Hybrid Broadband Seismograms for Seismic Scenarios of the Po Plain Sedimentary Basin (Northern Italy)

Abstract

The sedimentary basin of the Po Plain, northern Italy, is a region of low to intermediate seismicity that has gained attention in the recent years because of the May 2012 seismic sequence reaching magnitudes Mw~6. Its repercussions in Emilia- Romagna, Lombardia and Veneto, regions of a high density of population and industries, lead to an overall damaging cost of about 13 billion euro. Because of scarce recent seismicity — hence relatively little knowledge based on instrumentally recorded data — but high exposure and hence substantial seismic risk, there is an obvious need to generate detailed seismic shakemaps, based on an accurate physical description of seismic wave propagation. Recent improvements in the knowledge of the three-dimensional structure of the earth's crust and developments of computational seismology allow now pursuing this objective. Nowadays, standard approaches for synthetic waveform simulation are able to reproduce accurately only different portions of the frequency spectrum separately: low frequencies with deterministic methods or high frequencies with stochastic methods. We can hardly find a single method that reproduces the entire spectrum. We apply and evaluate a hybrid deterministic-stochastic method[1] developed in the recent years at the San Diego State University and implemented in the Broadband Platform[2] at the Southern California Earthquake Center, for the definition of seismic shakemaps in northern Italy. The method consists of integrating deterministic low frequency signals together with stochastic high frequency signals, enabling to obtain the whole gamut of seismic shaking. More specifically, high frequency seismogram computation follows the S-S backscattering method whereas low-frequency signals are generated externally through the open-source software SPECFEM3D[3] that reproduce accurately the seismic response up to 1sec in a fine grid of 3 million spectral elements based on a regional 3D model, with high resolution description of the Po Plain sedimentary basin[4], and following a finite-source representation.