Frequency-dependent site factors for the Icelandic strong-motion array from a Bayesian hierarchical model of the spatial distribution of spectral accelerations

The earthquake ground motions of over 1700 earthquakes recorded on a small-aperture strong-motion array in south Iceland (ICEARRAY I) that is situated relatively uniform site condition characterized as rock, exhibit statistically significant spatial variation ground-motion amplitudes across the array. Both and microseismic horizontal-to-vertical spectral ratios (HVSR) have been shown to distinct some cases, bimodal peaks amplification, indicating resonance at periods 0.1–0.3 s, phenomenon has attributed surface layer lava rock lying above sedimentary layer, structure then repeated with depth under In this study, we implement Bayesian hierarchical model (BHM) seismic partitions residuals into event term, station event–station term. We analyzed compared peak acceleration (PGA) 5% damped pseudo-acceleration response spectrum (PSA) oscillator T = 0.05–1.0 s. results show terms, dominate total variability However, terms are increase period range s most stations different extents, leading an overall those periods, captured by larger inter-station standard deviation. As measure how much deviate from average, they act proxies for localized effects amplification factors. These results, improve our understanding key factors affect small area ICEARRAY I. This approach can help accuracy hazard assessments local scales, which turn could contribute more refined risk engineering decision-making.