No abstract

[1] We estimate the spatial variation of the seismic parameter t* using teleseismic (epicentral distance = 30°-85°) P wave spectra of about 200 deep (focal depths > 200 km) earthquakes recorded by 378 broadband seismometers in the United States and Canada. Relative P wave spectral ratios up to 1 Hz for about 63,000 station pairs with high signal-to-noise ratio and impulsive P waveforms are inverted for t* P by least squares inversion. The continental-scale t* P pattern correlates to the age of geological terrains and the seismic, heat flow, gravity, and magnetic variations across North America. Predominantly low values of t* P are obtained in stable central North America (SNA), and high t* P values are obtained for stations in the tectonically active western part of the continent (TNA). This variation is similar to that observed previously in short-period amplitude anomalies, spectral ratio variations, and ScS reverberations. On average, we resolve a contrast in t* P between SNA and TNA of about 0.2 s. We resolve regional variations in t* P , which correlate with tectonics. Relatively low t* P is associated with currently active subduction below Alaska. Relatively high t* P is found in SNA below the Appalachians and the Gulf Coast. The consistency between t* P and tectonics suggests that the observed variations in t* P are, on the scale of around 200-500 km, predominantly due to intrinsic attenuation. The similar patterns in t* P and predicted values for a recent global attenuation model confirm this further. The compatibility with the t* P computed for attenuation estimated via a thermal interpretation of shear wave velocity anomalies illustrates that variations in seismic velocity are predominantly due to physical effects with a strong attenuation signature, most likely temperature or a combination of temperature and water content.

No abstract

No abstract

S U M M A R YSince W. J. Morgan proposed that intraplate volcanism at some Pacific hotspots is caused by hot plumes rising from the lower mantle, geophysicists have been actively pursuing physical evidence for mantle plumes. Several seismic studies have mapped low-velocity anomalies below a number of hotspots. However, the association of low-velocity structures with plume tails has remained controversial given the debate on whether lower-mantle plumes impart observable traveltime or amplitude perturbations on seismic waves. Using high-resolution numerical simulations of plume ascent through the mantle and their effects on waveforms, we demonstrate that the delay of shear waves by plume tails at depths larger than 1000 km are immeasurably small (<0.2 s) at seismic periods commonly used in waveform analysis. Therefore, we conclude that narrow lower mantle plumes are not detectable.