HR: 0800h
AN: DI41A-1264    [Abstracts]
TI: Upper mantle structure beneath the Japan Islands imaged by receiver function -With Hi-net tiltmeter data-
AU: * Tonegawa, T
EM: tonegawa@eps.nagoya-u.ac.jp
AF: Nagoya Univ, Furo-cho, Chikusa-ku, Nagoya, 464-8602 Japan
AU: Hirahara, K
EM: hirahara@eps.nagoya-u.ac.jp
AF: Kyoto Univ, Sakyo , Kyoto, 606-8502 Japan
AU: Shibutani, T
EM: shibutan@rcep.dpri.kyoto-u.ac.jp
AF: Kyoto Univ, DPRI, Gokasyo, Uji, 611-0011 Japan
AU: Shiomi, K
EM: shiomi@bosai.gp.jp
AF: NIED, 3-1, Tenodai, Tsukuba, 305-0006 Japan
AB: In the Japan subduction zone where the Pacific plate is subducting, we can investigate the dynamics of the upper mantle beneath the subduction zone by imaging the detailed structure beneath the Japan Islands. Recently, the undulation of the seismic discontinuities in the upper mantle beneath the Japan Islands has been well detected by receiver function (RF) analysis based on the broadband seismograms (e.g. Tonegawa et al., 2005). However, the density of broadband stations expanding throughout the Japan Islands is still insufficient (130 stations in Japan) to investigate the relation between the slab and 410 or 660 km discontinuities. The high sensitivity accelerometer (hereafter tiltmeter) network has recently been deployed with high-density spacing (700 stations in Japan) by NIED, and these tiltmeters can be used as long-period seismometer. If we can use the recordings obtained by the tiltmeter to disclose the structure of the upper mantle, it is possible to dramatically improve our knowledge on the dynamics of the upper mantle. In this study, we examine whether these recordings are available to demonstrate the seismic discontinuities in the upper mantle by calculating the RF from tiltmeter recordings. We select a tiltmeter station and a F-net (broadband network expanding in the Japan Islands) station, whose distance is approximately 4 km, and compare the recordings with velocity seismogram by changing the frequency band. Those RFs would be similar when applying the low-pass filter of longer than 6 sec. Based on this trial, we decide to set the low-pass filter of 6 and 10 sec, when producing the RFs from the tiltmeter recordings. Since tiltmeter seismograms have only the horizontal recordings, we need to obtain the vertical component or sourcetime function. Therefore, we make the sourcetime function by stacking all of vertical components obtained by F-net broadband stations, that is, we consider these stacked waveforms as sourcetime functions. For making the transect, assuming all later phases of the direct P wave are the Ps phases converted at depths, we migrate the time domain RF to the depth domain RF by referring to IASP91 velocity model. As a result, we obtain approximately 50,000 RFs with good SN ratio, though Tonegawa et al (2005) could use only less than 10,000 RFs. Moreover, when making the transect along a line, in contrast to a width of ñ200 km used in Tonegawa et al (2005) because of the low density spacing of the broadband station, this study can realize the clear imaging down to 1,200 km depth with a width of ñ40 km. The top surface and the lower boundary of the descending slab can be imaged down to 400 km and 600 km depth, respectively. The 410 km and the 660 km discontinuities are clearly traced, and our result also shows the depression of the 660 km discontinuity affected by the cold material of the stagnant slab. However, the undulation of the 410 km discontinuity affected by the slab penetration cannot be imaged. The distortion from the penetration of the slab to the 410 km discontinuity seems to be limited to the small scale compared to that from the stagnation of the slab to 660 km one, and hence the RF with longer than 6 sec has presumably insufficient resolution to detect such a fine structure.
DE: 1213 Earth's interior: dynamics (1507, 7207, 7208, 8115, 8120)
DE: 7208 Mantle (1212, 1213, 8124)
DE: 7240 Subduction zones (1207, 1219, 1240)
SC: Study of Earth's Deep Interior [DI]
MN: Fall Meeting 2005