range of 50-90 km that dip to the north suggesting northward subduction or a younger thermal feature. To the north along profile P4, the Paleozoic platform and EEC are divided by the Polish basin. In the area of the Polish basin, the P wave velocity is very low ($<$ 6.1 km/s) down to depths of 15-20 km, indicating that a very thick sedimentary sequence is present. We suggest two possible tectonic interpretations of the velocity models: (1) Baltica indented Avalonia, obducting its upper crust and underthrusting its lower crust and (2) a rifted margin of Baltica underlies the Polish basin. The CEL03 profile is parallel to the TESZ and demonstrates that crustal variations along the feature are as complex as the ones along it. This profile suggests that several terranes lie along the TESZ but whether they are silvers of Baltica or Gondwanan in origin remains to be determined. Together all these experiments are providing an unprecedented 3-D image of the lithospheric structure of the Central Europe, which has major tectonic implications.
http://paces.geo.utep.edu
S52B-04 11:05h
Integrating global and regional datasets for tomography in North America
The velocity structure of the upper mantle under North America is of great interest as the USArray project's deployment of a dense grid of broadband seismographs begins. The data from the stations of \linebreak USArray will provide a regional-scale complement to the large volume of high-quality seismic data now available from the Global Seismographic Network of the Incorporated Research Institutions for Seismology. We explore approaches for combining regional and global datasets for tomography of the North American upper mantle. We use a large dataset of surface-wave phase-delay measurements (35~$\leq T \leq$~350~s) made at stations of the IRIS GSN, the U.S. National Seismic Network (USNSN), and the Canadian National Seismograph Network (CNSN). We supplement this dataset with phase-delay measurements made at a selected set of PASSCAL array deployments, including MOMA, BEAAR, and RISTRA. We determine a regional model of the three-dimensional radially anisotropic shear velocity structure under North America that is consistent with long-wavelength, global models of the upper mantle, but that resolves structure on a wavelength of a few hundred kilometers throughout most of the continent. The correspondence between major geological features and those imaged in our mantle model is generally good. Radial anisotropy is observed to be weaker under cratonic regions than under the Basin and Range, where the amplitude of radial anisotropy reaches 4--6%.
S52B-05 11:40h
Rayleigh-wave multi-pathing along the west coast of North America
We have created a movie of surface ground motions for the November 3, 2002, Denali fault earthquake based upon spectral-element simulations using crustal model CRUST2.0, mantle model S20RTS, topography and bathymetry model ETOPO5, and a finite-fault slip model. The movie features two anomalous wave packets that travel along the west coast of the North American plate following off-great-circle paths. These wave packets are Rayleigh waves with dominant periods around 20 sec, which are also found in seismograms recorded by the Southern California Seismic Network. One of these packets is the direct surface wave, whose group arrival-time changes laterally as dictated by the shape of the Oregon coast. The other packet is a surface wave reflected by a lateral interface underneath the Rocky Mountains. A linear reflector parallel to the Canadian coast offsetting it by a few hundred kilometers can explain its arrival time, but the offsetting distance derived from the synthetic seismograms puts the reflector 350 km northeast of the result obtained from the data, indicating a need to update the crustal and mantle models in this area. Records of other big earthquakes (Mw>6) are analyzed to better constrain this reflector and forward calculations using the Spectral element method (SEM) will be preformed to evaluate the final result.
S52B-06 11:55h
IRIS and the S-velocity structure of the North American upper mantle
Owing to its US-based origin and resulting seismogram holdings the Data Management Center (DMC) of the Incorporated Research Institutions for Seismology (IRIS) has greatly facilitated waveform tomographic studies worldwide, and for North America in particular. We report on one such undertaking, in which nearly one and a half thousand seismograms from the IRIS DMC and the Canadian National Seismic Network have been interactively analyzed and used in a Partitioned Waveform Inversion for a tomographic model for the three-dimensional S-velocity structure of the North American upper mantle. A predecessor (NA95) of this new model is consistent with global tomographic models and revealed additional detail such as an upper-mantle component of subducted Farallon lithosphere, an enigmatic structure for the Wyoming lithosphere, and a V-shaped dent in the new England cratonic lithosphere. These details in turn helped spark additional IRIS activity in the form of further analyses of the data holdings of the DMC and PASSCAL experiments addressing these details. The new model provides relatively high-resolution images of the high-velocity rigid root beneath the Canadian shield and central US, which extends to depths of 200-300 km, the low velocities beneath the tectonically active Cordillera and the continent west of it, which also reach depths of 200-300 km, and details herein such as those mentioned above. Below these structures, high-velocity features in the transition zone are not as steep as but in line with the dipping high-velocity Farallon slab imaged in the lower mantle with tomographic methods that include teleseismic body waves. Increased accuracy in the new model, relative to its predecessors, is largely a result of extending the data base that constrains it. We checked the effects of using sensitivity kernels that cover elliptical areas around the great circles and found that they do not lead to better a posteriori data fits.
S52B-07 12:10h
A High-Resolution, Shear-Speed Model of the Upper Mantle
Thanks to the expansion of seismographic networks within the last two decades and to the efforts being put at the IRIS data center into archiving, quality control, and distribution of the data, millions of top-quality earthquake recordings are now readily available from the center. One of the present challenges is to exploit more fully the resolving power of this extraordinary volume of data so as to increase further the resolution of tomographic imaging of the Earth. Here we process long-period, vertical-component seismograms recorded globally in 1994-2002 and available from the IRIS archives. We use the Automated Multimode Inversion of surface and S wave forms (Lebedev and Nolet, 2003), a technique aimed at (i) processing very large amounts of data, by making use of full automation; (ii) extracting the maximum of information from an S---surface wave train while ensuring that relevant approximations hold; (iii) balancing this information when relating it to Earth structure (in particular by weighting waves of different amplitudes and different types) so as to constrain unbiased Earth models. Waveform inversion enables mapping both the structure in the upper 200-300 km of the mantle (constrained primarily by the fundamental mode) and the transition-zone structure (constrained by regional S and multiple-S waves). The data sampling given by the 80,000 automatically selected and inverted seismograms is sufficient to achieve a resolution of a few hundred kilometers, varying with sampling density. The structure in the upper 100-300 km of the model correlates with the distribution of known tectonic elements, including backarc basins, spreading centers, and most cratons. Many of the recognized hotspots are underlain by low velocity anomalies extending as deep as the transition zone, which is consistent with a deep origin for the intraplate volcanism observed. We shall discuss selected features of the model in the presentation.
S52B-08 12:25h
3-D Shear Wave Speed Structure Beneath the Philippine Sea Plate
The Philippine sea is a marginal basin in large part opened through the two episodes of back-arc spreading. The evolution history of the Philippine Sea plate should be reflected in the upper mantle structure. However, the spatial resolution achieved by previous studies is not good enough to discuss seismological structures in terms of the plate tectonic history. Recently, several stations in oceanic islands have been installed and long-term broadband seismic observations on the seafloor have been conducted as a part of the Ocean Hemisphere Project (OHP). These seismic observations, in addition to existing data, enable us to reveal the seismic structure of the Philippine Sea plate with an unprecedented resolution. We measured phase velocities of the fundamental and first three higher modes of Rayleigh waves for the source-station pairs within a latitudinal range from $-20^\circ$S to $45^\circ$N and a longitudinal range from $110^\circ$E to $165^\circ$E, using a fully non-linear waveform inversion method by Yoshizawa and Kennett (2002). The measured multi-mode phase velocities are inverted to a 3-D shear wave speed structure using the three-stage inversion technique by Yoshizawa and Kennett (2004), which allows us to incorporate the effects of finite frequency as well as ray path deviation from the great-circle. The inverted model has a good resolution in the upper 250km of the mantle, showing a persistent feature of subduction of the Pacific plate against the Philippine Sea plate along the Izu-Bonin-Mariana trenches. Subduction of the Philippine Sea plate beneath the Philippine islands can also be seen. The thickness of the Philippine Sea plate is in general significantly thinner than the Pacific plate and the shear wave speed in the asthenosphere is very slow. The southern, older part of the Philippine Sea plate is much thicker than the rest. The slowest shear wave speed of the Philippine Sea plate can be found along its spreading axis at the Mariana trough.