Anisotropy in the Northwest Pacific Subduction Zone from Shear-Wave Splitting Analysis Based on S-net Seismic Data

The Northwest Pacific subduction zone, characterized by complex plate interactions and active tectonics, is a key area for geodynamics research. These tectonic movements generate seismic anisotropy, causing shear waves to split into orthogonal fast and slow components. Analyzing shear-wave splitting at surface stations allows for inferring fast directions and splitting times, offering insights into slab deformation, mantle flow, and stress field during subduction. The S-net seafloor observation network provides an ideal setup for studying anisotropy within the subducting slab. Using high-SNR S-net stations near trenches, we focus on anisotropy within subducting slabs and sub-slab mantle, excluding influences from overriding slabs and mantle wedges.       

This study focuses on the anisotropy of two major subduction zones: the Japan Trench and the Izu-Bonin subduction zone. For the Japan Trench subduction zone, 11 S-net stations located east of the trench and 8 seismic events in the Japan Sea were selected. The events, with magnitudes of 3.6 < M_JMA< 4.4, and focal depths of 373.8–444.78 km, had ray path lengths of approximately 853.19–1138.50 km, with only a small portion propagating through the sub-slab mantle. Using the minimum eigenvalue minimization and waveform rotation cross-correlation methods, 20 reliable shear-wave splitting measurements were obtained with a predominant fast direction of NNW-SSE, splitting times ranging from 0.1–0.86 s (average 0.363 s, median 0.32 s), and anisotropy intensities of 0.002%–0.017% (average 0.008%).

For the Izu-Bonin subduction zone, 16 S-net stations at its northern end and 5 seismic events from its central and southern segments were analyzed. The events have magnitudes of 4.1 < M_JMA < 5.6 and focal depths of 399–464 km. The ray path lengths are within 712–1101 km. The splitting measurements on different rays are classified into two types based on the length of sub-slab paths: 1) for those smaller than 222 km, 17 reliable measurements are obtained with the predominant fast direction of NNW-SSE, splitting times of 0.08–0.6 s (with an average of 0.226 s and a median of 0.18 s), and anisotropic intensities of 0.001%–0.02% (with an average of 0.006%); 2) for those greater than 222 km, 9 reliable measurements are obtained with the predominant fast direction of NNW-SSE, splitting times of 0.12–0.86 s (with an average of 0.34 s and a median of 0.34 s), and anisotropic intensities of 0.003%–0.02% (average 0.008%).

According to paleomagnetic studies, the paleo-spreading direction of the western Pacific Plate was NNW-SSE, consistent with the fast directions obtained from the three types of results in this study. This alignment suggests that the anisotropy within the subducting slab primarily originates from "fossil" anisotropy retained during the slab's formation and subduction. Since these rays sample more sub-slab mantle paths, they carry more sub-slab mantle anisotropy characteristics, indicating that the anisotropy intensity in the sub-slab mantle is greater than the "fossil" anisotropy preserved within the subducting Pacific Plate. The wide range of splitting times across the three types of results reflects the heterogeneous nature of anisotropy in the region.