Deep-Earth Structure from PKP Precursors and Other Scattered Body Waves

Monday, 15 December 2014: 5:30 PM
Nicholas J Mancinelli and Peter M Shearer, University of California San Diego, La Jolla, CA, United States
Studies now agree that the time- and range-dependence of globally averaged PKP precursor amplitudes in the 0.7 to 2.5 Hz frequency band can be explained by weak (0.1% r.m.s.) small-scale (~6 km) velocity heterogeneity distributed throughout the lowermost mantle. Individual recordings of PKP precursors, however, exhibit significant variability about the global average. We cull from our global dataset particular regions with dense source or station coverage to look for smooth changes in precursor amplitudes. We also perform slant stacks on longer-period (1 to 10 s) precursors recorded by continental-scale arrays to distinguish between source- and receiver-side scattering.

To address the question of larger-scale heterogeneity in the lithosphere and mantle, we compare long-period seismic records (17 to 150 s) with SPECFEM3D GLOBE synthetics. We find that the data exhibit more power between the main phases (e.g., P, PP, S), which may be explained by greater intermediate-scale (50 to 200 km) heterogeneity than accounted for by 3D global tomography models. In order to confidently model the amplitude stacks of the long-period records, we test the particle-based "seismic phonon" method against computationally intensive SPECFEM3D_GLOBE solutions. For an isotropic source at 20 km depth and a fractally heterogeneous earth (bandlimited from 50 to 3000 km), the two methods produce reasonably similar synthetics. This agreement strengthens our confidence in the phonon method at periods below 30s.