Sismologie différentielle : caractérisation et imagerie dynamique avec les ondes diffuses. – SisDif

Classical waves -including acoustic, seismic, radio and optical ones- propagating in heterogeneous media are subject to multiple scattering processes that disable conventional imaging and communication techniques. Nevertheless, diffuse waves (coda waves in the case of seismology) are perfectly deterministic and contain crucial information about the propagation medium. The purpose of our proposal is to characterize, image and monitor heterogeneous media with acoustic or seismic waves thanks to, and not despite, multiple scattering. Our fields of application concern the earth crust and lithosphere, volcanoes and active seismic faults. Three features make our approach quite original. 1) We will study acoustic or seismic waves following a differential approach (thus exploiting the spatial and temporal derivatives of the wavefield in amplitude and phase). This approach has been subject to feeble attention in acoustics and seismology, and is rendered possible thanks to the very high density of sensors nowadays available. 2) We will simultaneously conduct theoretical developments, numerical works, laboratory and field experiments. We will eventually propose new techniques and experimental configurations that can be applied to seismology. 3) The phenomena that we are studying have multi-length scales (in space and time) and exist along a range of frequencies ranging from mHz to MHz. The project falls into three work packages: I. We will study the dynamics of equipartition of diffuse waves at the global scale. This workpackage is based on the evaluation of the relative compressional and shear energy of coda waves. This observation is derived from the spatial derivatives of the wave field. The dynamics of equipartition are connected to the scattering strength of the earth, especially in the lithosphere in the long period range of interest. From the equipartition dynamics, we will characterize the level of heterogeneity of the lithosphere with the scattering mean free path at the scale of 10 km to 1000 km. This workpackage relies on the confrontation of real seismic data collected from worldwide seismic networks to numerical simulations (spectral element method). II. We will detect, characterize and locate weak changes in multiple scattering media thanks to temporal variations of the coda waveforms. Note that these changes cannot be detected with direct waves and conventional repeated travel time tomography. This workpackage is based on theoretical and experimental works in the laboratory. Our goal is to develop experimental configurations and data processing procedures that will eventually be applied to seismology around volcanoes and active faults. III. We will study the spatial correlation of the phase derivatives on a volcano, at the scale of 100m to 10 km. By phase, we mean the complex angle of the analytic signal, which is constant in homogeneous media but varies rapidly in heterogeneous ones. The spatial correlation of the phase is directly connected to the scattering mean free path of elastic waves in the subsurface. The first advantage of this technique inherited from optics is that it is independent from the absorption. The second advantage is that it does not depend on the sources parameters (duration, polarization) These work packages are connected to each other's through their common differential approach to diffuse waves, and will benefit from cross-fertilization. Our project is based on cutting edge scientific developments, and will apply concepts developed in optics, quantum mechanics and mesoscopic physics to acoustics and seismology. Every team member is involved in each work package. Each workpackage will require a collaborative effort from each partner in his specific field of expertise. The team is composed of three young researchers and one PhD student, and is located in Grenoble (except L Margerin, located in Aix, and whose task consists in numerical simulations and data processing. This task can be partially realized with remote access to Grenoble computers). We share and mix trans-disciplinary competences, including seismology, acoustics, theoretical and condensed matter physics and numerical simulations. By the end of this project, our goal is to attain a complete autonomy in terms of funding, but also in terms of project management and scientific initiative. All the members of the team are already working together (monthly meetings) and most of us share common publications [5, 22, 32]. The team has realized preliminary studies in every part of the projects. A strong financial support from the ANR is absolutely necessary to develop our team and to successfully pursue this head start.