Distribution of the measured Sdiff travel times

We collected 1796 SKS, 1729 SKKS and 3861 Sdiff travel times which sample the Pacific region. In order to measure the travel time anomalies, synthetic waveforms from the 1D model (PREM) are first created for each trace. The anomalies are obtained by taking the cross correlations between PREM synthetic and observed waveforms. The waveforms are bandpass filtered between 100 and 17 second.

Figure 2.61(a) shows a distribution of measured Sdiff travel time anomalies with respect to PREM. Travel time anomalies are plotted at the midpoints of the diffracting portions of the Sdiff phases. The anomaly distribution has a good correlation with the tomographic model in the D'' layer. The figure indicates that the S wave structure in the D'' layer primarily contributes to the observed Sdiff anomalies.

Figure 2.61 (b) through (d) show the Sdiff travel time anomalies with respect to azimuth or back azimuth for some selected events or stations. Observed travel times of SKS and SKKS phases and synthetic travel times of Sdiff, which are calculated by ray theory, are also plotted. Lack of correlations of the travel time anomalies between Sdiff and other phases indicates that the Sdiff travel time anomalies are due to heterogeneities within the lower mantle. This is because the paths of Sdiff and SKKS are close to each other in the upper mantle but they are different in the lowermost mantle. The paths of SKS and Sdiff are more separated compared to Sdiff and SKKS in the upper mantle; however, the lack of correlation between SKS and Sdiff travel time anomalies can still indicate if the Sdiff anomalies are caused by near source or station structure or the lower mantle structure.

Figure 2.61 (b) shows that trends of Sdiff travel time anomalies are well predicted in the Northern Pacific. The paths sample the border of the Pacific superplume. We have previously reported that ray theory gives larger positive travel time anomaly estimations compared to a more exact method, such as the spectral element method, because of the lack of the finite frequency effects. With the finite frequency corrections, the positive anomalies of synthetic travel times become a few seconds smaller from what is shown in the figures. Figure 2.61(c) shows the data set with a rather steep change in Sdiff travel time anomalies with respect to azimuth, observed in the central Pacific. The paths sample inside the Pacific superplume. They indicate a possibility that the superplume is a gathering of multiple separated slow regions rather than a single big blob. Figure 2.61(d) shows one of the cases where the model over predicts the travel time anomalies.

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: www@seismo.berkeley.edu
© 2007, The Regents of the University of California