Teleseismic P-wave Travel-time Residuals Research Articles

A strong seismic velocity contrast in the shallow mantle across the Zagros collision zone (Iran)

SUMMARY Surface wave dispersion measurements are interpreted jointly with the inversion of teleseismic P-wave traveltime residuals along a dense 620-km long temporary seismic profile across the Zagros to investigate its upper-mantle structure. The S-wave model determined from Rayleigh wave dispersion in the Zagros fold and thrust belt has high velocities from 4.5 ± 0.2 km s −1 below the Moho to 4.9 ± 0.25 km s −1 at 200 km depth, which are comparable to a shield-like structure. Beneath the suture region from the Main Zagros Thrust (MZT) to the UrumiehDokhtar volcanic arc, S-wave velocities are lower than beneath the Zagros in the top 50 km of the upper mantle, with a minimum of 4.4 ± 0.2 km s −1 at 80 km depth. From 150 km and deeper, S velocities are as high as beneath the Zagros. We suggest that part of the velocity difference at shallow depth is due to higher mantle temperatures and/or higher fluid content beneath the northern half of the profile, but that velocities are too high to support the hypothesis of mantle lid delamination under the suture zone. Teleseismic P traveltime relative residuals display a long-wavelength variation along the transect, with a difference of 1.1 s between negative residuals in the Zagros Simple Folded Belt and positive residuals in Central Iran. This difference backprojects into a 6‐7 per cent lateral variation of P-wave velocity in the shallow upper mantle, with higher VP beneath Zagros and lower VP beneath Central Iran. The main short wavelength variation of the residual is located in the suture region, with late P arrivals in the region of the MZT and early arrivals in the Sanandaj-Sirjan zone (SSZ). Using synthetic models of VP perturbations, we show that the high velocities of the Arabian platform have to extend laterally at least to the SSZ to fit the observed P delays. This model also predicts Rayleigh wave phase velocities, which are within the error bars of the observed dispersion. It supports the model of crustal-scale overthrusting at the MZT.

Litho—asthenospheric structures of northern Italy as inferred from teleseismic P-wave tomography

Regional three-dimensional inversions of teleseismic P-wave travel time residuals recorded by high-frequency regional and local seismic networks operating along the Western Alps and surrounding regions were carried out and lithosphere and upper mantle P-wave velocity models down to 300 km were obtained. Residuals of more than 500 teleseismic events, recorded by 98 fixed and temporary seismic stations, have been inverted. The comparison between real residuals and the ones obtained from tomographic model indicates that the method is able to solve the feature of the regional heterogeneities. Where the resolution is good, coherent lithospheric and upper mantle structures are imaged. In the shallower layers, high- and low-velocity anomalies follow the structural behaviour of the Alpine-Apenninic chains showing the existence of very strong velocity contrasts. In the deepest layers, velocity contrast decreases however two deep-seated high-velocity structures are observed. The most extended in depth and approximately trending NE-SW has been interpreted as a wreck of the oldest subduction responsible of the Alpine orogenesis. The second one, connected to the northwestern sector of the Apenninic chain, appears to vanish at depths greater than 180 km and is probably due to still active Apenninic roots. Cross-sections depict the spatial trend of perturbations and in particular outline the sub-vertical character of the Alpine and Apenninic anomalies. Under the Ligurian Sea, the 3-D inversion confirms the uplift of the asthenosphere in agreement with the tectonic evolution of the basin.

Critical evaluation and geological correlation of teleseismic phase data from Indian seismological stations

We evaluated the quality of seismic phase data from Indian seismological stations through the analysis of teleseismic travel times reported during 1976–83 and infer that only WWSSN stations (NDI, SHL, POO, KOD) apart from GBA and HYB can be rated satisfactory while the majority of stations (more than 40) produce very poor quality data sets. Detailed analysis of teleseismic P-wave travel time residuals shows that while the average structure of the upper mantle beneath India has high velocity (negative residuals) there are marked lateral variations. In particular, three zones of anomalous positive residuals (low velocity) are observed: one beneath the north western part of the Deccan trap, the second covering the southernmost peninsula (granulite terrain) and a third rather localized one, to the north of Delhi coinciding with Delhi-Haridwar ridge. New Delhi exhibits strong negative residuals in the E-SE quadrant along with negative station anomaly, implying that it is underlain by an anomalous high velocity crust/upper mantle. The negative residuals observed over India, continue beneath the Himalaya till the south of Lhasa but change sign further northward, suggesting the northern limit of the Indian upper mantle structure.

A teleseismic and petrological study of the crust and upper mantle beneath the geothermal anomaly Urach/SW-Germany

Teleseismic P-wave travel time residuals recorded by a mobile, shortperiod network are interpreted using the composition of crustal and upper mantle xenoliths. A three-dimensional inversion of the P residuals revealed a small-scale anomalous domain with a velocity reduction of 3% in the crust and a broad low-velocity anomaly of 4% beneath the Moho down to a depth of 50 km. Xenoliths of the Miocene Urach volcanic field (“UVF”, 1300 km 2; 355 eruptive centres, mostly diatremes) comprise phlogopite-bearing, clinopyroxene-rich mantle xenoliths, evidence for large ion lithophile element (LIL) metasomatism in an originally depleted harzburgitic mantle, reflecting chemical modification of the upper mantle beneath the UVF. The metasomatism caused partial melting in the spinel peridotite stability field, indicated by glasses in xenoliths, and a diapiric uprise of the partially molten metasomatized mantle. The observed reduction of the P-wave velocity of 4% in the lithospheric mantle can be explained by a local enrichment of phlogopite accompanied by increased temperatures. The crustal low-velocity body coincides very well with the postulated volume of an ancient intracrustal magma chamber or system of chambers required by the composition of xenoliths, the magnetization of the diatremes and the spatial distribution of the volcanic eruption centres.

Azimuthal anisotropy beneath the Bohemian Massif from broad-band seismograms of SKS waves

Shear-wave splitting in SKS phases recorded by the two broad-band stations KHC (South Bohemia) and KSP (Lower Silesia) is used to determine the anisotropic properties of the lithosphere beneath the marginal regions of the Bohemian Massif. Directions of fast velocity of 80° and 120° clockwise from North were obtained for KHC and KSP, respectively. These directions are consistent with the predominantly E-W direction of fast velocity found by different authors using similar techniques in some regions of Western Europe. They disagree, however, with the directions found for the investigated region by other authors on the basis of teleseismic P-wave traveltime residuals. The value of the time delay between the two splitted shear waves is close to 1 s for both stations implying that the thickness of the anisotropic zone in the upper mantle essentially exceeds 80 km.

Tomographic image of the Juan de Fuca Plate beneath Washington and western Oregon using teleseismic P‐wave travel times

Backprojection tomography is applied to teleseismic P‐wave travel‐time residuals to produce a detailed image of the upper mantle beneath Washington and western Oregon. Beneath the Cascades volcanic arc from central Oregon to central Washington the subducted Juan de Fuca plate is imaged as a high‐velocity quasi‐planar feature dipping ∼65°E. Resolution analysis shows that beneath southern Washington, where resolution is best, the steeply dipping slab extends to a depth of only about 300 km. The slab may end near this depth, or low‐velocity structure in the vicinity of the slab or a shallowing of the slab dip may produce this result. Beneath Oregon, where resolution is poor, the slab extends steeply into the mantle to at least 150 km depth. The uppermost mantle beneath northeastern Washington is high in velocity, preventing the resolution of the slab geometry there.

P-wave travel-time residuals in the Newfoundland Appalachians

Teleseisms and quarry blasts were recorded at 24 temporary seismograph sites along a line across the northern Dunnage Zone of Newfoundland. Azimuthally dependent functions were also determined for teleseismic P-wave travel-time residuals at the permanent St. John's (STJ) and Corner Brook (CBK) seismographs. In order to minimize the effects of reading errors and structure far from the receivers, absolute residuals at the field sites were then obtained by correlating the records with those from STJ and CBK and using the residual functions for these two stations. The P-wave residuals in north-central Newfoundland tend to be up to half a second later than at STJ and CBK, with the more positive values towards the east of the region. The direction of fastest approach for rays varies from southeast at STJ to southwest towards the western side of the Dunnage Zone. The travel-time residuals are generally similar to variations observed elsewhere in the Appalachians.The residuals and gravity data indicate that the crust is about 40 km thick over most of the Dunnage Zone, possibly increasing by several kilometres under the Gander Zone and western Avalon Platform, which could be consistent with a former subduction zone in the region. The low crustal velocity of 6.2 km s−1 may support an allochthonous model for central Newfoundland, with average crustal thickening of about 10 km compared with adjacent platform areas.

P-wave travel-time residuals and the crust and upper mantle lateral inhomogeneities in Africa

Teleseismic P-wave travel-time residuals of shallow earthquakes and nuclear explosions were analyzed for 27 African stations. The distribution of positive and negative residuals was determined for Northeast Africa, northern and southern parts of the East African Rift Valleys and South Africa. The existence of lateral inhomogeneities in the upper mantle was found in these regions. African rifts are characterized by lower velocity structure of the uppermost mantle than that of the surrounding areas.

Teleseismic delay times, Bouguer anomalies and inferred thickness of the African lithosphere

Teleseismic P-wave travel time residuals (delay times) have been determined for 38 African seismograph stations using a total of 104 presumed Russian underground nuclear explosions. The delay times ( T) for seismograph stations situated on Precambrian crust within the interior of Africa, where surface erosion is minimal, appear to be linearly related to both the station elevation ( E) and Bouguer anomaly ( B) such that: T=(1.12±0.32)E−(1.81±0.44)(1) T=−(0.013±0.005)B−(1.81±0.60)(2) where T is in seconds, E in kilometres and B in milligals. Equation 2 has been used to predict delay times more generally using the smoothed Bouguer anomaly map of Africa. The delay time map is tentatively used to derive a map of lithosphere thickness which takes into account lateral variations of velocity within both the lithosphere and asthenosphere. The lithospheric thickness model indicates a major zone of thin lithosphere (or thick asthenosphere) is associated with the East African Rift system and its continuation into southern Africa. Concomitant seismicity and volcanism further suggest that incipient separation of the continental plate is taking place along the axis of this zone.

A dynamic model of the Hellenic Arc deduced from geophysical data

Combined gravity and seismic data from Greece and the adjacent areas have been used to explain the high seismicity and tectonic activity of this area. Computed 2-D gravity models revealed that below the Aegean region a large “plume” of hot upper-mantle material is rising, causing strong attenuation of the crust. The hot “plume” extends to the base of the lithosphere and has very probably been mobilized through compressional processes that forced the lithosphere to sink into the asthenosphere. The above model is supported by: high heat flow in the Aegean region; low velocity of the compressional waves of 7.7 km/sec for the upper mantle; lower density than normal extending to the base of the lithosphere; teleseismic P-wave travel-time residuals of the order of +2 sec for seismic events recorded at the Greek seismic stations; volcanics in the Aegean area with a chemical composition which can be explained by assuming an assimilation of oceanic crust by the upper mantle; deep seismicity (200 km) which has been interpreted by various authors as a Benioff zone.

Teleseismic p-wave traveltime residuals and deep structure of the Aegean region

Teleseismic P-wave traveltime residuals have been measured at the Greek seismic stations with respect to the Herrin 68 tables. In spite of the large scatter, some insight into crustal and upper-mantle structure of the Aegean region can be gained. The average absolute residuals (observed minus Herrin traveltimes) are of the order of + 2 s. The most plausible interpretation is an efficient low-velocity zone in the upper mantle. Simple estimates of densities and subsequently gravity with the aid of Birch's law suggest that the Aegean region is underlain by hot expanded upper mantle, perhaps involving partial melting. The relative P residuals (observed minus Herrin traveltime differences between a station and Athens) are generally positive and can be interpreted with lateral variations of the LVZ or of the crust. The latter interpretation is supported in some cases by seismic refraction data. The azimuthal variation of the relative residuals at stations on the non-volcanic arc bears a distinct relation with the arc orientation. At Archangelos (Rhodes) where we “see” through the Benioff zone, the residuals from N to W are between -1 and -2 s and indicate a high-velocity slab sinking below the Aegean sea. At Vamos (Crete), Valsamata (Kephallenia), and Joanina (Pindus Range) the largest (smallest) residuals are along directions parallel (perpendicular) to the arc. This can be interpreted by crustal thickening under the sedimentary arc and/or by velocity anisotropy with the maximum perpendicular to the arc. On the whole, our study supports the hypothesis that the Aegean region is a trench—island-arc—marginal-sea system.