Search

The first meters of the underground media, the Critical Zone, involve many human activities. It is also the site of disturbances due to climate change through variations in the level of the water table, hydric and mechanical modifications that increase erosion and the risks of damage or even collapse. Therefore, monitoring of the subsurface and its anthropogenic structures is currently a major challenge. Therefore, geophysical methods, especially seismic methods, which allow to evaluate the mechanical parameters of the environment, are interesting. Among them, the MASW technique based on Surface Seismic Waves estimates a 1D-profile of S wave velocities. The inverted data corresponds to the dispersion of the phase velocity (V ph ) or group velocity. However, the accuracy of the MASW inverse solutions does not discriminate Vs variations lower than a few percent. These limits are due to the current level of measurements uncertainty but also to the sensitivity of the data to the medium parameters. To overcome this key issue, we propose three methodological alternatives. The first proposes to invert the frequency derivative of Vph . The two others are based on the inversion of differential data. In the first one, it is the statistical distance between the main lobes of dispersion diagrams assimilated to histograms. In the second, it is an analytical formulation by linear approximation of Vph . These proposals show a contribution of the resolution for typical two-layer models of the subsurface, based on synthetic data and laboratory measurements by laser interferometry on resin and concrete models., Les premiers mètres du sous-sol, la Zone Critique, concentrent de nombreuses activités humaines. La ZC est aussi le siège de perturbations en raison du changement climatique à travers des variations du niveau des nappes, de modifications hydriques et mécaniques qui augmentent l'érosion et les risques de dommages voire d'effondrements. Aussi, la surveillance de la subsurface et de ses structures anthropiques est actuellement un enjeu majeur. De ce fait, les méthodes géophysiques, notamment sismiques permettant d'évaluer les paramètres mécaniques du milieu sont intéressantes. Parmi elles, la technique MASWbasée sur les Ondes Sismiques de Surface estime un profil 1D de vitesses d'ondes S. Ladonnée inversée correspond à la dispersion de la vitesse de phase (Vph ) ou de groupe.Cependant, la précision des solutions inverses par MASW ne discrimine pas les variations de Vs de quelques pourcents. Ces limites sont dues auniveau d'incertitude des mesures mais aussi à la sensibilité de la donnée aux paramètres du milieu.Pour lever ce verrou, nous proposons trois alternatives méthodologiques. La premièrepropose d'inverser la dérivée fréquentielle de Vph . Les deux autres reposent sur l'inversion de données différentielles. Dans la première, il s'agit de la distance statistique entre les lobes principaux des diagrammes de dispersion assimilés à des histogrammes. Dans la deuxième, il s'agit de formulation analytique parapproximation linéaire de Vph . Ces propositions sont appliquées à des modèles bicouches, tant sur des données synthétiques que sur des mesures de laboratoire sur du béton et par interférométrie laser sur des modèles en résines.

The first meters of the underground media, the Critical Zone, involve many human activities. It is also the site of disturbances due to climate change through variations in the level of the water table, hydric and mechanical modifications that increase erosion and the risks of damage or even collapse. Therefore, monitoring of the subsurface and its anthropogenic structures is currently a major challenge. Therefore, geophysical methods, especially seismic methods, which allow to evaluate the mechanical parameters of the environment, are interesting. Among them, the MASW technique based on Surface Seismic Waves estimates a 1D-profile of S wave velocities. The inverted data corresponds to the dispersion of the phase velocity (V ph ) or group velocity. However, the accuracy of the MASW inverse solutions does not discriminate Vs variations lower than a few percent. These limits are due to the current level of measurements uncertainty but also to the sensitivity of the data to the medium parameters. To overcome this key issue, we propose three methodological alternatives. The first proposes to invert the frequency derivative of Vph . The two others are based on the inversion of differential data. In the first one, it is the statistical distance between the main lobes of dispersion diagrams assimilated to histograms. In the second, it is an analytical formulation by linear approximation of Vph . These proposals show a contribution of the resolution for typical two-layer models of the subsurface, based on synthetic data and laboratory measurements by laser interferometry on resin and concrete models.; Les premiers mètres du sous-sol, la Zone Critique, concentrent de nombreuses activités humaines. La ZC est aussi le siège de perturbations en raison du changement climatique à travers des variations du niveau des nappes, de modifications hydriques et mécaniques qui augmentent l'érosion et les risques de dommages voire d'effondrements. Aussi, la surveillance de la subsurface et de ses structures anthropiques est actuellement un enjeu majeur. De ce fait, les méthodes géophysiques, notamment sismiques permettant d'évaluer les paramètres mécaniques du milieu sont intéressantes. Parmi elles, la technique MASWbasée sur les Ondes Sismiques de Surface estime un profil 1D de vitesses d'ondes S. Ladonnée inversée correspond à la dispersion de la vitesse de phase (Vph ) ou de groupe.Cependant, la précision des solutions inverses par MASW ne discrimine pas les variations de Vs de quelques pourcents. Ces limites sont dues auniveau d'incertitude des mesures mais aussi à la sensibilité de la donnée aux paramètres du milieu.Pour lever ce verrou, nous proposons trois alternatives méthodologiques. La premièrepropose d'inverser la dérivée fréquentielle de Vph . Les deux autres reposent sur l'inversion de données différentielles. Dans la première, il s'agit de la distance statistique entre les lobes principaux des diagrammes de dispersion assimilés à des histogrammes. Dans la deuxième, il s'agit de formulation analytique parapproximation linéaire de Vph . Ces propositions sont appliquées à des modèles bicouches, tant sur des données synthétiques que sur des mesures de laboratoire sur du béton et par interférométrie laser sur des modèles en résines.

A nonlinear acoustic method to assess the damage level of a complex medium is discussed herein. Thanks to the highly nonlinear elastic signatures of cracks or, more generally, internal solid contacts, this method is able to distinguish between contributions from linear wave scattering by a heterogeneity and contributions from nonlinear scattering by a crack or unbounded interface. The coda wave interferometry (CWI) technique is applied to reverberated and scattered waves in glass plate samples featuring various levels of damage. The ultrasonic coda signals are recorded in both the absence and presence of an independent and lower-frequency elastic “pump” wave, before being analyzed by CWI. The monitored CWI parameters quantifying changes in these coda signals, which therefore quantify the nonlinear wave-mixing effects between the coda and pump waves, are found to be dependent on the damage level in the sample. A parametric study is also performed to analyze the influence of sensor positions and average temperature on the method’s output. The reported results could be applied to the non-destructive testing and evaluation of complex-shape materials and multiple scattering samples, for which conventional ultrasonic methods show strong limitations.

The influence of pore water pressure has to be taken into account when the propagation of seismic waves in saturated soils is simulated, because it controls the strength decrease during the dynamic process and can induce phenomena such as cyclic mobility and liquefaction. The decrease of effective stresses leads to permanent deformation in the soil and damages to the surface structures. This research aims at validating a 1-directional propagation model of 3-component seismic waves (1D-3C) in a horizontally layered soil, in the case of saturated soil. A 3-D non-linear elasto-plastic model for soils is used and the variation of the shear modulus reduction curve with the pore water pressure is taken into account during the seismic event. In the case of multiaxial stress states induced by a 3C seismic motion, if no additional information is available, the modulus reduction curves in shear and compression, normalized with respect to the elastic moduli, are assumed identical, since the Poisson's ratio is supposed to be constant during the process. The multiaxial stress state in saturated soils induced by a 3C seismic motion is analysed for various stress paths. The 1D-3C wave propagation model is used to compute the seismic response of the Kushiro soil profile (Japan). The stratigraphy and geotechnical properties are deduced from in situ and laboratory tests provided by the PRENOLIN benchmark. Numerical results are in correct agreement with the records at the surface in terms of acceleration envelope

Site effects may be assessed using a standard soil classification parameter, VS30 (the harmonic average shear-wave velocity in the first 30 m); however, this index does not account for the complexity of the velocity profile, especially its variability at depth. In the present study, in addition to VS30, we propose consideration of the gradient of the VS profile from 0 to 30 m depth, denoted B30. A lower gradient value means low velocity increases with depth; a higher gradient indicates a rapid velocity increase with depth in the shallow layers. In addition, we consider the fundamental resonance frequency of the soil (f0), which has been shown to be a relevant parameter for site-effect assessment and which is obtained from the empirical site response. Using the Japanese KiK-net database, we analyze the variability of the VS profiles and the empirical borehole site responses of selected sites through the VS30, the velocity gradient B30, and f0. We select 289 sites for which the 1D linear numerical modeling is close to the empirical site response and a VS at the downhole station is greater than 1000 m=s. For a given VS30 class, B30, and f0 can be used to distinguish between two types of sites: deep sedimentary sites and sites with high velocity contrast at shallow depths. We find that, even if the gradient is calculated using shallow information, its use improves the site amplification characterization, compared to using only VS30, by reducing the intersite site-response variability. As expected, however, this improvement is limited for deep sedimentary sites. On the other hand, f0 is able to reduce the intersite response variability for deep sedimentary sites even though it is limited to specific VS30 classes. Thus, the combined use of VS30, B30, and f0 improves the assessment of linear site amplification.

In this article, we compare field estimates of near-surface attenuation, as captured by site-specific o-values (i.e., o0) with laboratory-based estimates of minimum shear-strain damping (Ÿmin). We propose models for Ÿmin based on o0 measured at selected stations of the KiK-net database, which are found to be generally larger than lowstrain damping values obtained from laboratory testing. The latter can only quantify intrinsic material damping, whereas other attenuation mechanisms such as scattering of the wavefield contribute to field-based estimates. In addition, we evaluate the difference in damping at the surface and at borehole stations to determine the contribution of shallow layers to attenuation as captured by o0-values at the surface. Thus, values of o0 are computed at the surface and at the downhole instrument depth. The difference between both values, correlates well with the averaged shear-wave velocity over the top 30 m of the profile, VS30, and with the depth to bedrock. Estimates of o0 for hard-rock and stiff sites in Japan are also examined and compared with other regional o0-values proposed for high VS30 materials in New Zealand, Greece, and Switzerland. Two values of o0, which are lower than the corresponding estimates for the aforementioned regions, are deemed potential descriptors of hard-rock conditions in Japan. The ability of the proposed o0-consistent damping models to predict ground motions using the vertical array data from the KiK-net sites has yet to be tested.

The analysis of the seismic wavefield is an essential pre-requisite to the study of seismic wave propagation which in turns helps improving our understanding of the physical processes behind the sources and our knowledge of the propagation medium. The objective of this thesis is to further develop signal processing techniques to more fully exploit the information brought by the 3 component stations and arrays in order to characterize the seismic wavefield. The thesis work focuses on polarization analysis, its extension to 3-component arrays and its joint use with classical array processing. A review of the existing methods that attempt to extend array processing to the 3-components leads to the observation that these methods are complex and their use is limited. Therefore, two alternative methods that associate array processing and polarization sequentially are suggested. In order to best exploit the polarization analyses, a standardized parametrization system describing the polarization is developed and associated with a visualization solution regrouping all the parameters necessary for the interpretation on one figure. Finally, a polarization analysis performed on data from the LSBB 3-component array demonstrates the possibility to use spatial coherency to assist with the interpretation of seismograms, L’analyse du champ d’ondes est un prérequis essentiel à l’étude de la propagation des ondes sismiques qui permet à son tour d’améliorer notre compréhension des processus physiques liés à la nature de la source et notre connaissance des milieux de propagation. L’objectif de cette thèse est de développer des techniques de traitement du signal afin d’améliorer l’exploitation des informations apportées par les stations et les antennes 3 composantes dans le but de caractériser le champ d’ondes sismiques. Elle se concentre sur les analyses de polarisation, leur extension aux antennes 3 composantes et leur utilisation conjointe avec des traitements d’antenne classiques. La thèse revient sur les approches existantes qui tentent d’étendre le traitement d’antenne aux 3 composantes. Ces méthodes existantes se montrent complexes et leur utilisation reste limitée, la thèse suggère deux méthodes alternatives associant successivement traitement d’antenne et polarisation. Afin d’exploiter au mieux les analyses de polarisation, un système standardisé de paramètres décrivant la polarisation est développé et associé à une solution de visualisation permettant de regrouper l’ensemble des paramètres essentiels à l’interprétation sur une figure unique. Finalement, une étude de polarisation sur l’antenne 3 composantes du LSBB (Laboratoire Souterrain Bas Bruit) démontre la possibilité d’utiliser la cohérence spatiale de la polarisation comme aide pour l’interprétation des sismogrammes

Numerical modeling of seismic wave propagation has been a major topic on ground motion studies using a number of different numerical integration schemes. The consideration of soil nonlinearity holds an important place in order to achieve simulations consistent with real observations for strong seismic shaking. Additionally, in the presence of strong ground motion in saturated soils, pore pressure becomes an important parameter to take into account for related phenomena such as flow liquefaction and cyclic mobility. In this study, first, one component (1C) - seismic wave propagation is modeled in linear and nonlinear media in 1D based on the spectral element numerical method. Viscoelastic and nonlinear soil rheologies are implemented by use of the memory variables technique and Iwan’s elastoplastic model, respectively. Then, the same study is extended to a 1D - three component (3C) model and a preliminary comparison on the effect of using 1C and 3C approaches is made. Then, the influence of excess pore pressure development is included in the 1D-3C model and the developped numerical model is applied to realistic case on the site of Wildlife Refuge Liquefaction Array (USA) which is affected by the 1987 Superstition Hills event. The ground motion modification for different assumptions of the soil rheology in the media and different input motions is studied. The calculated motion is found to be amplified on low frequency and damped in high frequency range due to excess pore pressure development. Furthermore, the soil is found to be more nonlinear under triaxial loading in 3C approach and more dilative due to higher nonlinearity. Despite the similitude in surface acceleration and velocity results, significant differences in surface displacement results of 1C and 3C approaches are remarked. Similar analyses are performed on two Japanese sites Kushiro Port and Onahama Port, which are influenced by the 1993 Kushiro-Oki and the 2011 off the Pacific coast of Tohoku earthquakes, respectively. It has been shown that the nonlinearity-related changes are not homogeneous all over the concerned frequency band and the influence of cohesionless soil behavior on wave propagation is highly dependent on model properties and loadingconditions. Lastly, the 2D SEM code is developped by taking into account soil nonlinearity and pore pressure effects similary to 1D-3C SEM code. The developped 2D SEM code is applied to a 2D sedimentary basin site where the basin geometry is asymmetrical and soil profile consists of layers with different nonlinearity properties. Total and effective stress analyses are performed on the 2D basin for P-SV and SH zave propagation models. The calculated surface motion is shown to differ significantly under synthetic and realistic input motion loading conditions and the resultant deformation in superficial layers is found to be very high in effective stress analysis compared to total stress analysis. Also, wave propagation takes longer time inside basin media and the reflections on bedrock-basin boundaries lead the soil in basin edges to higher nonlinearity. This study shows the possibility of modeling nonlinear soil behavior including pore pressure effects in seismic wave propagation studies by coupling different models with spectral element method. These analyses help identifying and understanding dominant phenomena occurring in superficial layers, depending on local conditions and input motions. This is of great importance for site-specific studies, La modélisation numérique de la propagation des ondes sismiques fait partie des études principales sur le calcul du mouvement sismique basées sur de différents schèmes numériques. La prise en compte du comportement nonlinéaire du sol est consideré désormais très important afin de pouvoir calculer la réponse du milieu cohéremment aux observations sous les sollications sismiques très fortes. En plus, le paramètre de pression interstitielle, qui pourrait emmener le sol aux phénomènes de liquéfaction, devient très important pour les sols saturés. Dans cette étude, dans un premier temps, la propagation des ondes sismiques est modelisée sur une composante (1C) dans les milieux linéaires et nonlinéaires en utilisant la méthode numérique des éléments spéctraux. Les rhéologies viscoélastique et nonlinéaire sont implementées par le méthode de technique des variables de mémoire et le modèle élastoplastique d’Iwan, respectivement. Ensuite, le modèle 1D - trois composantes (3C) est développé et une comparaison préalable sur l’effet de la considération des approches 1C et 3C est faite. L’effet de pression interstitielle est implementé dans le code 1D-3C et le site américain Wildlife Refuge Liquefaction Array (WRLA), qui a été frappé par le séisme de Superstition Hills en 1987 est étudié. Le changement de la réponse du sol sous les différents hypothèses de rhéologie du sol et de mouvement d’entrée est étudié. Le mouvement calculé est noté d’être amplifié pour les basses fréquences et atténué pour les hautes fréquences en raison de l’excès de pression interstitielle dans les sols liquéfiables. Par ailleurs, le sol devient plus nonlinéaire sous le chargement triaxial dans l’approche 3C et plus dilatant dû à la nonlinéarite élevée. En dépit de la similitude entre les accélérations et les vitesses en surface des approches 1C et 3C, une importante différence dans le déplacement en surface entre les deux approches est notée. Les analyses sont répétées pour deux sites japonais Kushiro Port et Onahama Port, qui ont été influencés par le séisme de Kushiro-Oki en 1993 et le séisme de la côte Pacifique de Tohoku en 2011, respectivement. Il a été montré que les changements apportés par la nonlinéarite ne sont pas identiques dans toute la gamme de fréquence concernée et l’influence du comportement des sols non-cohésives sur la propagation des ondes sismiques dépend fortement des propriétés du modèle et des conditions de chargement. Dernièrement, le code SEM est avancé en 2D en considerant les mêmes modèles implementés en 1D-3C pour la nonlinéarite du sol et les effets de pression interstitielle. Le code SEM 2D est mis en application dans un modèle de bassin sédimentaire dont la géometrie est assymmétrique et le profile du sol est composé des couches possédant différentes propriétés nonlinéaires. Le modèle est étudié par les analyses totale et effective pour les propagation des ondes P-SV et SH. La differentiation du mouvement calculé en surface est très importante sous les chargements avec les signaux d’entrée synthétique et réel. L’analyse effective résulte en plus de déformations dans les couches superficielles par rapport à l’analyse totale.De plus, la durée de propagation des ondes augmente à l’intérieur du bassin et les reflections aux frontières de bassin-rocher entraînent plus de nonlinéarite dans les coins du bassin. Cette thèse révèle la possibilité de la modélisation du comportement nonlinéaire du sol en prenant en compte l’effet de pression interstitielle dans les études de la propagation des ondes sismiques en couplant les modèles différents avec la méthode des éléments spéctreaux. Ces analyses contribuent à l’identification et la compréhension des phénomènes majeures qui se déroulent dans les couches superficielles en respectant les conditions locales et les mouvements d’entrées, ce quirend ce travail très important pour les études spécifiques de sites

The geometrical and mechanical features of alluvial deposits have a major influence on seismic-wave propagation and amplification. However, for alluvial basins located in densely urbanized areas, the surface structures such as buildings could influence seismic-wave propagation near the free surface. In this article, the influence of surface structures on seismic-wave propagation is analyzed numerically in the case of an actual two-dimensional (2D) shallow basin. At a local scale, the vibration of a surface structure can induce a seismic wave field in the surficial soil layers. At the scale of an alluvial basin, the site-city models considered herein show that the city effect can lead to a significant seismic wave-field modification when compared to the free-field case. The coincidence between the fundamental frequencies of the soil layers and eigenfrequencies of the surface structures is a key parameter to investigate site-city interaction. When comparing simplified sitecity models (Kham et al., 2006) to the basin-city model, the influence of the lateral heterogeneities on the site-city interaction is found to be significant. Indeed, the seismic wave field radiated by the city appears to be trapped within the alluvial basin, and specific directivity features are found for this wave field. The influence of site-city interaction on the free-field seismic hazard may then be significant. The effects of the site-city interaction are beneficial in some parts of the city or detrimental in other parts (especially city boundaries). These effects strongly depend on the urban configuration (city heterogeneity, building density, etc.). Finally, the full characterization of the seismic wave field in densely urbanized areas could often raise the need for investigating site-city interaction and consider such parameters as basin and city fundamental frequencies, building density and city arrangement, as well as basin effects combined with the seismic wave field radiated by the city.

Both the genesis and rates of activity of shallow intraplate seismic activity in central Chile are poorly understood, mainly because of the lack of association of seismicity with recognizable fault features at the surface and a poor record of seismic activity. The goal of this work is to detail the characteristics of seismicity that takes place in the western flank of the Andes in central Chile. This region, located less than 100 km from Santiago, has been the site of earthquakes with magnitudes up to 6.9, including several 5+ magnitude shocks in recent years. Because most of the events lie outside the Central Chile Seismic Network, at distances up to 60 km to the east, it is essential to have adequate knowledge of the velocity structure in the Andean region to produce the highest possible quality of epicentral locations. For this, a N–S refraction line, using mining blasts of the Disputada de Las Condes open pit mine, has been acquired. These blasts were detected and recorded as far as 180 km south of the mine. Interpretation of the travel times indicates an upper crustal model consisting of three layers: 2.2-, 6.7-, and 6.1-km thick, overlying a half space; their associated P wave velocities are 4.75–5.0 (gradient), 5.8–6.0 (gradient), 6.2, and 6.6 km/s, respectively. Hypocentral relocation of earthquakes in 1986–2001, using the newly developed velocity model, reveals several regions of concentrated seismicity. One clearly delineates the fault zone and extensions of the strike-slip earthquake that took place in September 1987 at the source of the Cachapoal River. Other regions of activity are near the San Jose volcano, the source of the Maipo River, and two previously recognized lineaments that correspond to the southern extension of the Pocuro fault and Olivares River. A temporary array of seismographs, installed in the high Maipo River (1996) and San Jose volcano (1997) regions, established the hypocentral location of events with errors of less than 1 km. These events are clustered along no particular lineament approximately 25 km away from the San Jose and Maipo volcanoes. Recurrence intervals, based on a frequency magnitude relationship for lower magnitude events, indicate that earthquakes with magnitudes of 4.7 and 7 have a repeat time of 1 and 1200 years, respectively. Focal mechanisms of the two largest events indicate horizontal maximum and minimum compressive stresses with σ1 varying from a NW–SE orientation in the north to E–W at the southern extreme.

International audience; Building high-resolution models of several physical properties of the subsurface by multi-parameter full waveform inversion (FWI) of multi-component data is one of the challenge of seismic imaging for the next decade. The physical properties, which govern propagation of seismic waves in visco-elastic media, are the P and S wavespeeds, density, attenuation and anisotropic parameters. Updating each of these properties is challenging because several parameters of different nature can have a coupled effect on the seismic response for a particular propagation regime (from transmission to reflection). This is generally referred to as trade-off or cross-talk between parameters. Moreover, different parameter classes can have different order of magnitude or physical units and footprints of different strength in the wavefield, which can make the inversion poorly conditioned if it is not properly scaled. These difficulties raise the issue of a suitable parameterization for multi-parameter FWI, where the term parameterization must be understood as a set of independent parameter classes that fully describe the subsurface properties. Many combinations of parameters can be viewed and this choice is not neutral as the parameterization controls the trade-off between parameters and the local resolution with which they can be reconstructed. Once this parameterization is selected, the subset of parameter classes in the parameterization that can be reliably updated during the inversion, must be identified to avoid over-parameterization of the optimization problem. The purpose of this tutorial is to provide a comprehensive overview of the promise, pitfalls and open questions underlying multi-parameter FWI. We first review the main FWI ingredient that controls the trade-off between parameters, namely the radiation pattern of the so-called virtual sources, and some tools for analyzing these trade-offs. Then, we present some illustrative examples of multi-parameter FWI, which should provide some guidelines to choose a suitable parameterization for FWI in visco-acoustic anisotropic media. We conclude by proposing a data-driven and model-driven workflow for visco-elastic anisotropic FWI of multi-component marine data, which has been inspired by a real data case study from Valhall.

43 p., 1 ann.; Cette étude a été réalisée dans le cadre la convention 2011 n° 2100472261 entre Ministère de l'écologie du développement durable et de l'énergie et le BRGM (action C6 " effets topographiques "). L'intérêt du ministère de l'écologie, du développement durable et de l'énergie est d'avoir des premières pistes et/ou recommandations sur l'applicabilité des coefficients d'amplification topographique (noté ST) des Eurocode 8, à savoir, par exemple : - Est-ce que les valeurs de coefficients ST proposées correspondent bien à la majorité des cas que l'on peut rencontrer ? - Y a-t-il un nombre de cas significatifs où l'amplification due aux effets topographiques est sous-estimée ou surestimée ? - Peut-on préciser certains termes qualitatifs tels que " à proximité de la crête ", " largeur de la crête notablement inférieure à la largeur de la base " ? Ce rapport répond à ces questions via une étude paramétrique comportant plus de 500 simulations 2D par la méthode des éléments finis spectraux (via le code numérique du BRGM EFISPEC2D). Cela représente environ 10000 heures de calcul (i.e., 417 jours). Grâce à ces simulations, nous montrons que l'ordre de grandeur des coefficients d'amplification topographique ST des Eurocode 8 est acceptable. Néanmoins des améliorations sont réalisables, à savoir par exemple (cf. chapitre 7 " Conclusions , recommandations et perspectives ") : - Une augmentation d'environ 10 à 20% des valeurs des coefficients d'amplification ST pour des buttes ou pentes ayant des angles de 10 à 25°. - Un ajustement de la dépendance spatiale des coefficients d'amplification ST à proximité des topographies (ex. une butte). Cette étude a été réalisée en tenant compte des conseils avisés de collaborateurs externes présents aux différentes réunions du projet. L'auteur tient à remercier Fabian Bonilla (IFSTTAR), Pierre-Yves Bard (IFSTTAR/ISTerre), Nathalie Glinsky (IFSTTAR/CETE), Céline Gélis (IRSN), Etienne Bertrand (CETE) et Ezio Faccioli (Studio Geotecnico Italiano).

International audience; In most geologic environments, accounting for anisotropy is necessary to perform acoustic full waveform inversion (FWI) of wide-azimuth and wide-aperture seismic data because of the potential dependence of wave speeds on the direction of the wave propagation. In the framework of multiparameter FWI, the subsurface parameterization controls the influence of the different parameter classes on the modeled seismic data as a function of the scattering angle and hence the resolution with which the parameters can be reconstructed and the potential trade-off between different parameters. We have evaluated a numerical procedure based on computation of the scattering patterns of the different parameters to assess the sensitivity of the seismic data to different parameterizations of vertical transverse isotropic media in the acoustic approximation. Among the different categories we have tested, a monoparametric FWI was found for imaging one wave speed with a broad wavenumber content, keeping the Thomsen parameters fixed, which have a small influence on the data relative to the wave speed. This raises the question of the initial information required in the background models of the Thomsen parameters to perform reliable monoparameter FWI. Alternatively, simultaneously inverting the horizontal and vertical wave speeds introduces limited trade-off effects because these wave speeds have significant influence on the data for distinct ranges of scattering angles, while the influence of the Thomsen parameter δ remains weak. With such parameterization, the short-to-intermediate wavelengths of the vertical velocity are updated from the short-to-intermediate scattering angles, while the long-to-intermediate wavelengths of the horizontal velocity are updated from the wide-to-intermediate scattering angles. We concluded that the choice of the subsurface parameterization can be driven by the acquisition geometry, which controls the scattering-angle coverage and hence the resolving power of FWI, and by the accuracy of the available initial FWI models. Read More: http://library.seg.org/doi/abs/10.1190/geo2012-0204.1

The author thanks the discussers for their interest in the original paper and for their valuable comments. Because the original paper is a review paper dealing with various numerical methods, it was not easy to gather and synthesize the numerous results related to the modeling of seismic problems through such methods as FEM, discrete galerkin method (DGM), finite-difference method, and boundary element method (BEM), among others. More detailed results on numerical dispersion for the FEM and spectral element method, absorbing layer methods (e.g., perfectly matched layer), or boundary element approaches are discussed in Semblat and Pecker (2009).

Landslides affect many clay slopes in the world and regularly threaten people in urban areas mountainous. These landslides are characterized by a slow velocity but they may suddenly liquefy and accelerate unexpectedly. The solid-liquid transition on the clay has been studied of Trièves region (French Alps) using rheological experiments. They have shown the yield stress thixotropic behavior with a viscosity bifurcation which can explain the catastrophic fluidization observed in the field. This loss of material stiffness can be followed by a drop in the shear wave velocity (Vs). Inclined plane test and field experiments (Pont-Bourquin landslides in Switzerland) have both shown a precursor drop of Vs indicating that it could be a good proxy for monitoring unstable clay slope., Les glissements de terrain argileux affectent de nombreux versants à travers le monde et menacent régulièrement les activités humaines dans les zones urbanisées montagneuses. Ces glissements sont caractérisés par des cinématiques souvent lentes mais ils peuvent brutalement se liquéfier et accélérer de manière imprévisible. Cette transition solide-liquide a été étudiée sur les argiles de la région du Trièves (Alpes Françaises) à l'aide d'études rhéologiques. Elles ont montré le caractère de fluide à seuil thixotrope avec une bifurcation de viscosité importante lors de la fluidification pouvant expliquer le caractère catastrophique de l'accélération observée sur le terrain. Cette perte de rigidité du matériau peut être observée par une chute de la vitesse des ondes de cisaillement (Vs). Des études réalisées en parallèle à la fois sur un modèle analogique de plan incliné et sur le terrain (glissement de Pont-Bourquin, Suisse) ont permis d'observer une chute de Vs précédent à cette fluidification montrant ainsi que Vs pourrait être un bon proxy pour la surveillance des instabilités de terrain argileux.

International audience; The Earth is an heterogeneous complex media from the mineral composition scale (10−6m) to the global scale ( 106m). The reconstruction of its structure is a quite challenging problem because sampling methodologies are mainly indirect as potential methods (Günther et al., 2006; Rücker et al., 2006), diffusive methods (Cognon, 1971; Druskin & Knizhnerman, 1988; Goldman & Stover, 1983; Hohmann, 1988; Kuo & Cho, 1980; Oristaglio & Hohmann, 1984) or propagation methods (Alterman & Karal, 1968; Bolt & Smith, 1976; Dablain, 1986; Kelly et al., 1976; Levander, 1988; Marfurt, 1984; Virieux, 1986). Seismic waves belong to the last category. We shall concentrate in this chapter on the forward problem which will be at the heart of any inverse problem for imaging the Earth. The forward problem is dedicated to the estimation of seismic wavefields when one knows the medium properties while the inverse problem is devoted to the estimation of medium properties from recorded seismic wavefields.

The assessment of an Excavating Damaged Zone (EDZ) due to underground works is important in the context of radioactive waste disposal within a geological formation. Seismic surveys can pro- vide, in a non-destructive manner, information directly related to the mechanical properties of this damaged zone. However, the classical refraction method is unusable for galleries where a concrete layer caps the top surface. Moreover, other possible seismic methods like surface wave analysis can become complex to implement in such a context since guided waves are present and pass through a high-velocity concrete layer. For these reasons, the present paper will focus on an experimental feasibility study of the MASW (Multi Channel Analysis of Surface Waves) method for assessing an EDZ below a concrete layer in the specific context of the Tournemire experimental platform, which was drilled into an argillaceous formation in the Aveyron department of central France. This facil- ity is operated by IRSN (French institute for Radiological Protection and Nuclear Safety). This experimental study has been conducted in conjunction with a modelling results analysis. During an initial stage, the measurements carried out in situ are analysed and it is demonstrated that a receiv- ing profile longer than 8 m is required to distinguish the various modes. Moreover, the comparison of a hammer source and a sweep signal generated by a vibrator source shows that the latter is more appropriate when building the dispersion diagram, particularly for fundamental and first higher mode curves. A second stage shows that according to the analysis of theoretical curves for profiles with similar seismic velocities of the Tournemire platform medium, the dispersion curves are sensi- tive to EDZ characteristics. In the paper's final part, the dispersion curves obtained from the experimental seismic data are inverted using a global approach. Results of the inversion of the fundamental mode only and of both the fundamental mode and the first higher mode are com- mented upon, evidencing an improved quality for this last approach.

Many major challenges of the twenty- first century in the economic, social, or environmental fields, require a careful characterization of the geological structures of the Earth. Seismics is probably the most widely used technique to explore the environment at large scales. Imaging and characterizing the subsurface from real seismic data however is a di fficult task, and the di fficulty increases since the wave propagation in complex and heterogeneous natural systems is poorly understood. It is therefore essential to identify the geological features, which signi ficantly in fluence the wave propagation in a given confi guration, and to understand their interaction with the seismic waves. To model the wave characteristics one has to account for the complexity of natural environments through which the waves propagate, but also for the contact laws between the geological media, and for the limitations introduced by the seismic waves themselves. This manuscript summarizes the work I have developed essentially on the understanding and on the modeling of the wave/interface interaction using multidisciplinary approaches. Besides basic knowledge of the physical mechanisms underlying the wave propagation in complex media (e.g., the vertical and the lateral seismic resolutions, the de finition of the 3D seismic refl ector, the interface scattering, and the in fluence of the anisotropy on surface waves), applied works dedicated to the oil and gas exploration or to the monitoring of geological storage of CO2 have also been developed., Certains besoins économiques, sociétaux ou environnementaux nécessitent une caractérisation approfondie des structures géologiques du sous-sol. Parmi les différentes méthodes géophysiques, la sismique est sans doute celle qui est le plus souvent privilégiée. Imager et caractériser le sous-sol à partir de données réelles acquises sur le terrain est cependant un problème inverse généralement di fficile à résoudre. Il l'est d'autant plus si le problème direct associé n'a pas été au préalable "maîtrisé" a minima, c'est-à-dire si l'on ne dispose pas d'une modélisation réaliste de la propagation des ondes. Pour cela, il semble essentiel, pour modéliser les données sismiques synthétiques a fin de les rendre comparables aux données réelles, d'identi fier et de comprendre en premier lieu les facteurs géologiques et/ou physiques de premier ordre qui in fluencent la propagation des ondes dans une con figuration donnée. Cela suppose qu'on prenne en compte à la fois les lois de comportement des couches géologiques traversées et les lois de contact entre ces couches, mais aussi les limitations introduites par le système d'observation constitué par les ondes sismiques. Ce mémoire décrit les travaux que j'ai menés essentiellement sur la compréhension et la modélisation de l'interaction ondes/interfaces en me basant sur des approches pluridisciplinaires. Outre l'aspect fondamental d'une meilleure connaissance des mécanismes physiques sous-jacents à la propagation des ondes dans des milieux complexes (e.g., les résolutions sismiques verticale et latérale, la dé finition du ré flecteur sismique 3D, la diffusion d'interface et l'in fluence de l'anisotropie sur les ondes de surface), des aspects plus appliqués, dédiés notamment à l'exploration de gisements énergétiques ou à la surveillance de réservoirs géologiques de stockage du CO2, ont été également développés.

International audience; Here we assess the potential of the visco-acoustic frequency domain full-waveform inversion (FWI) to reconstruct P-wave velocity (VP) and P-wave attenuation factor (Q) from surface onshore seismic data. First, we perform a sensitivity analysis of the FWI based upon a grid search analysis of the misfit function and several synthetic FWI examples using velocity and Q models of increasing complexity. Subsequently, we applied both the acoustic and visco-acoustic FWI to real surface wide-aperture onshore seismic data from the Polish Basin, where a strong attenuation of the seismic data is observed. The sensitivity analysis of the visco-acoustic FWI suggests that the FWI can jointly reconstruct the velocity and the attenuation factor if the signature of the attenuation is sufficiently strong in the data. A synthetic example corresponding to a homogeneous background model with an inclusion shows a reliable reconstruction of VP and Q in the inclusion, when Q is as small as 90 and 50 in the background model and in the inclusion, respectively. A first application of acoustic FWI to real data shows that a heuristic normalization of the data with offset allows us to compensate for the effect of the attenuation in the data and reconstruct a reliable velocity model. Alternatively, we show that visco-acoustic FWI allows us to reconstruct jointly both a reliable velocity model and a Q model from the true-amplitude data. We propose a pragmatical approach based upon seismic modelling and source wavelet estimation to infer the best starting homogeneous Q model for visco-acoustic FWI. We find the source wavelet estimation quite sensitive to the quality of the velocity and attenuation models used for the estimation. For example, source-to-source wavelets are significantly more consistent when computed in the final FWI model than in the initial one. A good kinematic and amplitude match between the early-arriving phases of the real and time-domain synthetic seismograms computed in the final FWI model provides an additional evidence of the reliability of the final FWI model. We find the recovered velocity and attenuation models consistent with the expected lithology and stratigraphy in the study area. We link high-attenuation zones with the increased clay content and the presence of the mineralized fluids.

The main scientific and engineering goal of this book is to deal simultaneously with soil dynamics/vibrations and wave propagation in soils (including seismic waves). These various fields are generally considered separately and the important links between them, both from scientific and practical points of view, are unfortunately not investigated. They are usually considered in separate disciplines such as earthquake geotechnical engineering, civil engineering, mechanics, geophysics, seismology, numerical modelling, etc. The objective of the book is to offer in a single publication an overview of soil dynamics and wave propagation in soils with emphasis on engineering applications. It starts from a wide variety of practical problems (e.g. traffic induced vibrations, dynamic compaction, vibration isolation), then deals with 1D and 2D/3D wave propagation in heterogeneous and attenuating media (with application to laboratory and in situ dynamic characterization of soils), gives an overview of various numerical methods (e.g. FEM, BEM) to simulate wave propagation (including numerical errors, radiation/absorbing conditions, etc) and finally investigates seismic wave propagation and amplification in complex geological structures (e.g. irregular topographies, alluvial deposits).

International audience; In this paper, the authors compare results obtained by semi-analytical and numerical approaches for the dynamic response of a poroviscoelastic soil under transient loads. The behaviour of the medium is governed by complete Biot formalism. The semi-analytical approach is based on Helmholtz decompositions and Fourier transforms, and yields exact solid and fluid displacements in the transformed domain. The numerical approach uses a C++ object oriented programming finite element–finite difference code. Both methods give concurring results. Moreover, influence of viscous coupling on the response of the ground and visualization of the compressional wave of the second kind are discussed

First arrival traveltime tomography aims at inferring a seismic wave propagation velocity model from first arrival traveltimes picked on seismograms. The velocity model inferred can be used directly to perform a structural interpretation of the subsurface or as an initial model for another seismic imaging method. This technique can be applied at different scales from geotechnical studies to seismology through oil exploration. The geophysicist know-how plays an important role in the difficult resolution of the nonlinear and ill-posed tomographic problem. Numerous studies have tried to ease and improve this resolution considering a physical or mathematical approach. Within the scope of this work, we wish to develop a pragmatic approach, i.e. we consider that the tomographic problem should be solved using an interactive algorithm whose tuning parameters are clearly identified. The interactive aspect of the algorithm facilitates the acquisition of the tomographic know-how because it allows performing, within a reasonable time, many simulations for different kinds of parameterization. The goal pursued in this work is the definition of a first arrival traveltime tomography algorithm that fulfills these specifications at best. Conventional first arrival traveltime tomography algorithms do not match our criteria of a pragmatic approach. Indeed, their implementation hardly takes benefit of the parallel architecture of current supercomputers in order to reduce the computation times. Moreover, their practical application implies a complex parameterization due to the resolution of the linear tomographic system. All these practical limitations issue from the original formulation of the algorithm based on a Gauss-Newton minimization method. The idea developed in this work is to formulate the resolution of the tomographic problem using a steepest descent method to overcome all these limitations. The key step of this formulation is the computation of the gradient of the misfit function with respect to the model parameters. We use the adjoint state method and a method based on an a posteriori ray tracing to compute this gradient. These two methods differ from their formulation, respectively nonlinear and linearized, and their implementation. Then, we clearly define the parameterization of the new algorithm and validate on a supercomputer its practical properties that are: a direct and efficient parallelization, a memory requirement independent of the amount of input data and a straightforward implementation. Finally, in order to validate the tomographic behaviour of this new algorithm, in terms of obtained results and stability, we present tomography results for 2-D and 3-D, synthetics and real, marine and land, seismic refraction acquisitions. A great number of simulations have been carried out thanks to the fast execution time of the algorithm, typically few minutes for 2-D simulations.; La tomographie des temps de première arrivée cherche à estimer un modèle de vitesse de propagation des ondes sismiques à partir des temps de première arrivée pointés sur les sismogrammes. Le modèle de vitesse obtenu peut alors permettre une interprétation structurale du milieu ou bien servir de modèle initial pour d'autres traitements de l'imagerie sismique. Les domaines d'application de cette méthode s'étendent, à des échelles différentes, de la géotechnique à la sismologie en passant par la géophysique pétrolière. Le savoir-faire du géophysicien joue un rôle important dans la difficile résolution du problème tomographique non-linéaire et mal posé. De nombreuses recherches ont entrepris de faciliter et d'améliorer cette résolution par des approches mathématique ou physique. Dans le cadre de ce travail, nous souhaitons développer une approche pragmatique, c'est-à-dire que nous considérons que le problème tomographique doit être résolu par un algorithme interactif dont les paramètres de réglage sont clairement définis. L'aspect interactif de l'algorithme facilite l'acquisition du savoir-faire tomographique car il permet de réaliser, dans un temps raisonnable, de nombreuses simulations pour des paramétrisations différentes. Le but poursuivi dans cette thèse est de définir, pour le cas spécifique de la tomographie des temps de première arrivée, un algorithme qui réponde au mieux à ces critères. Les algorithmes de tomographie des temps de première arrivée classiquement mis en oeuvre aujourd'hui ne répondent pas à nos critères d'une approche pragmatique. En effet, leur implémentation ne permet pas d'exploiter l'architecture parallèle des supercalculateurs actuels pour réduire les temps de calcul. De plus, leur mise en oeuvre nécessite une paramétrisation rendue complexe du fait de la résolution du système linéaire tomographique. Toutes ces limitations pratiques sont liées à la formulation même de l'algorithme à partir de la méthode de Gauss- Newton. Cette thèse repose sur l'idée de formuler la résolution du problème tomographique à partir de la méthode de plus grande descente pour s'affranchir de ces limitations. L'étape clé de cette formulation réside dans le calcul du gradient de la fonction coût par rapport aux paramètres du modèle. Nous utilisons la méthode de l'état adjoint et une méthode définie à partir d'un tracé de rais a posteriori pour calculer ce gradient. Ces deux méthodes se distinguent par leur formulation, respectivement non-linéaire et linéarisée, et par leur mise en oeuvre pratique. Nous définissons ensuite clairement la paramétrisation du nouvel algorithme de tomographie et validons sur un supercalculateur ses propriétés pratiques : une parallélisation directe et efficace, une occupation mémoire indépendante du nombre de données observées et une mise en œuvre simple. Finalement, nous présentons des résultats de tomographie pour des acquisitions de type sismique réfraction, 2-D et 3-D, synthétiques et réelles, marines et terrestres, qui valident le bon comportement de l'algorithme, en termes de résultats obtenus et de stabilité. La réalisation d'un grand nombre de simulations a été rendue possible par la rapidité d'exécution de l'algorithme, de l'ordre de quelques minutes en 2-D.

First arrival traveltime tomography aims at inferring a seismic wave propagation velocity model from first arrival traveltimes picked on seismograms. The velocity model inferred can be used directly to perform a structural interpretation of the subsurface or as an initial model for another seismic imaging method. This technique can be applied at different scales from geotechnical studies to seismology through oil exploration. The geophysicist know-how plays an important role in the difficult resolution of the nonlinear and ill-posed tomographic problem. Numerous studies have tried to ease and improve this resolution considering a physical or mathematical approach. Within the scope of this work, we wish to develop a pragmatic approach, i.e. we consider that the tomographic problem should be solved using an interactive algorithm whose tuning parameters are clearly identified. The interactive aspect of the algorithm facilitates the acquisition of the tomographic know-how because it allows performing, within a reasonable time, many simulations for different kinds of parameterization. The goal pursued in this work is the definition of a first arrival traveltime tomography algorithm that fulfills these specifications at best. Conventional first arrival traveltime tomography algorithms do not match our criteria of a pragmatic approach. Indeed, their implementation hardly takes benefit of the parallel architecture of current supercomputers in order to reduce the computation times. Moreover, their practical application implies a complex parameterization due to the resolution of the linear tomographic system. All these practical limitations issue from the original formulation of the algorithm based on a Gauss-Newton minimization method. The idea developed in this work is to formulate the resolution of the tomographic problem using a steepest descent method to overcome all these limitations. The key step of this formulation is the computation of the gradient of the misfit function with respect to the model parameters. We use the adjoint state method and a method based on an a posteriori ray tracing to compute this gradient. These two methods differ from their formulation, respectively nonlinear and linearized, and their implementation. Then, we clearly define the parameterization of the new algorithm and validate on a supercomputer its practical properties that are: a direct and efficient parallelization, a memory requirement independent of the amount of input data and a straightforward implementation. Finally, in order to validate the tomographic behaviour of this new algorithm, in terms of obtained results and stability, we present tomography results for 2-D and 3-D, synthetics and real, marine and land, seismic refraction acquisitions. A great number of simulations have been carried out thanks to the fast execution time of the algorithm, typically few minutes for 2-D simulations.; La tomographie des temps de première arrivée cherche à estimer un modèle de vitesse de propagation des ondes sismiques à partir des temps de première arrivée pointés sur les sismogrammes. Le modèle de vitesse obtenu peut alors permettre une interprétation structurale du milieu ou bien servir de modèle initial pour d'autres traitements de l'imagerie sismique. Les domaines d'application de cette méthode s'étendent, à des échelles différentes, de la géotechnique à la sismologie en passant par la géophysique pétrolière. Le savoir-faire du géophysicien joue un rôle important dans la difficile résolution du problème tomographique non-linéaire et mal posé. De nombreuses recherches ont entrepris de faciliter et d'améliorer cette résolution par des approches mathématique ou physique. Dans le cadre de ce travail, nous souhaitons développer une approche pragmatique, c'est-à-dire que nous considérons que le problème tomographique doit être résolu par un algorithme interactif dont les paramètres de réglage sont clairement définis. L'aspect interactif de l'algorithme facilite l'acquisition du savoir-faire tomographique car il permet de réaliser, dans un temps raisonnable, de nombreuses simulations pour des paramétrisations différentes. Le but poursuivi dans cette thèse est de définir, pour le cas spécifique de la tomographie des temps de première arrivée, un algorithme qui réponde au mieux à ces critères. Les algorithmes de tomographie des temps de première arrivée classiquement mis en oeuvre aujourd'hui ne répondent pas à nos critères d'une approche pragmatique. En effet, leur implémentation ne permet pas d'exploiter l'architecture parallèle des supercalculateurs actuels pour réduire les temps de calcul. De plus, leur mise en oeuvre nécessite une paramétrisation rendue complexe du fait de la résolution du système linéaire tomographique. Toutes ces limitations pratiques sont liées à la formulation même de l'algorithme à partir de la méthode de Gauss- Newton. Cette thèse repose sur l'idée de formuler la résolution du problème tomographique à partir de la méthode de plus grande descente pour s'affranchir de ces limitations. L'étape clé de cette formulation réside dans le calcul du gradient de la fonction coût par rapport aux paramètres du modèle. Nous utilisons la méthode de l'état adjoint et une méthode définie à partir d'un tracé de rais a posteriori pour calculer ce gradient. Ces deux méthodes se distinguent par leur formulation, respectivement non-linéaire et linéarisée, et par leur mise en oeuvre pratique. Nous définissons ensuite clairement la paramétrisation du nouvel algorithme de tomographie et validons sur un supercalculateur ses propriétés pratiques : une parallélisation directe et efficace, une occupation mémoire indépendante du nombre de données observées et une mise en œuvre simple. Finalement, nous présentons des résultats de tomographie pour des acquisitions de type sismique réfraction, 2-D et 3-D, synthétiques et réelles, marines et terrestres, qui valident le bon comportement de l'algorithme, en termes de résultats obtenus et de stabilité. La réalisation d'un grand nombre de simulations a été rendue possible par la rapidité d'exécution de l'algorithme, de l'ordre de quelques minutes en 2-D.

Colloque avec actes et comité de lecture. Internationale.; International audience; La résolution des équations de l'élastodynamique par la méthode des éléments de frontière (BEM) conduit à un système linéaire plein. Des travaux récents sur les équations de Helmholtz et Maxwell ont établi la capacité de la méthode multipôle rapide (FM) à réduire la complexité de la BEM à N log N par itération d'un solveur de type GMRES. Cet article présente la transposition de l'approche FM-BEM à l'élastodynamique 3D dans le domaine fréquentiel. La précision et l'efficacité de la méthode sont illustrées sur des exemples numériques mobilisant jusqu'à N=O(1 000 000) inconnues nodales.

The solution of the elastodynamic equations using integral formulations requires to solve full and non symmetric systems. The use of an iterative solver like GMRES lowers the complexity of the solution (number of operations) to order N2, where N is the number of degrees of freedom (DOFs). The most expensive computational task is the matrix-vector product. For Helmholtz and Maxwell equations, the fast multipole method (FMM) is known to dramatically reduce that cost. This is achieved by (i) using a multipole expansion of the relevant Green's tensor, which allows to reuse element integrals for all collocation points, and (ii) defining a (recursive, multi-level) partition of the region of space enclosing the domain boundary of interest into cubic cells, allowing to optimally cluster influence computations. Moreover, the matrix of the system is not stored with the FMM. So, the complexity (both in CPU time and memory) is found to be N log2 N per iteration. The fast multipole formulation of the boundary element method for 3D elastodynamics in frequency domain is presented in this article. Numerical efficiency and accuracy are assessed on the basis of numerical results obtained for problems having known solutions. Finally, the present FMM-BEMis demonstrated on seismology-oriented examples, namely the study of the diffraction of a plane wave or a point source by a canyon. The influence of the size of the meshed part of the free surface is studied, and computations are performed for nondimensional frequencies higher than those considered in other studies, with which comparisons are made whenever possible.

Les catastrophes récentes ont confirmé la nécessité d'une meilleure compréhension de l'impact des tremblements de terre sur les villes. Dans ce contexte ont été développés des travaux expérimentaux, théoriques et de modélisation numérique visant à mieux prévoir le mode de propagation des ondes sismiques et leurs effets sur le bâti.

The Racha earthquake of 1991 was the strongest of the earthquakes recorded in the Caucasus. It was studied by an international epicentral expedition. Unique data gathered by this expedition included records of aftershocks whose swarm was very intense. A 3-D velocity model is obtained from analysis of these data by the method of local earthquake tomography. The ancient crystalline basement and the sedimentary-volcanic Mesozoic-Cenozoic cover could be identified from cross sections of the P wave field. The eastern and western boundaries of an uplift in the basement of the Dzirulskii Massif are delineated. Linear low velocity heterogeneities correlating with the active Kakheti-Lechkhumi fault zone and two Trans-Caucasian linear fault zones are discovered in the basement. The cloud of aftershock hypocenters is shown to correlate with a reflector coinciding with the cover-basement interface.

Structural and geotechnical engineers have been aware for a long time of soil-structure interaction (SSI) phenomena, which modify the seismic response of massive or tall buildings erected on soft soils. On their side, seismologists have known for a long time that it is not a good idea to install seismological stations close to trees. During the last decades, it has also become clear how large the effects of surface heterogeneities, commonly called 'site effects' (SE, concerning soft soils as well as topographic features), can be. On this basis, it is legitimate to wonder whether a large building on a soft soil can contaminate the ground motion in its immediate vicinity (phenomenon hereafter abbreviated as 'CGMB', Contamination of Ground Motion by Buildings). Going one step further, we may ask about the overall effect of such contamination in a densely urbanized area. The aim of this chapter is to present an overview of the results that substantiate the plausibility of this kind of 'global' interaction between all the buildings of a city and its subsoil, which we call hereafter 'site-city interaction' (SCI).

We propose a procedure for the detection of converted phases in seismograms between the P- and S-arrivals. This procedure has been successfully applied to waveforms of different stations from the 1991 passive experiment around the Aigion area (Greece). Three major converted phases have been detected at two stations on the northern Gulf of Corinth. The first phase has been interpreted as a P-to-S transmitted wave occurring at an interface located at a depth of 3 km in the northern part of the Psaromita Peninsula. The two later phases that have a more complex polarisation might come from deeper interfaces. To cite this article: D. Latorre et al., C. R. Geoscience 336 (2004).

We present a detailed 3-D seismic velocity model for the upper crust, beneath the Erzincan Basin using tomographic inversion applied to the bodywave arrivals of the 1992 March 13 Erzincan earthquake (M-l = 6.9) aftershock sequence. High quality digital data collected by various institutions is used, consisting of 1240 hypocentres with rms errors less than 0.10 s, recorded by 25 stations on average. This provides a total of approximately 21 000 P- and 10 000 S-arrival times. The final solution is well constrained and reveals sharp boundaries bordering the basin. The shape of the basin is found to be deeper on the north and narrower at the eastern border. The velocity contrast related to the basin extends down to a depth of 3 km, but is hardly perceptible at 6 km and completely disappears at 9 km. The thickness of low-velocity unconsolidated soft sediments (V-p < 2000 m s(-1)) is estimated to be shallower than 3 km. The hypocentres were relocated using the inferred 3-D model and correlated with the fault segments that were observed in field studies. In addition to the main segment of the North Anatolian Fault (NAF), on the north of the basin, we identified a secondary faulted zone located along the southern border of the basin and merging with the Pulumur Fault Zone. We observed that this southern zone has a limited extension towards east (similar to 20 km) and therefore played only a minor role in the evolution of the basin. Despite the complex interaction between the North Anatolian Fault, the Ovaci k Fault and the Pulumur Fault Zone, it is clear that the Erzincan Basin has developed predominantly by the pull-apart mechanism. We conclude that the main continuation of NAF to the east is the branch on the north following the Euphrates valley, and constitutes, locally, the primary seismic hazard factor.

A network of 60 seismographs was deployed across the Andes at approximately 23.5°S. The array was centered in the backarc; atop the Puna high plateau in NW Argentina. P and S arrival times of 426 intermediate depth earthquakes were inverted for 1-D velocity structure and hypocentral coordinates. Average velocities and v sub p/v sub s in the crust are low. Average mantle velocities are high but difficult to interpret because of the presence of a fast velocity slab at depth. Although the hypocenters sharply define a 35° dipping Benioff zone, seismicity in the slab is not continuous. The spatial clustering of earthquakes is thought to reflect inherited heterogeneties of the subducted oceanic lithosphere. Additionally, 57 crustal earthquakes were located. Seismicity concentrates in the fold and thrust belt of the foreland and Eastern Cordillera, and along and south of the El Toro-Olacapato-Calama Lineament (TOCL). Focal mechanisms of two earthquakes at this structure exhibit left lateral strike-slip mechanisms similar to the suggested kinematics of the TOCL. We believe that the Puna north of the TOCL behaves like a rigid block with little internal deformation, whereas the area south of the TOCL is weaker and currently deforming. (Résumé d'auteur)

Seismic approaches reveal the lithospheric structure and geodynamic contexts of Piton de la Fournaise and Etna. A set of tomographic experiments constrains their internal structures with a resolution providing insight into edifice evolution and eruptive behavior. The hotspot context of La Réunion island thus appears to differ from the Hawaiian one by a reduced plate flexure, large proportion of low velocity-density in the edifice, crustal underplating under the presently active part, which may be seen as the level of transposition from primary melts into the more tholeitic products erupted. Structural heterogeneities of the preexisting plate are documented which may control magmatism in addition to slow plate/hotspot motion. Two completely different tomographies establish a high-velocity central core under the active volcano summit area which contrasts with the surroundings interpreted as eruptives or transported material and rises from the top of the oceanic plate. Another distinct high-velocity body to the East, which top has been drilled, is imaged to be rooted in the plate. So-called rift-zones of the active volcano are not marked by velocity anomalies, hence do not appear to be magmatic. High resolution location of earthquakes, heterogeneity in Vp/Vs and magnitude-number evolution can be correlated for the preeruptive magma migration to the most recent eruption. Mt Etna is not in a simple geodynamic context typical for volcanism. Marine reflection seismics reveal extension of crustal scale, which has succeeded to the compression related to the subduction at the Calabro-Sicilian arc. It is interpreted as marking a change in slab coupling, with which the development of Etna appears coeval... (D'après résumé d'auteur)

SUMMARY The Racha-Dzhava earthquake (Ms= 7.0) that occurred on 1991 April 29 at O9 : 12: 48.1 GMT in the southern border of the Great Caucasus is the biggest event ever recorded in the region, stronger than the Spitak earthquake (Ms=6.9) of 1988. A field expedition to the epicentral area was organised and a temporary seismic network of 37 stations was deployed to record the aftershock activity. A very precise image of the aftershock distribution is obtained, showing an elongated cloud oriented N105", with one branch trending N310" in the western part. The southernmost part extends over 80 km, with the depth ranging from O to 15 km, and dips north. The northern branch, which is about 30 km long, shows activity that ranges in depth from 5 to 15 km. The complex thrust dips northwards. A stress-tensor inversion from P-wave first-motion polarities shows a state of triaxial compression, with the major principal axis oriented roughly N-S, the minor principal axis being vertical. Body-waveform inversion of teleseismic seismograms was performed for the main shock, which can be divided into four subevents with a total rupture-time duration of 22s. The most important part of the seismic moment was released by a gentle northerly dipping thrust. The model is consistent with the compressive tectonics of the region and is in agreement with the aftershock distribution and the stress tensor deduced from the aftershocks. The focal mechanisms of the three largest aftershocks were also inverted from body-wave records. The April 29th (Ms=6.1) and May 5th (Ms= 5.4) aftershocks have thrust mechanisms on roughly E-W-oriented planes, similar to the main shock. Surprisingly, the June 15th (Ms = 6.2) aftershock shows a thrust fault striking N-S. This mechanism is explai'ned by the structural control of the rupture along the eastdipping geometry of the Dzirula Massif close to the Borzhomi-Kazbeg strike-slip fault. In fact, the orientation and shape of the stress tensor produce a thrust on a N-S oriented plane. Nappe tectonics has been identified as an important feature in the Caucasus, and the source mechanism is consistent with this observation. A hidden fault is present below the nappe, and no large surface breaks were observed due to the main shock. The epicentral region is characterized by sediments that are trapped between two crystalline basements: the Dzirula Massif, which crops out south of Chiatoura, and the Caucasus Main Range north of Oni. Most, if not all, of the rupture is controlled by the thrusting of overlapping, deformed and folded sediments over the Dzirula Massif. This event is another example of blind active faults, with the distinctive feature that the fault plane dips at a gentle angle. The Racha Range is one of the surface expressions of this blind thrust, and its growtb is the consequence and evidence of similar earthquakes in the past. II 5

Nous avons analysé les données des stations portables des expériences BANJO, SEDA et PISCO afin de déterminer les paramètres de séparation des ondes S (direction de polarisation rapide F, retard dt) sous les réseaux. Les réseaux suivent une ligne EW à 20° à travers les Andes, et une ligne globalement NS de 16° à 26°. Le long de la ligne EW, F est le plus souvent orienté EW, alors qu'ailleurs il est le plus souvent parallèle à la zone de subduction. Nous en concluons que le signal principal vient de sous la plaque Nazca, ce qui indique un champ d'écoulement complexe avec une composante parallèle à la zone de subduction. Nous suggérons que le champ d'écoulement est relié à la structure des Andes à grande échelle. (Résumé d'auteur)

Le déploiement d'un réseau sismique large-bande passif dans les Andes et au Tibet fournit un excellent échantillonnage des deux plus grands plateaux à la surface du globe. Les propriétés crustales moyennes des deux plateaux estimées à partir des deux expériences sismologiques sont significativement différentes. L'Altiplano est caractérisé par une croûte uniforme de 65 km d'épaisseur avec des faibles vitesses (P : 6,0 km/s) et un rapport de Poisson (RP) d'environ 0,25. Le Tibet montre des variations N-S systématiques de l'épaisseur crustale et du rapport de Poisson : 75 km avec un RP de 0,25-0,27 au sud, 70 km avec un RP de 0,27-0,28 aux latitudes centrales, et 55 km avec un RP > 0,3 au nord. Alors que l'Altiplano apparaît être en équilibre isostatique, d'autres forces doivent agir pour maintenir le Tibet à son altitude uniforme. (Résumé d'auteur)

L'atténuation intrinsèque et l'atténuation "scattering" sont calculées pour des séismes qui se sont produits entre 1993 et 1996 à moins de 30 km de distance épicentrale de la station MEJ du réseau sismologique d'Antofagasta. Alors que l'atténuation intrinsèque ne montre aucune variation temporelle, l'atténuation "scattering" augmente à MEJ avant le tremblement de terre d'Antofagasta du 30 juillet 1995 et regagne en 1996 son niveau de 1993. (Résumé d'auteur)

Un réseau large-bande passif déployé dans les Andes Centrales a enregistré de nombreux séismes régionaux. La modélisation des séismogrammes montre que la croûte de l'Altiplano est caractérisée par des vitesses P faibles, un rapport de Poisson de 0,25, et une épaisseur crustale de 65 km. L'épaisseur crustale varie à travers l'orogène, de 70-74 km sous les cordillères Real et occidentale, à 43-47 km dans le subandin, et à 32-38 km sous la plaine du Chaco. Ces propriétés moyennes sont en accord avec un raccourcissement d'une lithosphère peu rigide entre deux lithosphères plus fortes. (Résumé d'auteur)

Les temps d'arrivées d'ondes P et S de séismes locaux ainsi que les ondes P de téléséismes ont été utilisés pour définir les modèles de vitesses (1D et 3D) pour la croûte et le manteau supérieur, sous les Andes équatoriennes. Comme dans les Andes Centrales boliviennes, sous les cordillères, les vitesses sont plus rapides qu'au niveau de la partie centrale des Andes, Altiplano en Bolivie et dépression interandine en Equateur, le moho étant plus profond sous ces zones. (Résumé d'auteur)

A partir de l'analyse des données sismologiques de l'expérience Lithoscope BOLIVIE/CHILI en 1994, nous étudions les variations latérales d'amplitudes des phases régionales pour en déduire des caractéristiques de la structure crustale. Notamment, nous cherchons à préciser s'il existe effectivement une anomalie de propagation des ondes Lg perpendiculairement à la direction des unités géomorphologiques. (Résumé d'auteur)

La distribution de vitesses dans le modèle 3D indique l'existence d'une zone à faible vitesse entre 30 et 40 km de profondeur et la présence d'un prisme asténosphérique qui sépare le "slab" de la racine de la Cordillère des Andes. Des contraintes extensives et compressives s'observent simultanément autour de 100 et 200 km de profondeur. (Résumé d'auteur)

Observed patterns of high-frequency seismic wave attenuation suggest that near 22°S, the upper mantle structure beneath the central Andean plateau changes along strike to the south. Contrasting regions of high and low upper mantle seismic wave attenuation beneath the plateau are identified based on striking azimuthal variations in the character and frequency content of P and S waves propagating beneath the plateau to a portable seismic network deployed in Jujuy Province, Argentina (24°S, 65°W). Ray paths from intermediate depth earthquakes located north and northwest of the network transmit seismic waves with a higher frequency content than ray paths from earthquakes at similar depths and distances but located west and south of the network. The estimated apparent Q values fall into two categories: Qp >500 and Qs >350 for the high-Q paths, and Qp