Towards the detection of the mantle-reflected P wave using seismic interferometry

Abstract:

The seismic interferometry (SI) is the technique that extracts the Green's function between a station pair, by calculating the cross-correlation function (CCF) of random seismic wavefield (e.g., Snieder et al., 2013). Recent studies tried to extract the body waves by the SI. For example, some groups reported reflected P wave at the 410/660 km discontinuities (P410P/P660P) (e.g., Poli et al., 2012). This study aims to extract the P410P/P660P by applying the SI to seismic data recorded by Hi-sensitivity Seismograph Network Japan (Hi-net) and discuss the effect of the source distribution of microseisms on the retrievals.

The procedures of the CCF calculation are as follows: The dataset in this study is the vertical components of 240 Hi-net stations in Southwest Japan (2007–2018). The 1-day-length waveform eliminated the coherent logger noise, was split into 1024-second-length segments and selected based on the mean square amplitude in 0.05-0.1 Hz and 0.1–0.2 Hz. We calculated CCFs for every pair of stations by stacking over all the selected segments with spectral whitening from 0.1 to 0.2 Hz.

For discussing the effect of the source distribution, we calculated the 4th root vespagram (Rost and Thomas 2002), which showed the P410P in the 0–300 km offset and the weak P660P in the 50–100 km offset. To examine the azimuthal dependence of the retrievals, we calculated the vespagram for groups of CCFs chosen by their azimuth. Those with azimuths 0–30°, 90–120°, and 150–180°show P410P.

We compared these vespagrams with slowness distribution for events of teleseismic P-wave microseisms, inferred from Hi-net data (Nishida and Takagi, IASPEI, 2017). P410P in CCFs with azimuths 0–30° and 150–180°could originate from the sources in the Northern Atlantic Ocean and Antarctic ocean, respectively. However, P410P in CCFs with azimuths 90–120°has no corresponding sources, which imply that they originate from scattering at a deeper part.

Finally, we applied the Common Middle Point (CMP) stacking (e.g., Stein and Wysession, 2003) to estimate the depth of discontinuity with the 1D JMA seismic velocity model (Ueno et al., 2001). The CMP stacking shows continuous 410 km discontinuity and locally 660 km discontinuity. However, the CMP stacking lacks the reflection related to the Pacific plate partly due to the tilting surface.