11:30 AM - 11:45 AM
[SSS12-04] Teleseismic P and S wave travel-time residuals at the seafloor S-net stations
Keywords:S-net, teleseismic events, relative travel-time residuals
The occurrence of interplate earthquakes in subduction zones is considered to be controlled by the stress field and structural heterogeneity at the plate boundary. However, due to the lack of seismic network in the oceanic areas, few studies have investigated the effects of structural heterogeneity in the subslab mantle. After the great 2011 Tohoku-oki earthquake (Mw 9.0), the National Research Institute for Earth Science and Disaster Resilience has installed a seafloor observation network for earthquakes and tsunamis along the Japan Trench (the S-net) in the East Japan forearc region, which provides us with an excellent opportunity to study shallow and deep structures beneath the forearc region.
Fan & Zhao (2021) determined 3-D P-wave tomography beneath the Japan Islands using arrival times of local and teleseismic events recorded at the Hi-net (High Sensitivity seismograph Network in Japan) stations and local earthquakes recorded at the S-net stations. They revealed sub-slab low-velocity anomalies (SLVAs) due to the inclusion of arrival times of some deep earthquakes recorded by the S-net in their dataset, which improved the resolution of the sub-slab structure. The SLVAs may reflect hot and buoyant upwelling from the deep mantle and may influence the stress field at the plate boundary and the earthquake generation. The arrival times of teleseismic events at the S-net are necessary to further constrain the deep structure beneath the ocean. In this study, for the first time, we measure a large amount of teleseismic P and S wave travel-time residuals at the seafloor S-net stations.
We selected teleseismic events (M > 6.5) recorded at 150 S-net stations during January 2017 to October 2022 with epicentral distances of 30-90°. Using the code AIMBAT (Automated and Interactive Measurement of Body-wave Arrival Time) developed by Lou et al. (2013) on the basis of the multi-channel cross-correlation technique (VanDecar & Crosson, 1990), we analyzed 48 teleseismic events and obtained 4156 P-wave and 4174 S-wave relative travel-time residuals. Then we calculated the mean relative travel-time residuals at the S-net stations in two ways. First, we averaged the relative travel-time residuals for all the teleseismic events recorded at each station. The residuals are negative near the Japan trench and positive off northern Tohoku. Next, the teleseismic events in four quadrants (NE, NW, SE and SW) are analyzed separately, and we calculated the mean relative residuals for each quadrant. The S-net stations off Kanto show large negative residuals for the SW events.
The S-net arrival times are affected by the seafloor sedimentary layer. Azuma et al. (2019) determined P and S wave travel-time corrections for each S-net station using the PS-converted wave at the bottom of the sedimentary layer. To evaluate the effect of the sedimentary layer on the relative travel-time residuals, we corrected the raw residuals with the station corrections. The changes in the pattern of the mean relative residuals before and after the correction are small, suggesting that the relative residuals mainly reflect the influence of crustal and mantle structures. In particular, the large negative residuals near the Japan trench may reflect the influence of the high seismic velocity of the Pacific plate. We think that the large negative residuals off Kanto for the SW events are mainly due to the large dip angle of the subducting Pacific slab to the south of Kanto, where the teleseismic rays travel longer within the slab.
We plan to conduct joint inversion of the local and teleseismic P and S wave travel-time data recorded at both the S-net and Hi-net stations, so as to image the detailed 3-D mantle structure beneath Japan and surrounding regions.
Fan & Zhao (2021) determined 3-D P-wave tomography beneath the Japan Islands using arrival times of local and teleseismic events recorded at the Hi-net (High Sensitivity seismograph Network in Japan) stations and local earthquakes recorded at the S-net stations. They revealed sub-slab low-velocity anomalies (SLVAs) due to the inclusion of arrival times of some deep earthquakes recorded by the S-net in their dataset, which improved the resolution of the sub-slab structure. The SLVAs may reflect hot and buoyant upwelling from the deep mantle and may influence the stress field at the plate boundary and the earthquake generation. The arrival times of teleseismic events at the S-net are necessary to further constrain the deep structure beneath the ocean. In this study, for the first time, we measure a large amount of teleseismic P and S wave travel-time residuals at the seafloor S-net stations.
We selected teleseismic events (M > 6.5) recorded at 150 S-net stations during January 2017 to October 2022 with epicentral distances of 30-90°. Using the code AIMBAT (Automated and Interactive Measurement of Body-wave Arrival Time) developed by Lou et al. (2013) on the basis of the multi-channel cross-correlation technique (VanDecar & Crosson, 1990), we analyzed 48 teleseismic events and obtained 4156 P-wave and 4174 S-wave relative travel-time residuals. Then we calculated the mean relative travel-time residuals at the S-net stations in two ways. First, we averaged the relative travel-time residuals for all the teleseismic events recorded at each station. The residuals are negative near the Japan trench and positive off northern Tohoku. Next, the teleseismic events in four quadrants (NE, NW, SE and SW) are analyzed separately, and we calculated the mean relative residuals for each quadrant. The S-net stations off Kanto show large negative residuals for the SW events.
The S-net arrival times are affected by the seafloor sedimentary layer. Azuma et al. (2019) determined P and S wave travel-time corrections for each S-net station using the PS-converted wave at the bottom of the sedimentary layer. To evaluate the effect of the sedimentary layer on the relative travel-time residuals, we corrected the raw residuals with the station corrections. The changes in the pattern of the mean relative residuals before and after the correction are small, suggesting that the relative residuals mainly reflect the influence of crustal and mantle structures. In particular, the large negative residuals near the Japan trench may reflect the influence of the high seismic velocity of the Pacific plate. We think that the large negative residuals off Kanto for the SW events are mainly due to the large dip angle of the subducting Pacific slab to the south of Kanto, where the teleseismic rays travel longer within the slab.
We plan to conduct joint inversion of the local and teleseismic P and S wave travel-time data recorded at both the S-net and Hi-net stations, so as to image the detailed 3-D mantle structure beneath Japan and surrounding regions.