Crustal Attenuation and Site Amplification Across the Central and Eastern United States from Lg Q Tomography
Abstract
We present a new model of crustal Lg-wave attenuation and site amplification across the central and eastern United States (CEUS), derived from joint tomographic inversions of vertical and horizontal seismic amplitudes. We use more than 75,000 broadband records from over 1,600 regional crustal earthquakes (Mw≥3.5, depth <30 km) recorded between 2007 and 2024 at more than 1,500 seismic stations. Using this dataset, we improved the imaging of the frequency-dependent seismic quality factor (Q), site terms, and source amplitudes across five center frequencies from 0.7 to 11.3 Hz. Our results show significant lateral variations in attenuation even in what has been considered the low attenuating Midcontinental region. Q0 values as low as ∼200 to 230 occur in sediment-rich, tectonically active regions. Stable cratonic interiors exhibit high Q0 values (up to ∼380) but contain lower Q subregions (some newly identified) associated with tectonic features in the crust and mantle. We also resolve the frequency dependence of attenuation via the η parameter, which varies from ∼0.7 in high-Q0 regions to >1.0 in areas with strong scattering. Site amplification terms show higher values (50%–100% increase) in intracratonic basins and coastal plain sediments and lower values (10% decrease to 20% increase) in regions with exposed or shallow bedrock. The horizontal-to-vertical site amplification ratio exhibits a polarity shift across the Coastal Plain boundary. Checkerboard and feature recovery tests demonstrate the model’s spatial resolution down to ∼100 km, and jackknife resampling confirms relative uncertainties below 2% across most of the domain. Compared to previous CEUS attenuation studies, our model benefits from denser ray-path coverage and improved inversion methodology. These findings provide critical input for refining seismic hazard assessments and updating ground-motion prediction models for the next-generation U.S. National Seismic Hazard Model.