Core Phases Observed with AlpArray

In most seismic tomographic models, the first P and/or S wave data generated by regional and teleseismic events are used to conduct tomographic inversion. Despite the abundance and precise measurement of the first body wave arrival times, the non-uniform distribution of their ray path leads to a lower resolution in the mantle below 1000km in depth. Curiously, there are particularly few ray paths sampling the lowermost mantle below dense seismic arrays, due to the limited incidence angle range of P and S waves. Previous studies have demonstrated the importance of core phases, resulting from reflection and/or conversion of seismic waves at the core discontinuities, in seismic tomography by improving the ray path coverage and constraining the structures in the lower mantle. Therefore, adding core-grazing phases (Pdiff, Sdiff) as well as core phases (e.g. PKP, PKIKP, SKS) in tomography could deliver high-resolution tomographic images of deep mantle structures in poorly resolved regions and may even reveal undiscovered features.

To increase the topographic resolution in the Alpine region, the AlpArray Initiative deployed about 250 temporary stations alongside the local permanent stations in the European Alps forming a greater AlpArray seismic network. This large-scale network provides a dense sampling rate and high-quality seismic data across the region, which gives us a unique opportunity to observe core phases coming from all directions in such a large aperture. We investigate the visibility of core phases observed with AlpArray and find that it is uniquely suited to observe high order core phases (P’P’, PcPPcPPKP, PKPPKPPKP) from sources in Alaska, Japan, and Sumatra in a distance range of 60-110 degrees. We show some array processing methods to improve the resolution of seismic observation and examine the waveforms in different frequency ranges. We find significant deviations in core phase amplitudes from predictions which are most likely linked to other structures directly above the core mantle boundary and can serve to test tomographic models in this depth region. The insight gained from this modelling is used to discuss the usability of core phases in future tomographic studies.