Shear wave travel time, amplitude, and waveform analysis for earthquakes in the kurile slab: Constraints on deep slab structure and mantle heterogeneity

Abstract

Shear wave travel times, amplitudes and broadband waveforms from 16 events in the Kurile slab are analyzed to place further constraints on the seismic velocity heterogeneity associated with the subducted oceanic lithosphere. Several previous studies have proposed that the Kurile slab penetrates into the lower mantle with little deformation other than a steepening of the dip angle, while other studies suggest that strong slab deformations occur with perhaps negligible lower mantle penetration. In an effort to detect any deep slab effects, we compare anomaly patterns covering a 60° azimuthal range straddling the strike of the slab for events in three different depth intervals (> 500 km, 350–500 km, and 100–200 km). S and ScS pulse broadening is used as a measure of waveform complexity. Complexity of broadband shear waves does not vary systematically across our azimuthal range as the propagation direction changes with respect to the postulated lower mantle slab model for this region. Strong residual sphere travel time variations are found for shear waves from events at all depths; however, these variations do not display the depth dependence apparent in previous P wave analysis that has been interpreted as requiring a change in slab geometry and lower mantle penetration. The same S wave residual pattern is apparent in sS depth phases, and is well predicted by a recent model of shear velocity structure beneath North America. A combination of deep mantle and near‐receiver aspherical structure can explain most of our observed shear wave travel time patterns, with only a minor source depth‐dependent component remaining after correction. Individual amplitude and travel time measurements are poorly correlated; however, ScS/S amplitude ratios and ScS‐S differential times are significantly correlated, consistent with geometric focussing and defocussing effects somewhere along the raypaths. The travel time behavior suggests that the responsible velocity heterogeneity is not located near the source. The absence of a strong deep slab effect on the shear waves indicates that deep slab heterogeneity is less pronounced for shear waves than expected on the basis of existing P wave models derived from travel time residual sphere analysis. Various interpretations include: estimates of δVs/δT and compositional effects previously used to predict shear velocity structure of the deep Kurile slab are in error, with actual lower mantle slab anomalies being much weaker than proposed; the Kurile slab is so distorted in the lower mantle that no source depth dependent trends are produced for our limited focal sphere coverage; or the Kurile slab does not penetrate into the lower mantle.