Crustal attenuation from USArray ML amplitude tomography

Abstract

Summary Seismic attenuation across the US is estimated using station ML magnitude data from the USArray. Station magnitudes are recalibrated back to amplitude and back projected in a two-dimensional tomography. Data represent the amplitudes of the horizontal components of the Lg phase. The western US shows regions of very high attenuation and contrasts with the lesser attenuation of the eastern US. Individual attenuation anomalies can be clearly tied to regional geology. Station gains show broad regional variations that match geographic regions. Most of the high attenuation areas are regions of high geothermal activity suggesting that intrinsic attenuation dominates over scattering attenuation. An exception is the central San Andreas Fault zone because it lacks any localized heat-flow anomaly. The US east of the Rocky Mountains is bland and contains none of the high attenuation regions of the western US. Instead the central US has low attenuation patches that do not obviously correspond to geologic province. Sediments of the Gulf Coast Plain, Willison Basin, and Michigan Basin do show up as intermediate attenuation while the Illinois Basin, Appalachian Basin, and other basins are not apparent. In Alaska, attenuation is generally less than the western US, but still much greater than the eastern US. In southeast Alaska, the Wrangell Volcanic Field causes a sizeable high attenuation zone. The volcanic Aleutian Mountains also have high attenuation. However, moderate to high attenuation also correlates with the tertiary sedimentary basins in Alaska. The North Slope Basin does not seem to attenuate. Thicker crust and mountain roots tend to show less attenuation, if anything, but this correspondence is most likely due to differences in temperature and seismic velocity. Heat, scattering, and young sedimentary basins create seismic attenuation in the continental crust.

Western US intraplate deformation controlled by the complex lithospheric structure

“…Using the seismically detected Moho depth 49 and LAB depth 38,42 , we deduce the lithospheric density structure by matching surface topography ("Methods" section). Additionally, we construct a 3D lithospheric effective viscosity structure based on recent seismic attenuation studies 50,51 ("Methods" section, Fig. 2g).…”

“…The lithosphere structure in our model is inferred from recent seismic studies 38,42,[49][50][51][68][69][70] . The initial lithospheric thickness model has a Moho depth model from CRUST 1.0 49 and a hybrid LAB depth model 14 derived from LITHO 1.0 38 and a body-wave tomography model 42 .…”

“…We further infer the fine-scale lithospheric effective viscosity structure in the WUS based on recent seismic attenuation studies 50,51 . Seismic attenuation represents anelastic properties of rocks 79,80 and could be linked to the steady state viscosity by linear viscoelasticity theory.…”

“…Since the seismic attenuation in the lithospheric mantle beneath the conterminous US shows regional-scale variations 51 , largely consistent with the effective viscosity variations in Cao and Liu 14 , we keep the effective viscosity in the lithospheric mantle unchanged. The fine-scale 3D crustal viscosity structure is converted from a recent crustal seismic attenuation study 50 . The crustal seismic attenuation shows the integrated effect through the whole crust.…”