Pierre Boué

L'utilisation d'un grand nombre de capteurs sismiques est de plus en plus courant pour imager l'intérieur de notre planète depuis sa surface pour la prospection sismique, jusqu'à sa structure profonde avec la sismologie continentale et globale. L'application d'un traitement d'antenne aux enregistrements issus de réseaux de capteurs permet l'extraction de nouvelles observables et une meilleure compréhension de la propagation des ondes dans les milieux complexes. Parmi ces méthodes, on s'intéresse particulièrement aux traitements simultanés en émission-réception de type double formation de voies (DFV). A l'échelle de la prospection sismique, la DFV est utilisée pour extraire des ondes de volume pouvant être masquées par des ondes de surface plus énergétiques. A l'échelle continentale, les réseaux de sources étant plus rares, on propose d'appliquer la méthode DFV à des signaux reconstruits par corrélation du bruit sismique ambiant. De la même manière que pour un couple de stations, la corrélation d'enregistrements continus permet d'évaluer la fonction de Green entre deux antennes réceptrices. Cette méthode est appliquée à des données du réseau Transportable Array (USArray) afin de mesurer et cartographier la vitesse de phase des ondes de surface au centre des USA. Enfin à l'échelle globale, une combinaison de plusieurs grands réseaux sismologiques est utilisée pour démontrer que la corrélation d'enregistrements continus, dans la gamme de périodes, 5-100s permet la reconstruction des ondes de volume à des distances télésismiques. Une analyse de la contribution respective du bruit ambiant, d'origine océanique, et des séismes est réalisée. On montre que les arrivées tardives des forts séismes, réverbérées à l'intérieur du globe, contribuent de manière importante à la reconstruction des phases profondes. Les ondes de volume reconstruites à partir du bruit ambiant constituent une nouvelle source d'information, complémentaire aux données issues des séismes, et pouvant être utilisée pour imager notre planète.

Seismic ambient noise spectra ubiquitously show two amplitude peaks corresponding to distinct oceanic wave interaction mechanisms called primary (seismic period (T) ~ 14 s) and secondary (T ~ 7 s) microseism. Seismic noise records are used in a wide range of applications including crustal monitoring, imaging of the Earth's deep interior using noise correlations, and studies on the coupling between oceans and solid Earth. All of these applications could benefit from a robust knowledge of spatiotemporal dynamics of microseismic sources. Consequently, seismologists have been studying how to model microseismic sources of ambient noise with the recent improvements in ocean wave models. Global sea state and its derivative products are now covering the past decades in models such as the WAVEWATCHIII hindcast. This paper introduces the Wave Model Sources of Ambient Noise (WMSAN, pronounced [wam-san]) Python package. This modular package uses standardized wave model outputs to visualize ambient noise source maps and efficiently compute synthetics of seismic spectrograms and cross-correlations for surface waves (Rayleigh) and body waves (P, SV), in a user-friendly way.

We demonstrate the effectiveness of seismic dense array surface wave focal spot imaging using USArray data from the western‐central United States. We study dispersion in the 60–310 s period range and assess the image quality of fundamental mode Rayleigh wave phase velocity maps. We apply isotropic spatial autocorrelation models to the time domain zero lag noise correlation wavefield data at distances of about one wavelength. Local estimates of the phase velocity, its uncertainty, and the regression quality imply overall better ZZ relative to ZR or RZ results. The extension of the depth resolution compared to passive surface wave tomography is demonstrated by the inversion of three clustered dispersion curves from different tectonic units. We observe anisotropic surface wave energy flux and the influence of body wave energy, but sensitivity tests at 60 s targeting the data range, correlation component, and processing choices show that the ZZ focal spots yield consistent high‐quality images compared to regional tomography results in the 60–150 s period range. In contrast, at 200–300 s the comparatively small scales of the imaged structures and the imperfect agreement with low‐resolution global tomography results highlight the persistent challenge to reconcile imaging results based on different data sources, theories, and techniques. Our study shows that surface wave focal spot imaging is an accurate, robust, local imaging approach. Better control over clean autocorrelation fields can further improve applications of this seismic imaging tool for increased resolution of the elastic structure below dense seismic arrays.

Body wave extraction from oceanic secondary microseismic sources with seismic interferometry provides alternative information to better constrain the Earth's structure. However, sources' spatiotemporal variations raise concerns about travel time measurement robustness. Therefore, we study the cross‐correlations’ stability during a single oceanic event. This study focuses on 3 days of data and three seismic arrays' combinations between 8 and 11 December 2014 during storm Alexandra, a “weather bomb” event in southern Greenland. We use the WAVEWATCH III hindcast to model P‐wave noise sources and assess the impact of short‐term source variations on cross‐correlations. Model‐based cross‐correlations compared to data show coherent delays to reference 3D Earth models (∼0–3 s) confirming the robustness of the source model which could explain minor travel time variations (≤1 s).

A document by Lisa Tomasetto. Click on the document to view its contents.

In certain geological settings such as sedimentary basins, the ground motion induced by an earthquake may be amplified by local site conditions. Estimating these site effects is important for seismic hazard assessment but can be difficult to do empirically due to the scarcity of site-specific field data in time and space, especially in low-to-moderate seismicity regions where the earthquakes needed for measuring the site effects have long return periods. In this study, we try to overcome these limitations and investigate an alternative approach based on ambient seismic noise and numerical simulations. More specifically, we use a 3D numerical model of seismic properties derived from Ambient Noise Surface-Wave Tomography (ANSWT) for 3D numerical simulations of seismic wave propagation, and consequently for a numerical estimation of seismic amplification in the basin. We illustrate the approach on a target site located in the French Rhône valley, where the Messinian salinity crisis has dug a paleo-canyon which is now filled by soft sediments in direct contact with a harder substratum, thereby providing typical conditions for significant site effects, as also observed by previous studies in the area. This work makes use of two dedicated datasets. On one hand, we use earthquake recordings acquired by a network of broadband stations deployed over the target site over 8 months, in order to estimate seismic amplification in the basin with respect to a rock-site reference via Standard Spectral Ratios (SSR), which we consider as our reference for evaluating our numerical results. On the other hand, we exploit one-month-long ambient noise recordings acquired by a dense array of 400 3C sensors. Prior to this work, this noise data was used to build a 3D shear-wave velocity ( V _S ) model of the target site via ANSWT, and also to estimate seismic amplification via noise-based Standard Spectral Ratios (SSRn). The obtained ANSWT model well reproduces the main geological structures of the basin, with lateral variations of velocities at depth depicting the deeper parts of the basin. However, our simulation results also show that some of its limitations related to surface wave sensitivity and resolution capability have an impact on the numerical amplification predicted in the basin. In particular, this ANSWT model lacks clear basin edges in order to efficiently trap seismic waves in the basin and to generate significant 3D wave propagation effects (diffractions, reflections, and generation of laterally propagating surface waves at the edges of the basin). As a result, the numerical amplification predicted in the ANSWT model remains dominated by a 1D response and does not reproduce the broadband character of the observed amplification at locations affected by significant 3D propagation effects. On the other hand, the numerical amplification predicted in the ANSWT model shows a good agreement with the observations at locations that seem less affected by 3D propagation effects, including in complex regions of the model where lateral variations must be taken into account. Our results therefore contribute to identify and better understand the potential and limitations of using ANSWT models for numerical site effect estimation. This study allows us to propose perspectives for future work to improve the approach, which remains promising for site effect assessment in low- to moderate-seismicity contexts.

Global seismological observations are sensitive to oceanic forcings, namely microseisms. In addition to dominant surface waves, these sources generate body waves that travel through the deep structures of our planet. Despite these sources’ inherent complexity, interferometric methods allow isolating coherent waves for imaging applications. For a given station pair, only specific microseism events contribute to the illumination of a specific target. We propose an opportune workflow based on ocean sea state models to extract robust interferences. This approach is illustrated with a strong microseism source in the North Atlantic Ocean, occurring around December 9, 2014.

Seismic velocities in rocks are highly sensitive to changes in permanent deformation and fluid content. The temporal variation of seismic velocity during the preparation phase of earthquakes has been well documented in laboratories but rarely observed in nature. It has been recently found that some anthropogenic, high‐frequency (>1 Hz) seismic sources are powerful enough to generate body waves that travel down to a few kilometers and can be used to monitor fault zones at seismogenic depth. Anthropogenic seismic sources typically have fixed spatial distribution and provide new perspectives for velocity monitoring. In this work, we propose a systematic workflow to seek such powerful seismic sources in a rapid and straightforward manner. We tackle the problem from a statistical point of view, considering that persistent, powerful seismic sources yield highly coherent correlation functions (CFs) between pairs of seismic sensors. The algorithm is tested in California and Japan. Multiple sites close to fault zones show high‐frequency CFs stable for an extended period of time. These findings have great potential for monitoring fault zones, including the San Jacinto Fault and the Ridgecrest area in Southern California, Napa in Northern California, and faults in central Japan. However, extra steps, such as beamforming or polarization analysis, are required to determine the dominant seismic sources and study the source characteristics, which are crucial to interpreting the velocity monitoring results. Train tremors identified by the present approach have been successfully used for seismic velocity monitoring of the San Jacinto Fault in previous studies.

Superficial geological layers can strongly modify the surface ground motion induced by an earthquake. These so-called site effects are highly variable from one site to another and still difficult to quantify for complex geological configurations. That is why site-specific studies can greatly contribute to improve the hazard prediction at a specific site. However, site-specific studies have historically been considered difficult to carry out in low-to-moderate seismicity regions. We present here seismological datasets acquired in the framework of the French–German dense array for seismic site effect estimation project in the heavily industrialized area surrounding the French Tricastin Nuclear Site (TNS). TNS is located above an ancient canyon dug by the Rhône River during the Messinian period. The strong lithological contrast between the sedimentary fill of the canyon and the substratum, as well as its expected confined geometry make this canyon a good candidate for generating site effects that are variable on short spatial scales. To investigate the impact of this geological structure on the seismic motion, we conducted complementary seismic campaigns in the area. The first main campaign consisted of deploying 400 nodes over a 10 × 10 km area for one month and aimed at recording the seismic ambient noise. A second seismic campaign involved the deployment of 49 broadband stations over the same area for more than eight months. This complementary campaign aimed at recording the seismicity (including local, regional, and teleseismic events). These different designs allowed us to target a variety of seismic data at different spatial and temporal scales. Beyond the interest for local operational seismic hazard applications, these datasets may be valuable for studying seismic wave propagation within complex kilometer-scale sedimentary structures. In this article, we present the deployment designs as well as initial analyses to provide information on the characteristics and the overall quality of the data acquired to future users.

Microseismic noise has been used for seismic velocity monitoring. However, such signals are dominated by low‐frequency surface waves that are not ideal for detecting changes associated with small tectonic processes. Here we show that it is possible to extract stable, high‐frequency body waves using seismic tremors generated by freight trains. Such body waves allow us to focus on small velocity perturbations in the crust with high spatial resolution. We report on 10 years of seismic velocity temporal changes at the San Jacinto Fault. We observe and map a two‐month‐long episode of velocity changes with complex spatial distribution and interpret the velocity perturbation as produced by a previously undocumented slow‐slip event. We verify the hypothesis through numerical simulations and locate this event along a fault segment believed to be locked. Such a slow‐slip event stresses its surroundings and may trigger a major earthquake on a fault section approaching failure.