Teleseismic Body Wave Attenuation in the Upper Mantle beneath the United States

Abstract:

EarthScope seismic data provide opportunities to examine mantle properties on a continental scale as the Transportable Array (TA) nears the end of its traverse across the contiguous United States. We use P- and S-wave amplitude spectra from all >M5.7 deep earthquakes recorded by the TA to examine seismic attenuation patterns in the upper mantle. More than 2 million inter-station P-wave spectral ratios were inverted for maps of relative t_P* variations across the U.S. in multiple frequency bands between 0.08 – 2 Hz. We plan to have corresponding S-wave results by meeting time. Maps of t_P* are strongly correlated (>0.8) for frequency bands of 0.08 – 2 Hz, 0.25 – 2 Hz, 0.08 – 1 Hz, and 0.25 – 1 Hz. The broader the frequency band examined (e.g. 0.08 – 2 Hz), the lower the magnitude in variations of t_P*; however, those broader frequency bands still exhibited geographic patterns similar to the narrow frequency bands. We compare our maps’ t_P* with seismic velocity models and constraints on crustal scattering to assess the physical origin of apparent attenuation. In the tectonically active and high heat flow domain of the western U.S., t_P* variations are moderately correlated with thermal variations predicted by tomography studies of seismic velocity. However, contrast in t_P* between western Cordillera and the cratonic interior is weaker than predicted by tomography. Additionally some areas of high attenuation are correlated with Precambrian tectonic boundaries within the Laurentian craton. The weak contrast between the western and eastern U.S. and correlations with Precambrian tectonics suggest that elastic scattering due to small-scale (~10 – 100 km) heterogeneity or compositional variations in the lithosphere are major contributors to t_P* estimates from deep earthquake spectral ratios. Moderate correlation of t_P* with estimates of mantle temperature within the western U.S. suggests deep earthquake spectral ratios do carry some evidence of intrinsic attenuation, but separating the effects of elastic scattering and intrinsic attenuation remains a challenge.