Abstract

The most direct method for constraining physical properties of the outermost core involves analysis of S nKS waveforms (where n is an integer). However, the physical nature of this region remains ambiguous in spite of its significance in geodynamic, geomagnetic, and seismic models of the Earth's deep interior. Global inversion for P-wave velocity in the outermost 200 km of the core from SKS and SKKS waveforms is examined here. The inversion process consists of constructing synthetic seismograms using normal mode theory, and solving for first-order perturbations to P-wave velocity. Spheroidal modes with periods between 33 and 100 s are chosen to model the waveforms and P-wave velocity perturbation is solved along individual raypaths, assuming a laterally homogeneous initial model. The dataset includes about 800 digital, long-period radial seismograms from earthquakes which have occurred globally. Seismograms were chosen for source-receiver distances of 110–130° in which SKS and SKKS are best isolated in time from nearby phases. We have attempted to remove the effects of mantle heterogeneity by incorporating the mantle velocity model MDLSH. Figures of P-wave velocity results, plotted at the midpoint between source and receiver, show large-scale patterns of positive and negative lateral velocity variation. There are also regions of inconsistencies, not simply explained by core-mantle boundary (CMB) topography or crisscrossing raypaths. There is no clear dependence of residuals on latitude. Incorporation of a modified version of MDLSH does not significantly change our solutions, suggesting that the resolution length scales of global mantle models are too large to remove important smaller-wavelength (less than 1000 km) mantle heterogeneity effects. Furthermore, raypaths of SKS and SKKS independently go through laterally different structure relative to normal mode wavelengths, starting well above the CMB. These results suggest that simultaneous waveform inversion for P- and S-wave velocities is a more reliable way of constructing a model of outermost core structure.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call