[S-SS26_30PO1] Crustal Structure

[SSS26-P04] Seismic structure beneath Kyushu island, Japan, inferred from S-wavevector receiver functions.

*Takuya UEDA¹, Hiroshi TAKENAKA², Takumi MURAKOSHI³, Taro OKAMOTO⁴ (1.Kyushu Univ., 2.Okayama Univ., 3.National Defense Academy, 4.Tokyo Institute of Technology)

Keywords:receiver function, crustal structure, top of plate, Kyushu region

The underground structure of Kyushu region is characterized by active subduction of the Philippine Sea plate (PHS) beneath the Eurasian plate and several active volcanos, for example, Aso, Kirishima, and Sakurajima volcanos along with the volcanic front, and Unzen volcano located Beppu-Shimabara graben. And also there are very thick sediments at several plains in Kyushu. Therefore the seismic structure beneath Kyushu Island is seemed to be very complicated and it is very important to understand the detailed structure, especially around Moho and the top of PHS. There are many previous researches on seismic structure beneath Kyushu Island. Travel time tomography method is very useful tool for imaging the subsurface structures. In the previous works, a lot of characteristic structures are identified by the tomography for example, low velocity structure beneath volcanic front. Receiver function analysis is also very useful tool to image the seismic velocity structures. We apply it to image seismic structure on Kyushu area. In this study, we use teleseismic records from Hi-net and F-net seismic stations in Kyushu, which are supplies by the National Research Institute for Earth Science and Disaster Prevention.If those seismic stations are located at the top or in the sedimentary layer, the records include strong effect of reverberation within the sedimentary layer, which makes the image of the structure unclear. To overcome this problem, we exploit the modified S-wavevector receiver functions (SWV-RFs). The SWV-RFs are derived by deconvoluting the upgoing S-wave component with the upgoing P-wave component of the records. For suppressing the sedimentary layer effect, we apply SWV-RFs for borehole records and move virtually the seismic sensor to the top of the basement layer, and calculate the SWV-RFs at that location [Takenaka and Murakoshi, 2010]. This method needs the structure model from the surface to the sensor location. We employ the Integrated Velocity Structure Model by the Headquarters for Earthquake Research Promotion. We take several cross sections in Kyushu Island to map the calculated SWV-RFs. We then interpret the continental Moho and low velocity regions in the mapped SWV-RFs. It can be seen that characteristic low velocity regions in mantle wedge, some of which may be related to magma. We also model some SWV-RF sections by the 2.5-D finite-difference method to confirm our imaging results.