S33A-1034
P-Wave Velocity Structure and Its Tectonics beneath Dabie-Sulu Region
The 3-D crustal structure of P-wave velocity in the Dabie-Sulu region and the vicinities was studied based on the data obtained by wide-angle seismic reflection and refraction surveys. The results suggest that the high velocity structure zones exist in the upper crust shallower than 20 km beneath the Sulu and Dabie regions. The cause of high velocity zones is attributable to the high pressure metamorphic (HPM) and ultra-high pressure metamorphic (UHPM) terranes with high velocity and density exhuming up to the upper crust and the surface in the Sulu and Dabie orogeny. While, anomalous zones of low velocity reveal in the lower crust near 30 km beneath the Sulu and Dabie regions, respectively. The velocities distributing south of the Qinling mountain are generally greater than those north of the Qinling in the crust deeper than 10 km. The Moho discontinuity is as deep as 38 km beneath the Dabie region, deeper than those in its surroundings obviously. The crustal thickness is about 32-33 km in the Sulu region. The Moho discontinuity beneath the Sulu orogenic region is deeper than those beneath its vicinities a little. The deep Moho discontinuity zone implies the low crustal velocity structure zone in those regions. The low velocity characteristics in the lower crust are probably related to the remnant crustal root of the old mountains due to the orogeny in the Sulu and Dabie regions. So the results in the present analysis suggest that the high or low velocity zones are attributable the orogeny in the Dabie-Sulu region. The high velocity anomaly zones in the upper crust or low velocity anomaly zones in the lower crust beneath the Sulu region are always located northeast of the northern segment of the Tanlu fault, however, the high or low velocity anomaly zones beneath the Dabie region are correspondingly located southwest of the southern segment of the Tanlu fault. Such distribution for the velocity anomalous zones looks to be attributable a left lateral motion along the Tanlu fault. The distribution pattern of velocity anomalous zones may show some evidence for the left strike-slip motion regime of the Tanlu fault.
S33A-1035
Shear-wave Velocity Structure in the Southern Korean Peninsula
In order to reveal the perspective of velocity structure in the southern part of the Korean peninsula, a seismic refraction profile was obtained along a 335-km NNW-SSE line in 2004. Seismic waves were generated by detonating 500~1000 kg explosives in drill holes at depths of 80~150 m. The seismic signals were recorded by portable seismometers at a nominal interval of 1.7 km. Shear velocity tomogram was derived using a series expansion method of traveltime inversion. The raypaths indicate existence of mid-crust interfaces at approximate depths of 2~3 and 15~17 km. The refraction velocity of the deeper interface, which may be the boundary between the upper and lower crust, is approximately 3.8 km/s. The Moho discontinuity with refraction velocity of 4.5 to 4.7 km/s has a maximum depth of 38 km under the southern central portion of the peninsula. The Moho becomes shallower as the central part of the Korean peninsula is approached.
S33A-1036
New insight into structural heterogeneity beneath Taiwan
To know whether the Eurasian lithosphere subducts beneath Taiwan is an important issue for a better understanding of mountain building, arc magmatism and plate collision in the western Pacific region. High- resolution 3-D velocity images are estimated at depths of 0-400 km beneath Taiwan by inverting a large number of arrival times from local and teleseismic events simultaneously. We used 215,676 P-wave arrival time data from 6782 shallow and intermediate-depth earthquakes that are located in and around the Taiwan Island. We also used 12,078 P-wave arrival times that are collected from 3-componenet seismograms of 1108 teleseismic events recorded by the networks installed by Taiwan, Japan and China. Our tomographic images provide further direct geophysical evidence for the tectonic models proposed by previous studies and revealed some new features of structural heterogeneity related to the subducted Eurasian lithosphere and the subducting Philippine Sea slab. Low-velocity anomalies beneath the active volcanoes are visible in the subduction zone of Taiwan, which might caused by the collision between the subducted Eurasian plate and the subducting Philippine Sea slab. In the southern portion of Taiwan, the Eurasian lithosphere is clearly imaged as a high velocity zone with a thickness of 65-80 km and subducted down to a depth of 300 km, whilst it has not been observed beneath North Taiwan. Despite that the existence of subducted Eurasia slab beneath Taiwan has been documented by Lellamant et al. (2001), the present study is the first one to provide high-resolution image and indicate that the Eurasian lithosphere stops at the depth of 300 km beneath South Taiwan but not under North Taiwan. Meanwhile, the present tomographic results are also coherent well with the geology and with plate reconstructions in the region. The previous study proposed that the plate convergence rate is constant at about 7 cm/yr (Seno et al., 1993), it takes about 4-5 Ma for the subducted slab of the Eurasian plate to reach a depth of 300 km. This is remarkably consistent with geological inferences that the orogeny of Taiwan was initiated in the early Pliocene ( about 5 Ma) (e.g., Teng, 1990, Kao et al., 2000). The results of physical modeling of the arc-continent collision in Taiwan also require a slab remnant down to about 300 km beneath the island (Chemenda et al., 2001). Therefore, it is possible that the observed high-V anomalies are associated with the subduction system since the inception of the regional collision. Our results also indicate that the Philippine Sea slab is subducting northwestward from the Ryukyu Trench down to a depth of 200 km, showing good agreement with the previous seismic, geochemical and geophysical studies. The plate convergence of the Eurasian plate varies from subduction beneath South Taiwan to colliding with the Philippine Sea slab under North Taiwan. These characteristics of the structural heterogeneities in the crust and upper mantle suggest that mountain building process in the central region of Taiwan, magmatism and seismictectonics in the subduction zone are mainly attributed to the collision between the subducted Eurasian lithosphere and the subducting Philippine Sea slab.
S33A-1037
The Lithospheric Velocity Structure in New Madrid Seismic Zone from Teleseismic Travel Time Data
Totally 2664 teleseismic P arrivals have been used to invert the 3D velocity fluctuations in New Madrid Seismic Zone (NMSZ). These new teleseismic dataset were manually collected from 122 earthquakes (magnitude >= 5.0 Mb) and 105 stations through the period 1999-2005. Teleseismic travel times were determined by a semi- automated method of Multi-Channel Cross-Correlation (MCCC). We used a shorter correlation window for impulsive arrivals and a longer correlation window for un-impulsive ones to achieve high quality readings. Using the measured travel times we tomographically mapped the 3D P wave velocity structure in NMSZ in a space of 500 x 500 x 200 km. The tomographic model shows that a consistent low velocity anomaly is strongly correlated with three main NMSZ faults, where dense seismicity concentrates. This low anomaly extends down to the bottom of the model (200 km) with a decreasing magnitude of perturbation. It is also oriented parallel to the major trend of NMSZ faults, or the Realfoot Rift Zone (RRZ). Our observed anomaly of relatively low velocity lies within a much larger anomaly of faster Pn velocity that spans entire NMSZ and surrounding regions. It requires that the absolute velocity of this relatively slow anomaly is more close to the average velocity in the uppermost mantle whereas the surrounding regions are substantially faster. The low velocity anomaly in the crust seems to correlate with both the active faults and high gravity in NMSZ. It suggests that a fault zone may extend throughout much of the crust.
S33A-1038
Regional Significance of Lower-Crustal Reflectivity in the California Continental Borderland
The California Continental Borderland (CCB) is key to understanding the Neogene tectonics of western North America, because it lies near the point where the East Pacific Rise is thought to have first encountered the continental margin. Thus, the CCB marks the probable site of the initiation of the San Andreas transform boundary and the migration of the Mendocino and Rivera triple junctions. The inner CCB is believed to have formed by exhumation during oblique right-lateral transtensional motion of the outer CCB. Two modes have been proposed for the exhumation of the inner CCB, either through metamorphic core complex (MCC)-type exhumation above a crustal-scale detachment or magmatic upwelling within a slab gap or window. The models differ in that one predicts the presence of fossil oceanic crust at the base of the CCB crust, whereas the other does not. Seismic investigations have yielded conflicting results that support both the existence of oceanic crust, west of the western Transverse Ranges (EDGE line RU-10) and the absence of oceanic crust beneath the inner CCB, south of Santa Catalina Island (Los Angeles Regional Seismic Experiments [LARSE] lines 1 and 2). In this work, we re-interpret data from the 1991 PRAN offshore/onshore seismic wide-angle reflection experiment near Oceanside, California in hopes of resolving questions surrounding the different models for inner CCB exhumation. Six of the eight receiver-gathers from the PRAN transect show two wide-angle reflections that originate at depths of approximately 14 and 22 km beneath the inner CCB. We interpret the upper reflector to be the top of an 8 km thick layer that has a P-wave velocity of 6.7 - 6.9 km/s and an average density of 3000 kg/m3. This layer thickens to 12 km beneath the Peninsular Ranges. The deeper reflector is from the Moho. We interpret the basal layer to be fossil oceanic crust, possibly thickened by magmatic underplating. The occurrence of a lower-crustal reflector beneath most of the outer and inner CCB, west of the western Transverse Ranges and south of San Clemente Island, may support exhumation above a detachment that soles into the top of oceanic crust. On the other hand, absence of this reflector beneath the inner CCB crust between San Clemente and Santa Catalina Island may suggest that the basal layer is locally broken or delaminated.
S33A-1039
Upper Crustal Velocity Model of the Goldstream Valley, Central Alaska
A series of five explosions were detonated in the schist bedrock of the Goldstream Valley region of central Alaska and recorded on more than 120 local sensors to develop 1-D and 3-D models of the upper crust. Simple refraction analyses reveal that both P- and S-wave arrival times are azimuth dependent, with the fastest velocities seen in the north and northeast for P-waves and in the southeast and northeast for S-waves. In one layer over a half-space P models, average velocities of 3.46 ± 0.41 km\backslashs for layer 1 and 4.79 ± 0.53 km\backslashs for layer 2 are seen. Two layers over a half-space models indicate average P-wave velocities of 3.05 ± 0.19 km\backslashs (layer 1), 3.90 ± 0.18 km\backslashs (layer 2), and 5.12 ± 0.35 km\backslashs (layer 3). Analysis of S-wave travel time data yields an average velocity for only one layer (2 or 3), with a value of 2.77 ± 0.47 km\backslashs. The azimuthal dependence is modeled with velocity anisotropies of ± 1-10% and ± 2-12% for P and S, respectively, at a fast direction of ~40-70°, which is consistent with the northeast trending fault pattern of the region. Poisson's ratio calculations reveal a well constrained average value of 0.283 ± 0.031 at distances ≥ 7 km and a more scattered value of 0.277 ± 0.062 at 2-7 km distance from the source. These values are consistent with known Poisson's ratios for mica-quartz schist and greenschist or amphibolite facies pelitic schist, which are the major geological components of the region. The highly scattered Poisson values at close distances are likely due to the higher porosity of the bedrock at low pressures. In addition to the simple travel time analysis, the S-wave velocity structure of the upper crust is determined through multiple filter analysis and inversion of 0.4-2 sec period Rg waves. The resulting shear wave velocity model is incorporated into the refraction analysis results to help constrain the 1-D and 3-D models of the area.
S33A-1040
Crustal Thickness Estimates for the Chugach-St. Elias, Alaska, Region From Receiver Functions
During the summers of 2005 and 2006, twenty-two broadband seismic stations were installed in the Saint Elias and Chugach terranes of south-central Alaska as part of the Saint Elias Erosion and Tectonics Project (STEEP). We computed receiver function estimates from these 22 stations plus 16 broadband stations in the Alaska seismic network. Initial results were computed using a conventional waterlevel method, but work is ongoing to use a new array-based deconvolution method. Receiver function estimates are stacked with a novel nonlinear stacking method that automatically sorts data by coherence with the array stack. We find that receiver functions in the interior are more internally consistent than comparable data from the STEEP area. We suspect this may be due to stronger scattering induced by the extreme topography and complicated crustal structure that characterize the study area. The stacked waveforms show systematic variations with distance from the coast, suggesting a dipping structure related to deformation of the Yakutat Block. Work is in progress to invert the stacked receiver functions for velocity models to build a map of crustal thickness for this region.
S33A-1041
An Investigation of Lithospheric Thickness Beneath Stations of the Canadian National Seismographic Network
Recent studies have identified Sp conversions as a promising tool for characterizing the Lithosphere/Asthenosphere Boundary (LAB). The S-receiver function holds advantage over more conventional P- receiver functions since the conversions arrive within a time window that is free of crustal multiples. The disadvantage is the greater background level of signal generated noise and a more restricted selection of suitable events. We have assembled an S-wave data set from the Canadian National Seismograph Network to identify and characterize the depth of the LAB across Canada. S-receiver functions for each station are generated by transformation of particle velocity to upgoing wavefield components, followed by least-squares deconvolution. Signal quality is monitored through a culling of data whose signal-to-noise ratio lies below 65 dB based on the log variance before and during the direct S-arrival. A summary depth profile is generated by stacking receiver functions along predicted travel time curves for a 1-D model. The method has been calibrated through comparison with recently published results on LAB signals at other locations around the globe. On average there ~40 S-receiver functions per station corresponding to epicentral distances between 55° and 85° to minimize interference with SKS and ScS. Back-azimuthal coverage is geographically limited to three distinct sectors, the North-West Pacific, South America, and the Himalayas. Our initial focus is on the transition from North American Cordillera to Canadian Shield across western Canada, where a large and abrupt change from thin to thick lithosphere is expected on the basis of previous elastic thickness, surface wave, and heat flow investigations. Preliminary results from our data set indicate clear Moho arrivals at depths consistent with those determined from P-receiver functions. Arrivals at times and slownesses consistent with an origin near the base of the lithosphere are also evident at some stations. The nature of these signals and their relationship to the LAB will be discussed.
S33A-1042
Concordia, Antarctica, Seismic Experiment for the International Polar Year
Concordia, Antarctica, Seismic Experiment for the International Polar Year (CASE-IPY for short) is part of a larger IPY initiative, the Polar Earth Observing Network (POLENET), which includes contributions from numerous countries including France, Italy, the United States and China. POLENET will focus on deployment of autonomous observatories at remote sites on the continents and offshore, coordinated with measurements made at permanent station observatories. With CASE-IPY, we plan to deploy 8-10 broad-band seismometers in East Antarctica, from the French/Italian station Concordia at Dome C, to the Russian station Vostok, with a possible extension towards the Chinese station at Dome A in collaboration with our Italian colleagues. We plan to deploy this profile during the 2008/2009 Antarctic summer season. In order to prepare for this deployment, we shall be installing three autonomous seismic stations close to the Concordia base this coming Antarctic summer (Dec 2007 - Jan 2008). Concordia is the site of an experimental permanent seismic observatory station, which has been operational since 2005. The extreme temperatures present at the site (-60°C) imply difficult operating conditions for the seismological equipment. The quality of data we obtain from this station has been steadily improving as we resolve the technical issues related to working at such low temperatures. The lessons we have learned from operating the Concordia station are being applied to the design of our CASE-IPY autonomous stations. We shall present details of the CASE-IPY project and results from the past two years of seismic observation at Concordia.
S33A-1043
Tomographic Mapping of Lg Propagation in Northern Europe
How Important is the Tornquist Zone? We look at crustal differences across the geologically significant Tornquist zone. Recordings of regional earthquakes in Northern Europe have been collected in attempt to find the Lg-surface wave propagation. We classify the Lg-wave propagation paths according to weather the Lg wave propagation is good, intermediate or poor. We have included the available data from the period 1994-2004, accepting earthquakes with at least three amplitude measurements. With the relatively dense areal coverage provided by paths crossing the region it is possible to invert for the pattern of crustal heterogeneity which gives rise to the observed character of the Lg-wave propagation. We are using data from permanent seismic stations.
S33A-1044
Moho Depth in Peninsular Italy From Teleseismic Receiver Functions
We analyzed teleseismic data recorded in the past 4 years by the Italian National Seismic Network and MedNet to retrieve the Moho depth beneath Italy. To improve coverage in areas where the sampling is low, we added data recorded by temporary experiments carried out in the region during the last few years (i.e. Retreat, CatScan, and others). Our data-set comprises more than 8,000 teleseismic waveforms recorded at 165 broad-band stations. The stations were classified in five quality categories, according to the sharpness of the Ps converted phases and to the presence and continuity of multiple phases. We extracted first-order information about the crustal properties (i.e. Moho depth and Vp/Vs ratio) using the technique developed by Zhu and Kanamori (2000). We modified the original stacking method using three different crustal Vp values. Also, the definition of the quality level for each single station has been used to weight the stacking result. A first idea on the Moho geometry beneath the Italian peninsula comes from the distribution of the stations as a function of the quality level. The stations deployed along the Tyrrhenian side of peninsular Italy and above the Apulian platform are characterized by high quality ranking, while medium- to low-quality data can be found everywhere along the Apennines. Data-set coming from the Southern Apennines and the Calabrian arc display very low quality. The main reason for the decreased data quality in these areas is the presence of sharp velocity discontinuities at shallow depth, such as the Apulian carbonate platform below the Apenninic allochtonous cover in the southern Apennines, or the complex upper crustal structure in the Calabrian arc. In the regions of simple crustal structure, corresponding to high quality ranking, the technique allows to determine clearly the Moho depth and the mean Vp/Vs value in the crust: in the peri-Tyrrhenian region the crust is as thin as 20 km; in the Apulia foreland region we find a thicker crust (30-35 km); finally, the Apenninic belt exhibits highly variable values of crustal thichness, generally between 35 and 40 km. Other regions along peninsular Italy show a complex pattern, and a unique clear Moho is not easily identified.
S33A-1045
Structure and Anisotropy Beneath Southern Italy
The crust and upper mantle structures beneath southern Italy are often associated with rollback and fragmentation of the Western Mediterranean subduction zone in the past 30 million years. In this study, we utilize broadband records from the Calabria-Apennine-Tyrrhenian/Subduction Collision-Accretion Network (CAT/SCAN) to probe the effect of the past and on-going plate motions beneath this region. Waveforms from two distinct frequency ranges are examined in detail using both forward and inverse waveform modeling approaches, and the resulting 1-D models for each path are subjected to a Monte-Carlo uncertainty test. By analyzing the Love and Rayleigh waves from two regional earthquakes during the temporary deployment between 2003 and 2005, we are able to retrieve information on the anisotropic seismic structure down to 200-km depth. Our study shows that the average seismic structure beneath Calabria/Apulia is significantly faster than that beneath the Apenninic mountain chain. Sharp changes in seismic velocities, regardless of wave polarizations, lend strong support for the distinct geologic histories of the major tectonic units. The difference between Love- and Rayleigh-wave models provides further constraints on the dynamic processes and crust/mantle fabric beneath the study area. Our preliminary results show relatively minor anisotropy beneath the southeastern Tyrrhenian Sea and Calabrian Arc region, which could be partially explained by the rapid changes in the alignment of olivine fast crystallographic-axes due to the complex arc geometry. The ray paths connecting the Ionian Sea and the Southern Apennines reveal strong anisotropy, where transversely polarized waves travel at higher speeds than vertically polarized waves. Both "frozen-in" and flow-induced anisotropy could contribute to the observed waveform differences between the Love and Rayleigh waves sampling this region.
S33A-1046
Anisotropy Studies in Central Greece
The Gulf of Corinth, located in Central Greece, is a tectonic graben characterized by high seismicity level. GPS measurements indicate extension of the Gulf in an approximately N-S direction, with a rate of 10 to 15 mm/year. The southern part of the Gulf is dominated by the presence of large active normal faults in an almost E-W direction, dipping north, resulting to the subsidence of the central part of the graben. Analysis of data recorded by the Cornet network, which is the permanent network of the University of Athens, revealed the existence of an anisotropic upper crust at the eastern part of the Gulf of Corinth. Anisotropy was also observed at the western part of Corinth Gulf, using data recorded by a temporary seismological network installed in the area. Furthermore, shear wave splitting analysis was performed in the region of Attica (to the NE of the Gulf) that hosts Athens, the capital of Greece, using aftershocks of the 1999 Athens earthquake (Mw=6.0) that caused 143 fatalities. The methods used for the determination of the splitting parameters are the polarization vector as a function of time (polarigram) and the hodogram. For each selected event the direction of polarization of the fast shear wave, the delay between the two split shear waves and the polarization of the source were measured. Concerning both parts of the Gulf of Corinth, the obtained mean values of anisotropy vary between N90° and N142°. In the region of Attica the mean values of the anisotropy direction of all stations vary between N95° and N100°, almost parallel to the azimuth of the Parnitha fault. The time delay between the split shear waves vary between 0.020s and 0.130s. The obtained anisotropy measurements are in agreement with the extensive dilatancy anisotropy (EDA) model, since the direction of anisotropy is independent from the event-station azimuth and perpendicular to the direction of extension. ACKNOWLEDGMENTS The present study was co-funded by the European Social Fund and National Resources – (EPEAEK II) PYTHAGORAS, contract No. 70/3/7306.
S33A-1047
Elastic-Anelastic properties beneath the Aegean inferred from long period Rayleigh Waves
This work is towards contributing to the better knowledge of the deep structure of the Aegean by introducing experimental elastic and anelastic parameters via the study of long period Rayleigh waves. For this scope path- average phase velocities and attenuation coefficients of fundamental Rayleigh waves crossing the Aegean were extracted over the period range 10-100 s. It is mean worth that it is the first time that anelastic parameters of the long period wavefield are determined for the region. The wavetrains were recorded at the broadband stations installed some years ago in the Aegean region for the SEISFAULTGREECE project. The stochastic inversion algorithm has been used to derive 36 path-average models of shear velocity and 19 path-average models of inverse shear Q down to 200 km. Average over the study region shear Q values at depths from 0 to 200 km range between 29±13. The observed low shear Q likely indicate that fluids reside in lower crustal, as well as upper mantle depths. Furthermore, the elastic and anelastic 1-D path-average models were combined in a continuous regionalization tomographic scheme to obtain a 3-D model of shear velocity variation down to 200 km and a 3-D model of inverse shear Q variation down to 120 km. The most prominent features in the tomograms are: a) A low shear velocity zone in the back-arc region, especially in the central and north Aegean. This region is located south of the North Aegean Trough (the western edge of the North Anatolian Fault) and correlates well with the derived anelastic tomograms which present high attenuation in this area. b) A high velocity/low attenuation zone in South Aegean indicating the subducted African lithosphere beneath the Aegean. The zone in central and north Aegean characterized by low velocities/high attenuation is compatible with a region of high extensional strain rates, recent volcanism and high heat flow. These observations suggest a hot or perhaps partially molten upper mantle and/or distributed deformation beneath the study region, probably related with the slab roll-back that has accompanied back-arc extension. ACKNOWLEDGMENTS The present study was co-funded by the European Social Fund and National Resources - (EPEAEK II) PYTHAGORAS, contract No. 70/3/7306.
S33A-1048
Variation in crustal structure in Iran and the surrounding region
We present a model for the topography of the Moho discontinuity for Iran and its surrounding regions. This is produced using data from field deployments within Iran by the University of Cambridge (UK) and the Universite Joseph-Fourier (FRA) in conjunction with International Institute of Earthquake Engineering and Seismology (Iran), in addition to data from IRIS and Geofone. We determine tomographic group velocity maps for periods between 10 and 60 s from multiple filter analysis of ~5500 seismograms. Because of the dense path coverage, these images have substantially higher lateral resolution for this region than is currently available from global and regional group velocity studies. Joint inversion of receiver functions and Rayleigh wave dispersion give accurate crustal velocity structures at 96 sites within Iran These provide a constraint for the less sharp crustal velocity profile produced by inverting the Rayleigh wave dispersion curve across all of Iran. We observe variations in the crustal thickness across the region, consistent with the surface topography. The thickest crust (55-60 km) is found beneath the central Zagros mountains, with the crust in the remainder of Iran having a thicknesses of 40-50 km. No significant increase in Moho depth is seen beneath the Alborz or Kopet Dagh mountains. The structure of the South Caspian Basin is presented with a different structure to that found in previous studies, with a crustal thickness of 50 km in the west, and beneath the Caucasus and Talesh mountains, in the middle part of the basin, over the course of the ~100km, this decreases to 40km, and continues to 35 km beneath the Turkmen Platform. Comparisons are also made between the joint inversion results, and accurate hypocentre depths for regional earthquakes. This shows most events occur in the upper crystalline crust (~10-20km depth), with few in the lowest velocity layer. Almost no events are located in the lower crust, and only in the Makran and Aspheron- Balkhan Sill do earthquakes appear in the Upper Mantle.
S33A-1049
Crustal Structure And Tectonic Evolution Of The Kaapvaal Craton
The joint inversion of Rayleigh wave group velocities and body wave receiver functions is used to constrain the shear wave velocity structure across the Kaapvaal craton and surrounding mobile belts. The joint inversion results indicate the presence of mafic lithologies in the lower crust predominantly around the northern and eastern parts of the Kaapvaal craton, Limpopo Belt and southern parts of the Zimbabwe craton. The rest of the south and central portions of the Kaapvaal craton including the adjacent post-Archaean belts show a lower crust that is probably intermediate-felsic in composition. The P-to-SV conversions and crustal reverberations (PpPmS) from the receiver function results show higher amplitudes for localities with a lower crust that is intermediate-to- felsic in composition. The results further show a thinner crust within the cratonic areas (typically 34 to 45 km). The post-Archaean regimes show a thicker crust: Limpopo Belt (42 to 53 km), Namaqua-Natal Belt (42 to 50 km), Bushveld Province (44 to 49 km). However, the Cape Fold and Kheis Belts show Moho depths that are also thinner and comparable to those within the cratonic areas (34 to 40 km). Results for this study suggest that most of the south-western and central parts of the craton could be intermediate-to-felsic in composition. It is plausible that the lack of the mafic lithologies in the lower crust in the central and south-western parts of the craton could have resulted from cycles of extensional deformation in the post-Archaean associated with a number of extensional events (e.g. Ventersdorp and Transvaal). Because of crustal thinning during these events, the lower crust, which was orignally mafic in compostion, could have been thinned a sufficient amount such that it cannot now be resolved seismically.
S33A-1050
Shear Wave Velocity Structure Of The Bushveld Complex
The Bushveld Complex is a prominent ultramafic intrusion located in the northern part of the Kaapvaal Craton. It is unknown if there is connectivity and continuity of the ultramafic units at depth across the complex and especially beneath the central zone. The object of this study is to model the structure of the complex to depths of around 10 km by first measuring Rayleigh wave group velocities from 1 to 10 second periods, and then inverting them to obtain a model of the variation of shear wave speeds with depth. The group velocity measurements are being made mostly using earthquakes (ML greater than 2.5) from the local gold mines that were recorded by the SASE (South African Seismic Experiment) broadband network (1997 to 1999) and on AfricaArray stations. Data from regional earthquakes surrounding the Bushveld Complex is also being used to improve azimuthal coverage. The velocity models should shed some light on the stratification and tectonic setting of the ultramafic layering and also reveal whether there is connectivity and continuity at depth for areas without surface exposure.
S33A-1051
Rayleigh wave tomography of southern Africa: Preliminary results from AfricaArray
In this study data from permanent AfricaArray broadband seismic stations in southern Africa are combined with data from GSN stations to create phase velocity maps of fundamental mode Rayleigh waves. Phase velocity measurements are made using the Forsyth and Li method, where incoming Rayleigh waves are approximated as two plane waves to account for the distortion of the incoming wave field by heterogeneities outside the array. The distribution of the AfricaArray stations allows for better imaging of the lithosphere near the craton edges than is possible using previous datasets. Rayleigh wave dispersion curves and two-dimensional phase velocity maps produced from the data recorded in the first year will be presented, along with a preliminary interpretation for lithospheric structure.
S33A-1052
Crustal structure beneath Southwestern Mexico
The MARS ( Mapping the Rivera Subduction zone ) project started in January 2006 deploying 50 broadband seismometers across southwestern Mexico for one and a half year duration. The stations were deployed in Jalisco, Michoacan and Colima states. The goal of the project is to understand the geometry of the Rivera and Cocos subducting plates and the effect of the subduction on the overriding plate. In this study, we employ the teleseismic receiver function technique to map out the lateral variation in Moho depth as well as the Vp/Vs ratio of the crust in this tectonically and magmatically active area. The ambiguity between the delay time of Ps and crustal Vp/Vs ratio is reduced by stacking later phases, the PpPs and PpSs + PsPs, for different values of Moho depth and Vp/Vs ratio (Zhu et al. ). An average crustal depth and crustal Vp/Vs ratio is obtained by finding the highest combination of parameters that give the largest amplitude stack. We find that the average Moho depth is 39 km but varies significantly from 25 to 45 km thick. The average crustal Vp/Vs ratio is 1.82 but is also variable ranging from 1.7 to 1.9. We will discuss correlations of crustal thickness and Vp/Vs ratio with crustal composition and magmatic activity.