Elasto-acoustic modelling and simulation for the seismic response of structures: The case of the Tahtal{\i} dam in the 2020 \.Izmir earthquake

TL;DR

This paper develops a comprehensive 3D elasto-acoustic model to simulate the seismic response of the Tahtali dam during the 2020 Izmir earthquake, integrating complex geological and structural factors for accurate risk assessment.

Contribution

It introduces a multi-scale, physics-based simulation framework using discontinuous Galerkin methods to accurately model dam seismic response and validate against real data.

Findings

Synthetic seismograms closely match observed data.

The model effectively captures complex wave interactions at structure-soil interfaces.

Comparison shows advantages over classical plane wave approaches.

Abstract

As a mean to assess the risk dam structures are exposed to during earthquakes, we employ an abstract mathematical, three dimensional, elasto-acoustic coupled wave-propagation model taking into account (i) the dam structure itself, embedded into (ii) its surrounding topography, (iii) different material soil layers, (iv) the seismic source as well as (v) the reservoir lake filled with water treated as an acoustic medium. As a case study for extensive numerical simulations we consider the magnitude 7 seismic event of the 30$^{\rm th}$ of October 2020 taking place in the Icarian Sea (Greece) and the Tahtali dam around 30 km from there (Turkey). A challenging task is to resolve the multiple length scales that are present due to the huge differences in size between the dam building structure and the area of interest, considered for the propagation of the earthquake. Interfaces between structures and highly non-conforming meshes on different scales are resolved by means of a discontinuous Galerkin approach. The seismic source is modeled using inversion data about the real fault plane. Ultimately, we perform a real data driven, multi-scale, full source-to-site, physics based simulation based on the discontinuous Galerkin spectral element method, which allows to precisely validate the ground motion experienced along the Tahtali dam, comparing the synthetic seismograms against actually observed ones. A comparison with a more classical computational method, using a plane wave with data from a deconvolved seismogram reading as an input, is discussed.