S33A-01
Upper Mantle Anisotropy in Brazil With SKS Splitting: Predominance of Asthenospheric Flow From Absolute Plate Motion
Upper mantle anisotropy, due to lattice preferred orientation of minerals, like olivine, can be caused by shear deformation from the last major orogeny (usually preserved in the lithosphere) or from the lithosphere/asthenoshere strain related to the present absolute plate motion. Anisotropy causes shear wave splitting, with the fast polarization direction usually sub-parallel to the flow direction of the upper mantle rocks. Upper mantle anisotropy in Brazil was studied with SKS and SKKS shear wave splitting, where the delay time and the fast polarization directions were measured with the method of Silver and Chan (1991). Measurements of splitting parameters have been made at 10 stations of the BLSP02 project, mainly in northern and NE Brazil, complementing the previous study of Assumpcao et al.(2006) in SE Brazil. Detectable anisotropy was found at all stations with delay times ranging from 0.5s to 1.6s (average about 1s). Fast polarization directions tends to be roughly EW oriented, on average, consistent with the absolute plate motion direction given by the HS3-NUVEL1A model. The present data set confirms the deviation from the absolute plate motion caused by upper mantle flow around the keel of the the Sao Francisco craton in SE Brazil. In other parts of Brazil, no significant large deviation from the absolute plate motion could be observed, except at stations PDCB (BLSP02) and RCBR (GT), near the northeastern coast, where a clearly NNE to NE fast direction was determined. This fast direction can be correlated with local geological structures near each station. Alternatively, the NE to NNE fast direction is consistent with the predictions of the upper mantle flow model of Conrad et al.(2007). However, while this flow model is roughly consistent with the splitting observations in SE and NE Brazil, the deviations in the Amazon craton are very large. The fact that most fast directions in Brazil tend to be roughly parallel to the absolute plate motion implies that the contribution to the shear-wave splitting from the frozen lithospheric anisotropy is relatively small, compared to other continental plates.
S33A-02
Regional Seismic Tomography in Brazil and Uncertainty Evaluation Through Jackknife Re- Sampling Method
We used the regional seismic tomography to study the upper mantle beneath SE and Central Brazil. This method is based on the inversion of P- and S-wave relative travel time residuals (VanDecar, 1991) obtained from more than 80 stations in an area of 20 x 20 degrees. The ~11000 P and PKP residuals and ~8000 S, ScS, SKS, and SKKS residuals have been obtained from waveform cross-correlations for up to 12 simultaneous stations. Our results show correlations of seismic anomalies with the main tectonic structures and reveal new anomalies not yet observed in previous works. High velocity anomalies in the western portion of the Sao Francisco Craton support the hypothesis that this craton was part of a major Neoproterozoic plate. Low velocity anomalies beneath the Tocantins Province (mainly fold belts between the Amazon and Sao Francisco cratons) are interpreted as due to lithospheric thinning. Assumpcao et al. (2004) showed a good correlation between intraplate seismicity and low velocity anomalies in this region. The slab of the Nazca Plate is observed as a high velocity anomaly beneath the Parana basin (at 700-1200 km depths). At these depths, large low velocity anomalies appear accompanying the slab. Synthetic tests show that these anomalies are artifacts of the inversion generated by the presence of the slab. We use the Jackknife re-sampling method to evaluate the robustness of the tomographic results with respect to the data. The main advantage is that it is not necessary to assume a particular error distribution, since the model variability is accessed directly from the data variability. The approach is based on a random removal of a small percentage of the data (1%) to generate various new subsets, which are inverted to evaluate the model variability. These local estimates include inherently the highly variable ray coverage and measurement errors and can provide confidence in the interpretation of anomalies. This measure should not be interpreted as the resolution. As expected, the Jackknife approach shows that the inversions are less robust at shallow depth and at the margins of the study volume. The model variability is also used as an additional criterion to determine the optimum number of iterations in the inversion.
S33A-03
Study of a Buried Graben in the Intracratonic Parana Basin, Brazil, With High Frequency Receiver Functions
Teleseismic receiver functions are routinely employed to study the structure of the deep crust and uppermost mantle beneath seismic stations. Receiver functions are built from secondary waves generated by the interaction of a direct P-wave with near-receiver discontinuities. Because of the long trajectory between the source and the receiver, it is necessary to low-pass filter the waveforms at about 1 Hz to observe secondary phases from subsurface discontinuities emerge above the background noise. Here we extend the receiver function method to study the topography of the sediment-bedrock interface beneath five stations in the northern part of the intracratonic Paraná basin. The stations cross two suspected grabens hidden beneath the sediments, as previously revealed by Bouguer gravity anomalies (~16-30 mGal). We utilize nearby Andean earthquakes with good signal-to-noise ratios up to ~5Hz to obtain receiver functions at unusually high frequency contents (Gaussian width ~10 s-1) and to infer the S-velocity variation with depth down to the crystalline basement by jointly modeling the high-frequency receiver functions with low-period Rayleigh-wave group velocities. The 1D models reveal an average regional basement depth of ~3.0 km, in good agreement with basement depth estimates at a nearby well, and the stations located above the gravity lows show deeper basement depths in agreement with the graben hypothesis. Interestingly, our S-velocity models reveal the graben basement is ~4.0 km shallower than predicted from the independent gravity modeling. We show this discrepancy can be explained if sedimentary rocks of lower density, as suggested from an empirical velocity-density relationship applied to our S-velocity models, are assumed. A horst-graben structure beneath the northern part of the Paraná basin integrates both gravity and seismic data sets as long as lower densities (and shallower depths) are assumed.
S33A-04
Discriminating Hidden Structures Beneath the Paraná Basin With Receiver Functions
We investigate the seismic structure of the crust and uppermost mantle beneath the intracratonic Paraná basin of SE Brazil by analyizing P- and PP-wave receiver functions at 14 broadband stations within the basin. The stations were deployed by the University of São Paulo during the 1992-1995 and 2002-2005 Brazilian Lithosphere Seismic Project experiments and cross the surface projection of several independently proposed locations for a suspected cratonic nucleous presently hidden under the basin's sediments. Discriminating among competing hidden structures is critical for our understanding of the subsidence and formation mechanisms of the basin. Analysis of the phase moveout of the Moho interaction phases observed in the computed receiver functions reveals crustal thickness and bulk Vp/Vs ratio variations in the 40-48 km and 1.71- 1.76 ranges, respectively. Joint modeling of receiver function waveforms with independent local surface-wave dispersion velocities yields S-wave velocity variations with depth beneath each station that suggest the highest Vp/Vs ratios are the result of mafic underplating roughly along the basin axis. The overall pattern displayed by our measurements is thus consistent with a hidden structure consisting of several cratonic blocks separated by underplated rift segments, which suggests the locus of subsidence was defined by a preexisting zone of weakness and driven by the gravitational pull of underplated mafic material.
S33A-05
Non-Volcanic Seismic Tremor in the Chile Triple Junction Region: Active Subducted Transform Faults?
We present analyses of episodic non-volcanic seismic tremor recorded at the Chile Ridge Subduction Project
temporary seismic network, operated jointly by the University of Florida and the Universidad de Chile, Santiago.
The 57 station network was deployed to study subduction of the Chile Ridge spreading center in the Chile Triple
Junction region. Tremor activity during the first project year (2004-5) includes 4 or 5 peaks of high activity with a
maximum duration of 48 hours. We observe progressive increases and decreases of activity before and after the
high activity peaks, the frequency content of tremor signals is less than 15 Hz, and we observe low magnitude
earthquakes within the tremors on rare occasions, as has been seen in Cascadia and Japan. Tremor source
regions were determined using the source scanning algorithm of Kao and Shan (2004). The tremors are mainly
concentrated in the northern Taitao Peninsula, in the neighborhood of the now-subducted Taitao transform fault,
which is the subsurface Nazca-Antarctica plate boundary east of the Chile Triple Junction. At depth, the tremors
are distributed between the surface and around 40 km depth, and thus occur mainly in overriding South America.
The Taitao transform fault was the site of an unusual slow seismic event around one month after the great 1960
Chile earthquake. Kanamori and Stewart (1979) observed 1.5-2 hours of continuous Rayleigh wave radiation
consistent with strike-slip motions on the transform fault, but without associated body wave radiation. We
speculate that the tremors are also a result of quasi-continuous slow strike-slip events on the subducted Taitao
transform fault and associated deformation and fluid flow in overriding South America.
http:seismology.geology.ufl.edu/chile
S33A-06
Ancestral Structure of the Neuquén Basin, Supported by an Innovative Deep Seismic Reprocessing
Seismic-tracings comprising both the eastern and western sectors of Sierra de los Chihuidos, showed the deep structure of the Neuquén basin, Argentina. Deep reprocessing of historical industrial seismic-lines supplied interpretive information down to about 30-33 km. Consequently, seismic data reprocessed with "self-truncating extended correlation" confirmed an objective way for acquiring deep-seismic information where standard Vibroseis records are available. In addition, the FMED algorithm was an appreciated nonlinear mathematical tool to improve seismic resolution. Original results accomplished with the above emphasized techniques, revealed a list of concepts summarized along the subsequent comments. An acoustic contrast at about 24 km depth must be the top of the lower Crust. An oblique reflector between 16 and 18 km depth must be assumed as the local image of the master shear that controlled the extension system during the Late Triassic-Early Jurassic period. A sub-master fault dipping about 8° W, surely have been controlling the evolution of `Las Cárceles' area. An important inversion event initiated during the Bathonian-Callovian, sensibly affected the western sector of `Las Cárceles' (that is the site contiguous to the Neuquén river). Significant deposition of synrift sediments (Precuyo Group) originated in contiguous scarp degradation was detected on the western side of `Los Chihuidos' arch, at about 7 km depth. A Pliensbachian-Toarcian bipolar inversion developed during the transition to the Cuyo Group was evidenced in the western area. In the same sector, a middle Jurassic postrift episode is characterized by a deltaic depositional system prograding to the west with accentuate high energy. A deep discontinuity was related with the ancestral origin of the Basin, its seismic tracing permitted to match field results with a scale tank experiment simulating orogenic collapse. Bulk extension of the ancestral thickened crust could be only justified if a relative free boundary is adjacent to the ancestral orogenic domain. In such case, the idea of rollback of the western subducting slab would emerge as the most credible hypothesis.
S33A-07
Seismicity pattern and first b-value mapping of the Caribbean - South American plate boundary in North-eastern Venezuela
The implementation of the National Seismological BB Network in Venezuela gave way to much more precise determinations of earthquake hypocentres. This applies especially to the crustal events in North-eastern Venezuela related to the major El Pilar fault and its sub-parallel fault system, expanding over a width150 to 200 km. However, the improved depth calculations effected also the deeper seismicity locations. A pronounced seismicity gap separating crustal from deeper seismicity could be detected for the first time. The deeper seismicity exhibits a column-like shape reaching a depth of approximately 120 km thus having a width of ~ 50 - 70 km. These features coincide well with observations from the reflection/refraction/teleseismic transect at 64 deg W from the BOLIVAR (Broadband Ocean-Land Investigation of Venezuela) project. The seismicity column is located beneath the area where the profile denotes a step in Moho depth and the seismic gap agrees with the area around the Moho. We used the same seismicity data to do a spatial seismic b-value study. Spatial variations in the seismic b-value obtained from the Gutenberg-Richter relation are known to map out changes in the state of stress and/or material properties. This capability of b has been used to determine structural heterogeneity in faults and related stress inhomogeneities. In general, low b-values are related to increased stress or applied pressure whereas high b-values more likely indicate low stress environments or volumes of high crack densities. The b-value map we derived shows statistically significant variations. The deeper seismicity column is characterized by pronounced low b-values (0.5 and lower) indicating a high stress environment or high strain rates. This supports the hypothesis that the step change in Moho depth observed in the 64deg W profile marks the lower crustal plate boundary between the Caribbean and South America and coincides with a shear tear through the South American lithosphere. We propose that the actively propagating shear tear produces the column-like seismicity pattern and creates low seismic b-values along the entire width of the tear.
S33A-08
Station Corrections for Venezuelan Seismic Network
In order to improve earthqueake locations, anomaly time corrections were determined for each of the 35 broadband seismic station of the Venezuelan National Network. The study is defined for the region between latitudes 5-14 degrees N and longitudes 73-60 degrees W; this area includes Venezuela, The Caribbean Sea, Trinidad and Tobago, Netherlands Antilles and northeast of Colombia. The anomaly time corrections for the stations were obtained by using a data base of 111 well localized local earthqueakes recorded with a magnitude Mw greater than 3.0 for 2-year period from January 2005 to January 2007. These station correction were calculated using software VELEST, which inverts arrival time data to produced one-dimensional velocity models and stations corrections using Join Hypocenter Determination technique. Negative and positive corrections were obtained; the minimum value (-0.95) was observed at station ITEV, while the maximum value (0.85) at ORIV. A plot of the improved seismicity map will be plotted and compared to the existing one for the region.
S33A-09
Spatial distribution of hypocentres and crustal model in Joao Camara, Northeastern Brazil
An important seismic cycle that begun in August 1986 occurred in João Câmara, Northeastern region of Brazil, which lasted for several years producing more than 60.000 small to medium magnitude earthquakes. The more important event of this cycle (mb 5.1 and intensity VII MM) occurred on November 30, 1986. The epicentral area is on the border of the Late Cretaceous sedimentary Potiguar basin that overlays the Precambrian shield. Large Late Precambrian SW-NE transcurrent ductile shear zones divide the shield in this region. These shear zone have been reactivated in the Cretaceous and again in the Cenozoic as brittle structures. Data for this study were collected from a local nine-station network installed by Cardiff and UFRN universities, obtained during June to September 1987. Using a simultaneous inversion method, hypocentres of some earthquakes of that sequence and a local structural model, were obtained with the VELEST program. Those hypocenters were also relocated with the HYPODD program for comparing the results obtained with both methods. An elongated SW-NE feature around represents the seismogenic source 40 km long. Stable solutions for the simultaneous inversion were obtained after about 400 iterations for the northern portion and 53 for the southern portion. The mean structural model beneath the seismographic network show tree layers with increasing VP from around 5.8 km/s to around 6.2 km/s, at the bottom of the third layer in 6.5 km of depth, showing also a probable low velocity layer between 3 and 5.5 km of depth. The hypocentres present depths between 0.9 and 9.0 km. The spatial distribution of those hypocentres and the composite focal mechanism solutions suggest a correlation with a dextral transcurrent fault, passing close to the east of Joao Camara, with a small normal component segmented in at least three no- coplanar portions with a dominant strike of around 41°. The dip angle, oriented to the NW, changes from 76° in the northern portion to 66° in the southern one; whereas the slip angle increases from 10° to 35° from northern to southern portions. The spatial distribution of those events seems to correlate well with the surface tectonic features described above, and with the regional stress pattern. Focal mechanism and breakout data indicate that the region is under a roughly E-W-trending SH max related to a strike-slip faulting regime.