Jean-Claude Lépine

Mount Etna dense array local earthquake P and S tomography and implications for volcanic plumbing

Inversion for the three-dimensional velocity structure of Mount Etna is performed with a data set of arrival times of P and S waves of local earthquakes from temporary dense arrays of three-component seismographs. A high-V p body revealed by the original tomography without nearby stations is confirmed, and its image is sharpened using new velocity constraints provided by refraction data. Synthetic tests of V p and V p /V s , and comparison with an independent artificial source tomography with a fundamentally different geometry consistently calibrate the significance threshold of the resolution indicators. The trustworthy part of the image shows a high-V p body centered under the southern part of Valle del Bove above the 6 km below sea level deep basement, which extends towards sea level and may be rooted in or through the crust. It has a large contrast of over 1 km/s with the surrounding sediments and sharp lateral limits and can thus be regarded as made of intrusive material of magmatic origin. The massive high-V p body is heterogeneous in V p /V s . The regions inside it where V s is relatively low can then be suspected of containing a proportion of melt or be fractured and act as pressure links or transport zones. Such features may be structurally linked and appear to be activated in eruptive phenomena. By taking into account the heterogeneities in structure and physical state retrieved by seismic tomography a succession of seismic events, deformational episodes, and geochemical variation in lavas can be discussed with respect to the well-observed eruptions.

Spatial distribution of hotspot material added to the lithosphere under La Réunion, from wide-angle seismic data

Wide-angle seismic lines recorded by ocean bottom and land seismometers provide a pseudo three-dimensional investigation of the crust and upper mantle structure around the volcanically active hotspot island of La Reunion. The submarine part of the edifice has fairly low seismic velocities, without evidence for intrusives. An upper unit with a velocity-depth gradient is interpreted as made of material erupted subaerially then transported and compacted downslope. Between this unit and the top of the oceanic plate, imaged by normal incidence seismic reflection, a more homogeneous unit indicated by shadow zones on several wide-angle sections may correspond to lavas of a different nature, extruded underwater in the earlier phase of volcanism. Coincident wide angle and normal incidence reflections document that the oceanic plate is not generally downwarping toward the island but doming instead toward its southeastern part, with limited evidence for some intracrustal intrusion. Deeper in the lithosphere, the presence of a layer of intermediate velocity between the crust and mantle is firmly established. It is interpreted as resulting from the advection of hotspot magmatic products, possibly partially molten, and of a composition for which the crust is a density barrier. The extensive wide-angle coverage constrains the extent of this body. It does not show the elongated shape expected from plate drift above a steady hotspot supply. Alternative propositions can hence be considered, for example, that La Reunion is caused by a solitary wave of hotspot material or by a young hotspot. The size of the underplate, 140 km wide and up to 3 km thick, corresponds to less than half the volume of the edifice on top of the plate.

Geodetic evidence for rifting in Afar a brittle-elastic model of the behaviour of the lithosphere

Abstract An important episode of rifting occurred in November 1978 in southwest Afar, in the first subaerial section of the accreting plate boundary between the African and Arabian plates. Horizontal rifting of more than 2 m took place, with vertical displacements of about 1 m, earthquakes of magnitude up to 5.3, and a fissural volcanic eruption of basaltic lavas. Very precise geodetic measurements were carried out in order to study this crisis and strains of the order of 3 × 10−4 were measured, both tensile and compressive. This paper presents an analysis of the mechanical behaviour of the lithosphere. It is shown that an elastic-brittle model with a rebound mechanism fits very well the data, and it is suggested that such a model, with magma injection in the resulting open fissures, should be used to describe accretion at plate boundaries.

Modes of raising northeastern Tibet probed by explosion seismology

Abstract New wide-angle reflection and refraction seismic data provide constraints on the structure of the upper lithosphere, and test models of its evolution to raise the northeastern part of the Tibetan Plateau. Amplitudes observed for reflections from the crust–mantle boundary are sufficiently large to suggest that there is no significant partial melt in the deep crust. The data show an increase of the crustal thickness between terranes from north of the Kun Lun Fault into the Qang Tang of central Tibet, and a contrast among their intracrustal images and compositions. In the north, P and S velocities are consistent with a dominantly felsic composition and show that only the upper crust thickened. South of the Kun Lun Fault a thicker crust made of two layers could result from the superposition of the originally thin crust of the Bayan Har terrane on the lower part of the crust of the domain to the north, which upper crust it shoved and thickened. Different modes of crustal thickening, either by thickening of individual layers or superpositions and imbrication among them appear to work jointly to raise the topography.

Seismotectonics in the Republic of Djibouti, linking the Afar Depression and the Gulf of Aden

Abstract The Arta seismological array around the Gulf of Tadjurah monitored earthquake activity for over 15 years. Epicentral locations and focal mechanisms are compared with tectonic evidence from surface breaks, traces of active faults and the spatial distribution of swarms of aftershocks. Strike-slip motion on near vertical faults predominates. In the Gulf, left-lateral strike-slip on N150°E oriented faults is suggested, at variance with the emphasis in previous models on either: (1) a continuous plate boundary with spreading centres connected by ENE trending right-lateral transform faults; (2) a propagating rift oriented approximately east-west or (3) a few broad zones of extension between N150°E normal faults. Left-lateral, strike-slip motion occurring on a series of parallel faults results in clockwise rotation of the blocks between the faults. The type of deformation has similarities to that further west, but is not obviously controlled by a distinct pair of overlapping rifts, as proposed inland by Tapponnier et al. (1990). It is suggested that the stresses induced by the northeasterly drift of Arabia, alone, may deform the southern part of the more easterly oriented zone of lithosphere, derived of plume and ridge, which separates Arabia from Africa.

Episodes of pit-crater collapse documented by seismology at Piton de la Fournaise

Calderas are oval fault-bounded depressions in volcanic regions. The formation of a small analog, a summit pit-crater, has followed a ten-month cycle of eruptive activity at Piton de la Fournaise volcano, Reunion Island, Indian Ocean. In 1985, six short episodes of summit seismic activity, with shocks indicating shear failures, have been followed by eruptions in five cases. The eruptions occurred near the base of the summit cone which is at 2.5 km elevation, inside the Enclos Fouque, an ancient caldera 7 km in diameter. In another instance, the summit seismic crisis did not evolve to an eruption but was followed by a seismic crisis 4 km deeper. In March 1986, low-level activity both at shallow depth beneath the summit and around 4 km depth to the southeast, a simultaneous occurrence not previously observed, culminated by a summital swarm on March 18. Activity resumed its low level and five hours later a small and short-lived eruption occurred at mid-distance between the summit and the rim of Enclos Fouque. One day later a flank eruption began at 1000 m elevation, outside of the Enclos, 9 km southeast of the summit. Four days later, effusion began near sea level 14 km southeast of the summit. The shift of activity from the vent inside to that outside the Enclos occurred whilst both shallow summital and deeper flank seismicity were sustained but without notable shallow activity between one vent and the next. Then the seismic activity, which in all five previous instances stopped when the vents opened, persisted for ten days increasing in energy release rate. Activity only consisted of shallow shocks in the summit area whilst eruption was low on the flank, 9 km, then 14 km away. The focal mechanisms are different from all those previously determined. They correspond neither to shear failure nor tensile fracturing by fluid overpressure. They suggest cracking due to volume increase in a low fluid-flow environment which may be consistent with the simultaneous tapping of magma deeper in the edifice. A final paroxysmal seismic crisis lasting one and a half days led to the collapse of a 200-m diameter, 80-m-deep pit-crater inside the Dolomieu summital caldera on March 29. This crisis released an order of magnitude more energy than previous activity. The long-period signature of these seismic events indicates low stress drop with respect to source dimensions. Sources are shallow. However, rather than directly marking the progressive development of a depression at the surface, they appear to correspond to fractures opening above the retreating magma and progressive stoping of a cavity. Similarities with seismic observations during the last phase of the Fernandina, Galapagos, 8-km2 caldera collapse of June 1968 suggest that the same kind of mechanical behaviour occurred with a scaling of three orders of magnitude in seismic moments and in volumes of collapsed blocks between the two cases.

Complex images of Moho and variation of Vp/Vs across the Himalaya and South Tibet, from a joint receiver‐function and wide‐angle‐reflection approach

[1] Teleseismic receiver functions (RF) allow us to image the spatial variation of the crust-mantle boundary (Moho) along a tight array spanning from south of the Himalayas to the centre of the Tibetan Plateau. This approach is crosstested with wide-angle reflection imaging (WARR). Highlighted by each of the two independent methods, a complex architecture of the Moho with dipping and overlapping segments indicating lithospheric imbrication, is confirmed. The joint use of the two methods reveals an increase of the average crustal P-to-S-wave-velocity ratio from south to the centre of the Lhasa block. This may be due to lowered S-wave velocity confined in specific layers, that may be interpreted as partial melt. This accounts for half of the relative increase in the delay of direct teleseismic S-wave arrivals with respect to P-wave arrivals from south to north, suggesting a similar anomaly in the shallower mantle. INDEX TERMS: 0935 Exploration Geophysics: Seismic methods (3025); 7218 Seismology: Lithosphere and upper mantle; 9320 Information Related to Geographic Region: Asia. Citation: Galve ´, A., M. Sapin, A. Hirn, J. Diaz, J.-C.

Triple junction and ridge hotspots: Earthquakes, faults, and volcanism in Afar, the Azores, and Iceland

The Afar depression of the Horn of Africa, the Azores Archipelago, and Iceland are zones of plate divergence and mantle plume discharge. Strike-slip focal mechanisms dominate in major earthquakes. Among nodal planes a particular fault plane can be identified by precise epicentral location, from the strike of the aftershock zone, or by relation to field evidence of active faulting. These fault planes are orthogonal to the directions expected in previous models for transform faults along a continuous divergent plate boundary. Recent volcanic features are consistently at small angles with the fault planes. Rotation of blocks within a deforming zone is implied by the direction of fault planes, the sense of motion along them, and their location on a series of parallel features transverse to the deformed zone. The drift of oceanic or continental large lithospheric plates deforms a zone of lithosphere, deriving from both the plume discharge and ridge accretion while the zone is created.

Geodetic Results in Afar: The Rifting Episode of November 1978 in the Asal-Ghoubbet Rift

Abstract A seismo-tectonic and volcanic crisis occurred in November 1978 in the Asal-Ghoubbet rift, first subaerial section of the accreting plate boundary between the African and Arabian plates (Allard et al., 1979; Abdallah et al., 1979; Le Dain et al., 1980). The activity was located in the center of a geodetic network set up in the winter 1972–1973 by the Institut Geographique National in collaboration with the Institut de Physique du Globe de Paris. Simultaneously, a precise levelling line of about 100 km was established across the area (I.G.N., 1973). The resurveying of both the geodetic network and the levelling line was carried out after the crisis, between November 1978 and March 1979. Extensions up to 2.4 m and vertical displacements up to 0.7 m were measured. Operating techniques and results of the resurveying are described in Kasser et al. (1979) and Ruegg et al. (1979). Figure 1 shows the horizontal displacements (relating to point B and to the direction BT) and figure 2 shows the vertical displacements relating to the two external points. Tarantola et al. (1979, 1980) have shown that these results can be geodynamically interpreted by a mechanism of sudden breaking and elastic rebound after an elastic stretching of the crust due to the relative drift of the plates. The breaking is triggered by magmatic fracturing of the crust, with dykes injection from a magmatic chamber which has fed the basaltic fissurai eruption. The horizontal and vertical displacements outside the broken zone of the Inner Floor are predicted by a numerical model based on this interpretation which fit very well the experimental data.