Local Eikonal Tomography Using Ambient Noise Records From a Dense Array of Seismic Nodes Deployed in a Sediment-Filled, Deeply Incised Valley With an Extreme Subsurface Topography (Rhône Valley, Southern France)
Description:
Surface waves extracted from ambient noise recordings are often used to image the shallow subsurface structure through linearized tomographic approaches. These approaches require a-priori information such as a reference model and the trajectories of the surface waves between virtual sources and receivers, as well as the introduction of regularization equations which may reduce the resolution of the resulting group and/or phase velocity maps. In recent times, the eikonal tomographic approach has quickly gained popularity for dense seismic deployments, as it avoids posing and solving the linearized inverse problem. Instead, local phase velocities are computed from the gradient of the phase traveltimes measured between the virtual source and receivers. In this work, we applied the eikonal approach to data from a dense seismic deployment carried out in the heavily industrialized area of the Tricastin Nuclear Site (Rhône valley, southern France). This area is situated above a sediment-filled, deeply incised canyon with a extreme subsurface topography dug during the Messinian Salinity Crisis. The deployment consisted in 400 3-component nodes installed over a 10 x 10 km area, which recorded continuously during February-March 2020. The recorded ambient noise wavefield shows to be complex, i.e. higher mode propagation and non-uniformly distributed sources of ambient noise are clearly observed. Despite the wavefield complextiy we have been able to implement a practically fully-automated method that performs higher mode discrimination and measures the phase traveltimes of the fundamental mode from the phase spectra of the cross-correlations of the ambient noise recordings. We then applied the eikonal approach to the measured phase traveltimes to build a set of phase velocity maps for Rayleigh waves. The resulting maps cover frequencies ranging from 0.4 to more than 5 Hz and show a good correlation with the existing knowledge of the geological structures.
Session: Site-specific Modeling of Seismic Ground Response: Are We Quantitative Enough to Predict? [Poster]
Type: Poster
Date: 4/19/2023
Presentation Time: 08:00 AM (local time)
Presenting Author: Andrés Olivar-Castaño
Student Presenter: No
Invited Presentation:
Authors
Andrés Olivar-Castaño Presenting Author Corresponding Author andres.olivar-castano@uni-potsdam.de University of Potsdam |
Matthias Ohrnberger Matthias.Ohrnberger@geo.uni-potsdam.de University of Potsdam |
Marco Pilz pilz@gfz-potsdam.de GFZ Potsdam |
Annabel Händel ahaendel@gfz-potsdam.de GFZ Potsdam |
Pierre Boué pierre.boue@univ-grenoble-alpes.fr Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, UGE, ISTerre |
François Lavoué francois.lavoue@univ-grenoble-alpes.fr Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, UGE, ISTerre |
|
|
|
Local Eikonal Tomography Using Ambient Noise Records From a Dense Array of Seismic Nodes Deployed in a Sediment-Filled, Deeply Incised Valley With an Extreme Subsurface Topography (Rhône Valley, Southern France)
Category
Site-specific Modeling of Seismic Ground Response: Are We Quantitative Enough to Predict?