What was the level of ground shaking during the 2019 Mw4.9 Le Teil (France) earthquake ?

01/02/2022 - 15/07/2022
Laboratoire(s) de rattachement : ISTerre
Encadrant(s) : Mathieu Causse, Cécile Cornou
Co-encadrant(s) : Sreeram Kotha
Contact(s) : mathieu.causse univ-grenoble-alpes.fr
Lieu : ISTerre, Grenoble
Niveau de formation & prérequis : M1, background in Earth Sciences, seismology, signal processing
Mots clés : Seismic Hazard, Numerical simulation, Artificial Intelligence

Context. On November 11, 2019, a Mw 4.9 earthquake hit the region close to Montelimar (lower Rhône Valley, France), an industrial region that hosts several operating nuclear power plants. Occurring in a moderate seismicity area, this earthquake is remarkable for its very shallow focal depth (about 1 km depth) and the moderate to large damages it produced in several villages despite its moderate magnitude (Cornou et al., 2020). In the absence of seismological stations near the fault zone, it is necessary to develop techniques to estimate the level of ground motion it generated, especially in the fault vicinity where the largest damage were observed. A preliminary study based on numerical predictions pointed to an exceptional and destructive level of ground acceleration exceeding 1g in the immediate fault vicinity, confirmed by in-situ observations (broken stones, displaced funeral slabs) (Causse et al., 2020). Almost 2 years after the Le Teil earthquake, sets of recently acquired geophysical data (FREMTEIL project, 2021-2023) makes it possible to perform more refined sets of simulations, incorporating the effects of the superficial geological structures (site effects).

Objectives. In order to quantify the level of ground shaking during the Le teil earthquake, this internship will focus on performing suite of numerical ground motion simulations on a dense array of stations located above the fault. First, in order to properly model the superficial geological structures (site effects), geophysical measurements will be process so as to obtain site amplification functions at different positions in the fault vicinity, which will be incorporated in the numerical simulations. Second, Artificial Intelligence techniques (e.g. Random Forest algorithms) will be deployed to analyze the set of simulations and quantify the impact of the different parameters on the ground motion, including the seismic rupture characteristics (fault geometry, rupture propagation on the fault) and the wave propagation (large scale and superficial geological structrures). Such analysis is essential to determine what are the main physical properties that control near-source ground acceleration in such context, and hence what complementary information / additional effort is needed to improve ground motion predictions.

Reference
Causse, M., C. Cornou, E. Maufroy, J.R. Grasso, L. Baillet and E. El Haber (2021). Exceptional Ground Acceleration during the shallow Mw4.9 Le Teil earthquake, France, Communciations Earth & Environment, https://doi.org/10.1038/s43247-020-00089-0

Contact. mathieu.causse univ-grenoble-alpes.fr

Mis à jour le 22 septembre 2021