S11E-01 08:00h
Shear-Wave Structure Beneath Southern Africa From Rayleigh Wave Tomography
Southern Africa is characterized by the Archean Kaapvaal and Zimbabwe cratons with several Proterozoic mobile belts surrounding them. To investigate the evolution of the cratons, we have analyzed Rayleigh wave data recorded at the Southern Africa Seismic Experiment, which consisted of 82 broadband seismic stations and operated from April 1997 to July 1999. A two-plane-wave inversion technique is adopted in this study to account for the non-planar energy in the incoming wave field. Average phase velocities are obtained at 14 periods ranging from 20 s to 100 s and vary from 3.65 km/s at 20 s to 4.11 km/s at 50 s and 4.22 km/s at 100 s. These values agree well with Rayleigh wave phase velocities in the Kaapvaal craton in a global model. Fast anomalies at wavelengths of 200-400 km are imaged in the Kaapvaal and Zimbabwe cratons at periods less than 45 s and low velocities appear in the Bushveld-Complex province and the nearby mobile belts. At longer periods, high velocities on relatively small scales are observed both within and outside the cratons, suggesting that the cratonic lithosphere might have been eroded at great depths due to small-scale mantle convections. The lowest velocity is found under the Namaqua-Natal and Cape-Fold belts, the youngest region in the study area. We will resolve the 3-D shear-wave structure from these phase velocities and discuss the dynamics of the cratonic lithosphere in southern Africa.
S11E-02 08:15h
Evidence for a westward dipping upper mantle thermal anomaly beneath Kenya, East Africa from body wave tomography
Upper mantle structure beneath Kenya, East Africa, is re-examined using body wave tomography. Previous body wave tomography models of the Kenya rift did not include structure deeper than 165 km, leaving open the question of the depth extent and nature of the thermal anomaly beneath this rift. The data for this study come from the 1985 and 1989 KRISP project and the 2001-2002 Kenya Broadband Seismic Experiment. Relative P wave residuals have been obtained using a multi-channel cross correlation method and upper mantle structure has been imaged using the inversion scheme of VanDecar. Preliminary results show that thermally perturbed structure beneath the Kenya rift extends to depths in excess of 300 km, and that the thermal anomaly dips to the west beneath the Tanzania Craton. Our results for Kenya are consistent with a similar study to the south in Tanzania that also shows a deep-seated thermal anomaly that dips to the west under the Tanzania Craton. This finding corroborates previous studies suggesting that the locus of the thermal upwelling beneath East Africa is under the Tanzania Craton and not under the Kenya rift. If a mantle plume is invoked to explain the thermal anomaly, then the plume head must be situated under the Tanzania Craton, with warm plume material flowing around and up the sides of the thermally resistant cratonic lithosphere.
S11E-03 08:30h
Upper Mantle and Transition Zone Seismic Velocity Structure Beneath Ethiopia
Throughout much of the Cenozoic, Ethiopia has undergone extensive rifting, volcanism and uplift, though the origin of this tectonism remains ambiguous. Previous studies suggest that slow seismic velocities may extend through the upper mantle beneath this region, consistent with a lower mantle origin for the Cenozoic tectonism. To further understand the origin of the tectonism in Ethiopia, we analyze data collected from the 2000-2002 Ethiopia Broadband Seismic Experiment. We invert P and S wave travel time residuals to examine the upper mantle seismic velocity structure, and use receiver function analysis to map topography on the 410 and 660 km discontinuities. Results from our tomography study beneath Ethiopia reveal a broad westward dipping low velocity anomaly in the upper mantle beneath the Afar triangle and the Western Ethiopia Plateau that appears to extend to depths greater than 400 km. The low velocity anomaly found in our tomographic models appears to be similar to the broad, westward dipping low velocity anomaly beneath Africa (African Superplume) seen in some global tomographic models. This correlation suggests that the low velocity anomaly may extend from the lower mantle to the upper mantle. We have stacked receiver functions to examine the 410 and 660 km mantle discontinuities beneath Ethiopia using both 1D and 3D techniques. The 410 km discontinuity is clearly imaged and shows significant topography. In contrast, the 660 km discontinuity does not appear consistently across the study region. The structure of the discontinuities will be correlated with the P and S wave tomography models to assess further the possibility of a through-going mantle thermal anomaly.
S11E-04 08:45h
A new High-Resolution Model for the Velocity and Density Structure of Northwest Europe
A new high-resolution model for the crustal velocity structure of northwest Europe has been produced from wide-angle reflection seismic data. These data provide the best constraints on the P-wave velocity structure of the crust and depths to major interfaces (e.g.~Moho). Over 31000~km of wide-angle seismic profiles have been digitised to map the Moho depth and 3D variations in velocity structure. The profiles have been individually assessed and assigned uncertainties based primarily on the shot- and receiver-spacing, the data coverage and modelling method. Using 3D kriging, of the profiles, a regional model, with 40x40x1~km grid elements, has been produced. The velocity uncertainties assigned to the input data have been combined with the interpolation uncertainties to produce a quantitative estimation of the model uncertainty. Near surface sediment thicknesses were obtained from separate seismic reflection, borehole and geological databases since the resolution of near surface structure on most wide-angle seismic profiles is poor due to large shot and/or instrument spacing. Velocities were assigned to the sedimentary layer using a velocity-depth model constructed from normal incidence reflection data from around the UK. To the authors' knowledge the crustal model is unique in two respects: The high-resolution, resulting in greater detail in crustal structure, and the inclusion of quantitative uncertainties. The crustal velocity model has been converted to a density model using an empirical velocity-density function. This density model is currently being used to identify areas of buried loading and hence to investigate the isostatic balance of the crust and support for present day topography. Other possible uses of the model are: a starting model for tomography studies; improved earthquake location; and studies of crustal evolution.
S11E-05 09:00h
3-D Local Earthquake Tomography Within The Marmara Sea Region
High resolution P velocity structure of the Marmara Sea Region has been imaged using selected subsets of the 3700 Izmit and Duzce, Turkey aftershocks recorded by 77 seismic stations. We performed a step by step increase in complexity of inversion to equalize the path coverage and earthquake distribution. First, we used 440 homogenously distributed events to develop a minimum 1-D model for the entire region. This model was used as an initial model to the 3-D inversion. The 3-D inversion is a non-linear iterative algorithm and simultaneously solves the coupled hypocenter-velocity problem. We used the same 440 events to develop a smooth 3-D model with a coarse grid of 30 km. The station and event distribution is very dense in eastern Marmara. So, we performed an inversion with a finer grid and selected earthquakes which had a minimum of 8 P readings and maximum 180\deg azimuthal gap. We performed checkerboard test and tests with synthetic velocity models that represent the observed structure. In addition to synthetic tests we used all the other available information, such as resolution estimates, diagonal element of the resolution matrix and spread function to assess the solution quality of our model. Preliminary results show significant heterogeneities in the upper crust. We observe concentrations of seismicity of the relocated earthquakes in several distinct zones of the North Anatolian Fault.
S11E-06 09:15h
Determination of the 8° discontinuity beneath the major tectonic units of Central Europe from regional seismicity in Europe and northern Africa
We evaluate the existence and the depth of the '8° discontinuity' beneath the Alpine orogen using the natural seismicity of Europe and northern Africa as well as events induced by mining activity. For this analysis, the regional events (1) must have epicenters further than 1000 km from the structure being imaged, and (2) the magnitude of body waves must be higher than 4.0 to obtain a favourable signal to noise ratio. The events satisfying the above conditions have epicentres in Algeria, Spain, Bulgaria, Greece and in the Lubin Copper Basin in Poland. The last region is characterised by high seismicity resulting from mining activity. We base our analysis on P-wave traveltime residuals compared to the general iasp91 model. The 8° discontinuity seems to be attributed to the observed P-wave traveltime delays at epicentral distances around 800 km. The analysis of events from the Lubin Coper Basin and the events from other regions mentioned above, gives P-wave delays of 3 s at the Alpine stations in comparison with stations in the Variscan areas to further north. We attribute this variation in travel time to the difference between 'fast' and 'slow' uppermost mantle structures in Europe.
S11E-07 09:30h
3-D Teleseismic Tomography of the Crust and Upper Mantle Beneath Northern Tasmania, Australia
The TIGGER project is a multi-faceted seismic study of Tasmania and southern Victoria (SE Australia) undertaken by the Australian National University in 2001/2002. As part of this project, an array of 72 short period and broadband seismic recorders with a nominal spacing of 15 km was deployed across northern Tasmania for a period of five months. To date, nearly 6,000 relative arrival times from 100 earthquakes have been picked using a newly developed and robust adaptive stacking technique. The azimuthal coverage of teleseisms is generally good, with many events to the north and east (e.g.~Indonesia, Papua New Guinea, New Zealand, Fiji), although fewer from the south and west(e.g.~South Sandwich Islands, mid- Indian ridge). A new iterative non-linear tomographic inversion procedure based on the fast marching method (FMM), a grid based eikonal solver, and a subspace inversion scheme, is used to map traveltime residual patterns as P-wave velocity anomalies from an {\em ak}135 reference model. The 3-D model volume beneath the array is parameterized using cubic B-spline functions in spherical coordinates; a total of nearly 10,000 vertices at approximately 15 km grid spacing is used to describe the TIGGER model. Preliminary tomographic results from the TIGGER experiment show significant lateral variations in P-wave velocity structure within the Tasmanian lithosphere. Geological inferences made from these early results include: (1) Within the crust, the first-order E-W velocity variations strongly support the idea that eastern Tasmania is underlain by dense rocks with an oceanic crustal affinity, contrasting with the continentally derived lower crustal rocks of western Tasmania; (2) the Tamar Fracture System, often defined as a lithospheric scale discontinuity, probably does not exist; (3) the elevated crustal velocities beneath the Rocky Cape Group and Arthur Lineament, compared to the Tyennan Element and Mt. Read Volcanics to the east, also support a mafic-felsic compositional variation in the deep crust, possibly caused by a collision-obduction event in the Cambrian; and (4) the pattern of velocity variations within the crust does not have a simple relationship to that in the mantle lithosphere beneath.
S11E-08 09:45h
Body wave tomography of Iranian Plateau
The inverse teleseismic tomography approach has been adopted to study the P and S velocity structure of the crust and upper mantle across the Iranian Plateau. The method uses phase readings from earthquakes in a study area as reported by stations at teleseismic and regional distances to compute the velocity anomalies in the area. This use of source-receiver reciprocity allows tomographic studies of regions with sparse distribution of seismic stations, if only the region has sufficient seismicity. The input data for the algorithm are the arrival times of events located in Iran which were taken from the ISC catalogue (1964-1996). All the sources were located anew using a 1D spherical Earth model taking into account variable Moho depth and topography. The inversion provides relocation of events which is done simultaneously with calculation of velocity perturbations. With a series of synthetic tests we demonstrate the power of the algorithm to resolve both fancy and realistic anomalies using available earthquake sources and introducing measurement errors and outliers. The velocity anomalies show that the crust and upper mantle below the Iranian Plateau comprises a low velocity domain between the Arabian Plate and the Caspian Block, in agreement with models of the active Iranian plate trapped between the stable Turan plate in the north and the Arabian shield in the south. Our results show clear evidence of subduction at Makran in the southeastern corner of Iran where the oceanic crust of the Oman Sea subducts underneath the Iranian Plateau, a movement which is mainly aseismic. On the other hand, the subduction and collision of the two plates along the Zagros suture zone is highly seismic and in our images appear less consistent than the Makran region.