New Insights into Magnitude and Ground-Motion Modeling in Moderate-Seismicity Regions from Compiling a Consistent Dataset of Ground-Shaking Intensity Measures

ABSTRACT We present the compilation strategy and the key features of a unique dataset of ground-motion intensity measures (IMs) calculated from high-quality waveform data of earthquakes that occurred in Switzerland and its bordering regions since 2000. A total of 2.6 million waveforms were initially processed, and a substantial portion was eliminated based on strict quality criteria for waveform data and event/station metadata. The final dataset comprises IMs calculated from about 260,000 triaxial waveforms with uniform minimum bandwidth and signal-to-noise ratio generated by 22,000 earthquakes with local magnitudes lower than 5. All waveforms were consistently processed using the same automated workflow based on state-of-the-art, free, open-source seismological software to ensure homogeneity, objectivity, and reproducibility of the results. This cannot be achieved by relying only on the existing catalogs of station-magnitude data, which have been heterogeneously compiled during the last decades. Although the main objective of this study was to generate Wood–Anderson amplitudes (AWA) to calibrate new local magnitude models, several other ground-motion IMs largely used in engineering seismology (e.g., peak ground motions, response and Fourier spectra, and durations) were also calculated and included in the dataset, which can be used as an input for further studies in the domains of earthquake magnitude, ground-motion modeling, and seismic hazard and risk. Particular attention was paid to investigating the effects of filtering and rotating components on the resulting amplitudes and station magnitudes. Among other findings, we highlight the need to account for near-source and complex path effects in the development and assessment of the next-generation magnitude and ground-motion models in Switzerland and surrounding regions.