Effect of scattering onPcP/Pamplitude ratios at Lasa from 40° to 84° distance

Attempts to study the core-mantle boundary have traditionally been frustrated by the large scattered amplitude ratios of PcP to P phases. An analysis of these phases at the Large-Aperture Seismic Array (Lasa) shows that two sources of severe scattering are the crust and upper mantle structure under the receiver and the low-velocity zone near the source. Measured amplitudes at the 21 Lasa subarray sites show that P and PcP amplitude patterns from the same event are generally uncorrelated. As a result, PcP/P amplitude ratios have large standard deviations across Lasa, typically 50% of the mean PcP/P ratios. However, P patterns for two events several degrees apart are highly correlated, as are PcP patterns. Thus amplitude fluctuations across Lasa are caused by near-receiver scattering rather than deep mantle or core structure. Steered beams were formed of P and PcP phases for 35 events in the distance range 40°–84° with depths from 0 to 609 km. The PcP/P ratios obtained from these beams show a clear trend from ∼0.4 at a depth of 0 km for explosion data to ∼0.15 at depths greater than 200 km for all distances. This depth effect cannot be explained by laterally homogeneous low-Q models of the upper mantle. When the events are normalized to Lasa mb = 6.0, the resulting PcP amplitudes show no clear depth dependence. It is thus suggested that variations in P amplitude cause most of the scatter in PcP/P ratios obtained from beamed data. A horizontally laminated low-velocity zone consisting of alternating hard and soft layers could produce the observed variation of PcP/P amplitude ratios with depth, although other scattering mechanisms cannot be eliminated.