A seismic swarm which occurred in December 1994 near the Aoba volcano, Vanuatu, is investigated in detail by analyzing local and regional broad band seismic records. The seismic area is situated in a poorly instrumented region, although major seismic and volcanic events are possible. The data collected by four permanent short period stations (the nearest one being installed in the Espiritu Santo Island located 80 km from the volcano) allowed us to locate epicenters with a 2-3 km precision, and to document the present local seismicity. The quality of available regional broadband data allows us to determine reliable focal parameters, though the magnitude of the strongest event is 4.8. Data have been selected from the permanent Geoscope and Iris stations located in the south-western Pacific region, and from two temporary networks (one installed in New Caledonia and Vanuatu by I.R.D. and the E.O.S.T, and another one installed in Fiji and Tonga Islands by the Washington University, St Louis, USA). Results obtained either with local (epicentral distance d < 500 km) body waves or regional (500 < d < 2000 km) surface waves lead us to conclude that the swarm is associated with the volcanic edifice, but is not of volcanic type according to the definitions given by Minakami [Phys. Volcanol. (1974) 1] and Schick [Bull. Volcanol. 44 (1981) 491]. The observed mechanisms evidence an important local influence of the volcano itself on the stress field and the observed near horizontal P axis argues against a strong increase of phreatic activity. In addition, seismic and volcanic observations are analyzed in the framework of the interpretation of earthquake swarms occurring in the vicinity of active volcanoes. (C) 2001 Elsevier Science B.V. All rights reserved.

This paper exposes the various magnitude and focal mechanism computations done for the Saint-Beat (Pyrenees) earthquake of 4 October 1999 at 18:14 UT. The focal mechanisms have been obtained either by regional moment tensor inversion, for which the procedure is briefly described, or from first motion polarities. Although there is a disagreement about the direction of dip of the preferred fault plane, this one is almost sub-vertical and seems to exhibit a global east-west orientation, with a T axis oriented north to NNE and a P axis oriented SSE to south-east. Fire also deduced from the inversion a seismic moment of 4.3.10(15) N.m, corresponding to a Mu' magnitude of 4.4. This one equals the m(b) magnitude reported by the US Geological Survey and is between the extreme magnitudes Msz 3.7 and M-L 4.8 reported by the ReNaSS. (C) 2000 Academie des sciences / Editions scientifiques et medicales Elsevier SAS.

It is in theory possible to solve a full moment tensor from inversion of a few seismograms, using normal-mode data, surface waves or body waves. In fact, the isotropic component is usually set to zero in many inversions, in order to stabilize them. This approximation may be considered valid for tectonic earthquakes, but for other applications (such as the study of nuclear or volcanic explosions, deep earthquakes and induced seismicity), the determination of the volumetric component is a key point of the inversion. Our aim is to investigate under which practical conditions the determination of the isotropic component is feasible, and is mathematically and physically reliable. In the first part, we examine the question from a physical point of view and show that the classical interpretation of a full moment tensor for tectonic events implies rheological constraints that are not always realistic. We therefore propose an extended physical model which includes tectonic and non-tectonic volumetric variations. In the second part, we use the tools of inverse theory to infer mathematical constraints on the problem of full moment tensor inversions, from teleseimic surfacewave or body-wave spectra. In particular, we examine how much of the moment tensor can be solved, in relation to the eigenvalues, the condition number and the sampling of the inverse problem. In addition, the resolution and the correlation matrices show that, among a choice of possible constraints on the full tensor, a constraint on the isotropic component is most valuable. In the third part, we also show some applications of our theoretical developments to regional waveform inversions, using the 1992 April Roermond, the Netherlands, earthquake. In addition to physically reliable estimations of the tectonic and non-tectonic isotropic components in full moment tensor inversions, we finally propose extensions of the basic linear methods that can lead to particular models in subspaces of interest, such as tectonic models, or decompositions in a doublecouple plus a volumetric part. By revisiting carefully the determination and interpretation of moment tensors, we provide new perspectives in the estimation of the model and of its error, for a more flexible tectonic and physical interpretation of source mechanisms.

The Mw = 5.4 Roermond earthquake of April 13, 1992, is used as a 'test' earthquake for the development of source inversion methods at a regional scale in Europe. We combine structural modelling of the European continent (Du et al., 1997) with two source inversion methods derived from Sileny et al. (1992), and Mao et al. (1994). We show that following this strategy, it is possible to fully analyze the inverse problem of the hypocentral relocation, source mechanism and rupture history. We define and discuss our methodology on the basis of the inverse problem and of the associated tools. The results of our application to the Roermond earthquake are discussed in the light of other previously published solutions. Such an approach appears to offer a promising tool for the global description of seismic sources in regions well studied from the structural point of view, through waveform inversion of a few regional records.

Some theoretical and practical limits to linear moment tensor inversion of surface waves are analyzed in detail, in particular when one or few stations are used for rapid determination of source parameters. The theory is briefly outlined and steps of preprocessing, especially corrections for source history and propagation, are discussed in order to guarantee the validity of the moment tensor inversion. The inverse problem is first studied from a theoretical point of view. Then the feasibility of the moment tensor inversion is tested from synthetic computations and the formalism is improved in order to obtain a better system conditioning. Finally, some cases of practical nonuniqueness of the solution are shown and possible restrictions to the applicability of the method are discussed. All steps are illustrated with the example of theM s =6.9 Erzincan (Turkey) earthquake of March, 13, 1992.

The destructive earthquake of December 1988 in Armenia (M(S) = 6.9) was recorded on broad-band and very long-period channels at teleseismic distances by Geoscope and GDSN networks. These records are well distributed in azimuth, and allow a detailed study of the rupture process of this earthquake. The average focal mechanism obtained by P- and SH-wave modelling (phi = 300-degrees +/- 10-degrees, delta = 63-degrees +/- 5-degrees, lambda = 100-degrees +/- 20-degrees) is compatible with the mechanism obtained from very long-period surface waves and intermediate-period single-station determinations, as well as field observations. The mean depth of the rupture is also fixed by waveform modelling between 5 and 7 km which means that the rupture surface extends from the surface to a depth between 10 and 14 km, in agreement with aftershock depth distribution. The rupture is found to be complex, composed of a weak beginning or a small foreshock and two pulses well separated in time. The time delay between the two events is estimated for each station by waveform modelling and by spectral analysis. The azimuthal variation of this time delay is interpreted in terms of direction and velocity of rupture on the fault plane. A more detailed analysis of the source implies the use of additional information coming from aftershock studies and tectonics. We use forward modelling to investigate several rupture mechanisms. A three-source model gives an acceptable fit to the observed records but the western mechanism is at odds with observed tectonics and, furthermore, rupture propagation is not well simulated. A five-segment model of the source obtained from field seismotectonic data gives a better waveform fit, a time sequence of individual breaks that simulates a rupture propagating away from the hypocentre, and subsource mechanisms that are compatible with surface tectonics.