GNGTS 2022 - Atti del 40° Convegno Nazionale

GNGTS 2022 Sessione 2.1 239 VARIABILITY OF SOURCE PARAMETERS IN CENTRAL-SOUTHERN EUROPE M-H Yen 1,2 , D. Bindi 1 , R. Zaccarelli 1 , A. Oth 3 , B. Edwards 4 , F. Cotton 1,2 1 GFZ-German Research Centre for Geosciences, Germany 2 University of Potsdam, Germany 3 European Center for Geodynamics and Seismology, Luxembourg 4 University of Liverpool, England Abstract. Source parameter, stress drop, is a key parameter estimating strong ground motion, as it controls the level of peak ground acceleration (Hanks and Johnson, 1976). A non-parametric spectral decomposition of the Fourier amplitude spectra (FAS) is applied to isolate source, site and path contributions, solving regional attenuation models for two spatial domains across the Alps. The source spectra isolated from the other site contributions are fitted to a standard omega square model (Brune model) to determine the seismic moment and the corner frequency. These values are in turn used to compute the stress drop considering a circular rupture model with uniform stress drop. Finally, we discuss the variability of stress drop with earthquake size from southern to northern Europe and their dependencies on the origin of earthquakes (induced and tectonic). Introduction. The knowledge of source parameters is a crucial topic for the seismic hazard assessment, from ground-motion prediction equations to physics-based simulations. Stress drop is a key parameter in the estimation of strong ground motion, as it controls the level of peak ground acceleration (Hanks and Johnson, 1976). It determines the position of the corner frequency and the height of the high-frequency plateau in the Fourier amplitude spectrum of acceleration. Several studies of source parameters for different regions have been investigated such as Japan (Oth et al. , 2017), southern California (Baltay et al. , 2017, Trugman and Shearer, 2017), central Italy (Bindi et al. , 2019), and global datasets (Baltay et al. , 2019). However, modeling assumptions make the results difficult to compare between different studies, even in the same target regions. In this study, we estimate source parameters of events in central-southern Europe from the stable part (mainly Switzerland and Germany) to the active part (mainly central and northern Italy) of the continent. A non-parametric spectral decomposition approach is applied on the Fourier amplitude spectra (FAS) of events in the region to determine the source spectra, site contributions and attenuation propagation effects, considering two regional attenuation models. To remove the trade-off between source and site terms, we constrain a-priori the site amplification at one reference station contributing to both attenuation domains. The obtained source spectra are fitted to a standard ω 2 -model to evaluate the seismic moment and the corner frequency, which in turn are used to compute the stress drop. Finally, we discuss the variability of stress drop with earthquake size, regions, and source origins. Data and Processing . The recordings from 1990 to 2020 in the regions were downloaded and processed by means of stream2segment (Zaccarelli et al. , 2019) using as event catalog the International Seismological Center, ISC. The database hosts about 52 million segments in total from 178,000 earthquakes recorded by 4,771 stations belonging to 118 either temporary or permanent networks. Due to potential noises and outliers in the datasets from the automatic downloaded and processed by the software, the quality control (signal-to-noise ratio and outlier detection) are applied. Finally, we select 4,380 earthquakes with a magnitude larger than 2.5 in central-southern Europe recorded by 996 stations (Fig. 1). We match the location and origin times of selected earthquakes with the EMEC catalog to identify induced events for the further discussion of the scaling relationship of source origins. The FAS are computed by the vector sum of two horizontal components, which is independent of the sensor orientation. The considered frequency band is from 0.3 to 25 Hz.