Numerical simulation of wave propagation in the Grenoble basin: Benchmark ESG2006

Hideo Aochi (1, 2) , Julien Rey (1) , John Douglas (1)

1 BRGM - Bureau de Recherches Géologiques et Minières (BRGM)
2 Disaster prevention Research Institute

Numerical simulations of wave propagation in the Grenoble basin are undertaken using the finite difference method (FDM) in a staggered grid framework (4th order in space) for an elastic medium with a flat ground surface and perfectly matching layer absorbing boundaries. The seismic source is introduced in the form of a seismic moment tensor irrespective of whether a point source or expanding finite source model is used. This is the same methodology adopted in our previous studies (Aochi and Douglas, 2006; Douglas et al., 2006). As the FD grids in our code are equally spaced in both horizontal and vertical directions, the grid size should be fine enough to describe the Grenoble Basin where a low velocity (300m/s) zone is present near the surface (a few 100 m depth at maximum) while the surrounding bedrock has a velocity ten times higher. We carry out the simulations with different resolutions (a grid size of 100 m and of 50 m). Both of them require high performance computation resources. This benchmark test is an interesting application for seismic hazard evaluation because of the high contrast basin structure and its non-negligible dimensions. For the two large earthquake scenarios (S1 and S2 of magnitude 6), our numerical simulations are compared with empirical ground-motion models for peak ground velocity (PGV) to validate the calculations and to evaluate the local effect of the Grenoble basin. The equations derived by Campbell (1997) better match the simulations for the rock sites outside the basin than for the soil sites within the basin. It is remarkable that PGVs at some basin sites show changes of several times at the same distance from the earthquake fault. We also observe strong directivity effects from rupture propagation along the fault. The obtained scatter is compatible with other simulations which we have performed recently using dynamic rupture models and different crust structure models (Aochi & Douglas, 2006; Douglas et al., 2006). As the resolution of the basin model is not enough good and the inelastic attenuation is not fully taken into account, it is still difficult to quantitatively compare signal durations. However it is visible that the basin resonates a long time after the passage of body waves.

Domaines: Géophysique [physics.geo-ph]; Géophysique [physics.geo-ph]; Milieux et Changements globaux

Hideo Aochi, Julien Rey, John Douglas. Numerical simulation of wave propagation in the Grenoble basin: Benchmark ESG2006. Third International Symposium on the Effects of Surface Geology on Seismic Motion, Aug 2006, Grenoble, France. Paper Number: 32. ⟨hal-00714032⟩