The structure of the inner core inferred from short-period and broadband GDSN data

Abstract

Broadband displacement and velocity records of PKP phases are formed by the simultaneous deconvolution of the instrument responses of waveforms recorded by the short- and long-period channels of the Global Digital Seismograph Network. Revisions of standard earth models are inferred from the comparison of synthetic and observed waveforms. The synthetics incorporate a Q-model designed for the bandwidth of the data and a source model that accounts for the variations of waveform about the focal sphere due to the effects of finiteness and complexity of the rupture process. The interpretation of the data waveforms allows the position of the D-cusp to be fixed at 121 ± 1° and the C-cusp at 154 ± 2° for a surface source. The cusp positions are consistent with a P-velocity of 10.82 ± 0.02kms−1 on the underside of an inner core boundary at 1216 km and a P-velocity of 10.30 ± 0.05 km s−1 on its upper side, resulting in a P-velocity jump of Δα0.52 ± 0.07 km s−1. Based on the agreement of synthetic and observed waveforms that have been diffracted beyond the C-cusp there is no zone of small or negative P-velocity gradient or of strong anelasticity at the bottom of the outer core. Second-order features in the observed waveforms near the D-cusp suggest that the S-velocity may be nearly zero at the top of the inner core. Qα must increase with depth in the inner core in order to satisfy the pulse width of the PKP-DF phase at distances greater than 150°. In order to satisfy average properties of the inner core inferred from travel times and normal modes, these combined results require strong gradients in P-velocity, S-velocity and anelasticity in the upper 200–300 km of the inner core.