Stephen S. Gao

SKS splitting beneath continental rift zones

We present measurements of SKS splitting at 28 digital seismic stations and 35 analog stations in the Baikal rift zone, Siberia, and adjacent areas, and at 17 stations in the East African Rift in Kenya and compare them with previous measurements from the Rio Grande Rift of North America. Fast directions in the inner region of the Baikal rift zone are distributed in two orthogonal directions, NE and NW, approximately parallel and perpendicular to the NE strike of the rift. In the adjacent Siberian platform and northern Mongolian fold belt, only the rift-orthogonal fast direction is observed. In southcentral Mongolia, the dominant fast direction changes to rift-parallel again, although a small number of measurements are still rift-orthogonal. For the axial zones of the East African and Rio Grande Rifts, fast directions are oriented on average NNE, that is, rotated clockwise from the N-S trending rift. All three rifts are underlain by low-velocity upper mantle as determined from teleseismic tomography. Rift-related mantle flow provides a plausible interpretation for the rift-orthogonal fast directions. The rift-parallel fast directions near the rift axes can be interpreted by oriented magmatic cracks in the mantle or small-scale mantle convection with rift-parallel flow. The agreement between stress estimates and corresponding crack orientations lends some weight to the suggestion that the rift-parallel fast directions are caused by oriented magmatic cracks.

Southern African crustal evolution and composition: Constraints from receiver function studies

[1] Stacking of approximately 1500 radial receiver functions recorded at about 80 broadband seismic stations deployed in southern Africa reveals systematic spatial variations in the ratio of crustal P and S wave velocities (Φ), crustal thickness (H), and the amplitude of the converted Moho phases (R). The eastern Zimbabwe and the southern Kaapvaal cratons are characterized by small H (∼38 km), small Φ (∼1.73), and large R (∼0.15) values, suggesting that the relatively undisturbed Archean crust beneath southern Africa is separated from the mantle by a sharp Moho and is felsic in composition. The Limpopo belt, which was created by a collisional event at 2.7 Ga, displays large H (∼43 km) but similar Φ and R values relative to the cratonic areas. The Bushveld Mafic Intrusion Complex and its surrounding areas show large Φ (∼1.78), large H (∼43 km), and small R (∼0.11) values, reflecting the intrusion of mafic material into the original crust as a result of the Bushveld event at 2.05 Ga. Excluding the Bushveld, the spatially consistent and age-independent low Φ accentuate the difference between felsic crustal composition and more mafic island arcs that are thought to be the likely source of continental material. Within such an island arc model, our data, combined with xenolith data excluding mantle delamination in cratonic environments, suggest that the modification to a felsic composition (e.g., by the partial melting of basalt and removal of residue by delamination) is restricted to have occurred during the collision between the arcs and the continent.

Mantle deformation beneath southern Africa

Seismic anisotropy from the southern African mantle has been inferred from shear-wave splitting mea- sured at 79 sites of the Southern African Seismic Experi- ment. Thesedataprovidethemostdramaticsupporttodate thatArcheanmantledeformation ispreservedasfossil man- tle anisotropy. Fast polarization directions systematically follow the trend of Archean structures and splitting delay times exhibitgeologic control. Themost anisotropic regions are Late-Archean in age (Zimbabwe craton, Limpopo belt, western Kaapvaal craton), with delay times reduced dra- matically in o-craton regions to the southwest and Early- Archean regions to the southeast. While thin lithosphere can account for weak o-craton splitting, small or vertically incoherent anisotropy is a more likely explanation for the Early-Archean region. We speculate that this dierence in on-craton anisotropic structure is the result of two dierent continent-forming processes operating.

Shear wave splitting and mantle flow associated with the deflected Pacific slab beneath northeast Asia

[1] A total of 361 SKS and five local S wave splitting measurements obtained at global and regional seismic network stations in NE China and Mongolia are used to infer the characteristics of mantle fabrics beneath northeast Asia. Fast polarization directions at most of the stations in the western part of the study area are found to be consistent with the strike of local geological features. The dominant fast directions at the eastern part, beneath which seismic tomography and receiver function studies revealed a deflected slab in the mantle transition zone (MTZ), are about 100 from north, which are almost exactly the same as the motion direction of the Eurasian plate relative to the Pacific plate, and are independent of the direction of local geological features. The splitting times at those stations are about 1 s which correspond to a layer of about 150 km thickness with a 3% anisotropy. The shear wave splitting observations, complemented by the well-established observation that most of the eastern part of the study area is underlain by a lithosphere thinned by delamination in the Paleozoic era, can be best explained by the preferred alignment of metastable olivine associated with the subduction of the deflected Pacific slab in the MTZ, or by back-arc asthenospheric flow in the mantle wedge above the slab.

Asymmetric upwarp of the asthenosphere beneath the Baikal rift zone, Siberia

In the summer of 1991 we installed 27 seismic stations about lake Baikal, Siberia, aimed at obtaining accurately timed digital seismic data to investigate the deep structure and geodynamics of the Baikal rift zone and adjacent regions. Sixty-six teleseismic events with high signal-to-noise ratio were recorded. Travel time and Q analysis of teleseisms characterize an upwarp of the lithosphere-asthenosphere boundary under Baikal. Theoretical arrival times were calculated by using the International Association of Seismology and Physics of the Earths interior 1991 Earth model, and travel time residuals were found by subtracting computed arrival times from observed ones. A three-dimensional downward projection inversion method is used to invert the P wave velocity structure with constraints from deep seismic sounding data. Our results suggest that (1) the lithosphere-asthenosphere transition upwarps beneath the rift zone, (2) the upwarp has an asymmetric shape, (3) the velocity contrast is −4.9% in the asthenosphere, (4) the density contrast is −0.6%, and (5) the P wave attenuation contrast t* is 0.1 s.

Seismic anisotropy, mantle fabric, and the magmatic evolution of Precambrian southern Africa

The observed seismic anisotropy of the southern African mantle from both shear-wave splitting and surface wave observations provides important constraints on modes of mantle deformation beneath this ancient continent. We find that the mantle anisotropy beneath southern Africa is dominated by deformational events in Archean times occurring within the lithosphere, rather than present-day processes in the sublithospheric mantle. Consequently, the distribution and magnitude of anisotropy provide valuable data to constrain the mantle’s role in the tectonic evolution of this region. The pattern of mantle anisotropy reveals several noteworthy characteristics. First, mantle anisotropy is closely associated with the Great Dyke of the Zimbabwe Craton, with values of the splitting fast polarization direction, ϕ, parallel to the Dyke. This correspondence with the Great Dyke is likely not due to the present-day Dyke structure but instead is most probably due to the emplacement of the Dyke parallel to pre-existing mantle fabric within the Zimbabwe craton. This deformation thus predates dike emplacement and is no younger than Neo-Archean in age. Second, there is a spatially continuous arc of mantle anisotropy extending from the western Kaapvaal Craton to the northeastern Kaapvaal and Limpopo Belt. All along the arc, ϕ is subparallel to the trend of the arc. Given the crust/mantle chronology associated with these regions, the anisotropy likely represents deformation that occurred at ~2.9 to ~2.6 Ga during collisional accretion of both the western Kimberley and northern Pietersburg blocks onto the seismically isotropic eastern shield of the Kaapvaal, with accretion on the northern ramparts of the Kaapvaal ultimately culminating in the Neo-Archean Limpopo orogen. The anisotropy-inferred arc of deformation reveals diverse zones of both strong and weak coupling between the crust and mantle, as measured by the coherence between mantle deformation and geologically-inferred surface deformation. In particular, there is high coherence between surface and mantle deformation at the southwestern and northeastern ends of the arc, which implies strong crust-mantle coupling in these regions. Conversely, apparent decoupling exists in the northwestern portion of the arc, where northeast to southwest trending anisotropy cuts across north to south trending structures, such as the surface outcrop and aeromagnetic expressions of the Kraaipan Greenstone Belts. Independent seismic evidence from seismic reflection profiling supports the conclusions that these north-south-trending crustal features are superficial and confined to the upper crust. We present evidence that the mantle fabric producing seismic anisotropy constitutes fossil structure in the mantle that is subsequently reactivated, much like the more commonly acknowledged reactivation of crustal structures. In particular, we argue that Neo-Archean collisional orogenesis imparted a mechanical anisotropy to the mantle that controlled the subsequent magmatic history of cratonic southern Africa. We furthermore suggest that four major Precambrian magmatic events: the Great Dyke, the Ventersdorp, Bushveld, and the Soutpansberg, all represent extensional failure along planes oriented parallel to the local splitting fast polarization direction. Each of these events is interpreted to be a collisional rift, similar to the Baikal rift of northern Eurasia, where the stress field associated with collision produces extension and rifting for orientations at a small angle to the direction of the collision. Precise crustal geochronology associates both Ventersdorp and Great Dyke magmatism with the earliest and latest phases of the Limpopo collision, respectively. Similarly, the Bushveld magmatic event is temporally linked to the ~ 2.0 Ga reactivation of Neo-Archean structures in the Limpopo and surrounding areas by the Magondi Orogen, and the Soutpansberg is related to the ~1.9 Ga Kheis Orogen. Since the timing of these basaltic intrusions is controlled by temporal variations in lithospheric stress associated with orogenesis, it implies either that the melting process is genetically related to the evolution of the far-field collision, or that there was a semi-permanent reservoir of basaltic magma residing in the sublithospheric mantle during the ~1 billion-year time period spanned by these magmatic events. The existence of an extensive magma reservoir would argue for elevated temperatures just beneath the lithosphere during this time. Splitting delay times, δ t , a measure of the magnitude of anisotropy, reveal geologically controlled variations in the strength of anisotropy. In particular, the Meso-Archean Kaapvaal shield, the area that was not exposed to ~2.9 Ga and later deformational events, is effectively isotropic. We observe two areas where the anisotropic/isotropic transition is relatively sharp. The north-south boundary appears to coincide with the east-west trending Thabazimbi-Murchison Lineament. In the west, the boundary has been observed in the vicinity of Kimberley, South Africa, near the Colesberg Magnetic Lineament. The Eastern Shield has been relatively devoid of the kind of rifting and magmatic events seen elsewhere in cratonic southern Africa since the Meso-Archean, suggesting that the Eastern Shield lithosphere is mechanically stronger than surrounding areas. This relative strength difference may in part be due to the absence of the mechanical anisotropy inferred for the surrounding areas.

SKS splitting beneath southern California

Measurements of SKS phase splitting were obtained from nineteen seismic stations in southern California. The fast polarization directions are 53° at the southern end of the Great Valley, 82±8° in the western Transverse Ranges and northern Peninsular Ranges, 95±4° in Mojave Desert, and 70° on San Clemente Island. The splitting time ranges from 0.8 to 1.8 seconds, which is consistent with an anisotropic layer of 100 to 200 km thick for 4% anisotropy.

Seismic anisotropy beneath the Afar Depression and adjacent areas: Implications for mantle flow

[1] Shear wave splitting is a robust tool to infer the direction and strength of seismic anisotropy in the lithosphere and underlying asthenosphere. Previous shear wave splitting studies in the Afar Depression and adjacent areas concluded that either Precambrian sutures or vertical magmatic dikes are mostly responsible for the observed anisotropy. Here we report results of a systematic analysis of teleseismic shear wave splitting using all the available broadband seismic data recorded in the Afar Depression, Main Ethiopian Rift (MER), and Ethiopian Plateau. We found that while the ∼450 measurements on the Ethiopian Plateau and in the MER show insignificant azimuthal variations with MER‐parallel fast directions and thus can be explained by a single layer of anisotropy, the ∼150 measurements in the Afar Depression reveal a systematic azimuthal dependence of splitting parameters with a p/2 periodicity, suggesting a two‐layer model of anisotropy. The top layer is characterized by a relatively small (0.65 s) splitting delay time and a WNW fast direction that can be attributed to magmatic dikes within the lithosphere, and the lower layer has a larger (2.0 s) delay time and a NE fast direction. Using the spatial coherency of the splitting parameters obtained in the MER and on the Ethiopian Plateau, we estimated that the optimal depth of the source of anisotropy is centered at about 300 km, i.e., in the asthenosphere. The spatial and azimuthal variations of the observed anisotropy can best be explained by a NE directed flow in the asthenosphere beneath the MER and the Afar Depression.

Analysis of deformation data at Parkfield, California: Detection of a long‐term strain transient

Analysis of more than a decade of high-quality data, particularly those from the two-color electronic distance meter (EDM), in the Parkfield, California, area reveals a significant transient in slip rate along the San Andreas Fault. This transient consists of an increase in fault slip rate of 3.3±0.9 mm/yr during 1993.0 to 1998.0. The most reliable fault creep instruments show a comparable increase in slip rate, suggesting that the deformation is localized to the fault which breaks the surface. There was also an increase in precipitation around 1993. It is unlikely, however, that this anomaly is due directly to hydrology, as its spatial distribution is what would be expected for increased slip on the San Andreas Fault. The increase in slip rate corresponds temporally to a dramatic increase in seismicity, including the four largest earthquakes in the period 1984–1999 that occurred along a 6-km segment of the fault just to the north of the EDM network. There was also a previously reported anomaly in borehole shear strain [Gwyther et al., 1996] that closely corresponds temporally to the transient in EDM data. Solely on the basis of EDM data the transient can be modeled as a slip event on a 10-km-long segment of the fault. The calculated shear strains from this model, however, are not consistent with the observed ones. A compatible model can be found if there is increased aseismic slip to the northwest in conjunction with the four earthquakes. Support for this northwestern slip is provided by a recent study of slip rate based on microearthquake activity. We speculate that this northwestern event served to load the fault to the southeast, with the stress being partially released by the observed slip.

Making Reliable Shear-Wave Splitting Measurements

Shear-wave splitting (SWS) analysis using SKS, SKKS, and PKS (here- after collectively called XKS) phases is one of the most commonly used techniques in structural seismology. In spite of the apparent simplicity in performing SWS measure- ments, large discrepancies in published SWS parameters (fast direction and splitting time) suggest that a significant portion of splitting parameters has been incorrectly determined. Here, based on the popularly used minimization of transverse energy technique, we present a procedure that combines automatic data processing and care- ful manual screening, which includes adjusting the XKS window used for splitting analysis, modifying band-pass filtering corner frequencies, and verifying and (if nec- essary) changing the quality ranking of the measurements. Using real and synthetic data, we discuss causes and diagnostics of a number of common problems in perform- ing SWS analysis, and suggest possible remedies. Those problems include noise in the XKS window being mistaken as signal, non-XKS seismic arrivals in the XKS window, excessive use of null ranking, measurements from misoriented sensors and from sen- sors with mechanical problems, and inappropriate dismissal of usable measurements.