Improving High-Frequency Spectral Estimates in Spectral Ratio Analysis Using Between-Event PGA Differences

Abstract Assessment of ground-motion variability is crucial for seismic hazard analysis. However, the variability of peak ground accelerations (PGAs) and spectral stress drops has long been inconsistent, partly due to limitations of the single-corner-frequency (SCF) model. Recent studies indicate that accurate evaluation of the high-frequency spectral level, which often requires a double-corner-frequency (DCF) model, can reconcile this discrepancy. Building on these findings, consistency between PGAs and source spectra can serve as a constraint for improving high-frequency spectral estimates. This study proposes a simple method that extends SCF spectra estimated from the standard spectral ratio approach to DCF spectra by incorporating PGA differences between earthquake pairs. The developed method (1) requires only widely available PGA measurements as additional data, (2) yields source spectra that reproduce the scaling and variability of observed PGAs, and (3) enables estimation of DCF spectra from existing spectral ratio results with minimal extra analysis. These features highlight the potential of PGA data as a practical constraint for refining high-frequency source spectral estimates.