3-D SIMULATIONS OF SURFACE-WAVE PROPAGATION IN THE KANTO SEDIMENTARY BASIN, JAPAN .1. APPLICATION OF THE SURFACE-WAVE GAUSSIAN-BEAM METHOD

Authors
Citation
K. Kato et al., 3-D SIMULATIONS OF SURFACE-WAVE PROPAGATION IN THE KANTO SEDIMENTARY BASIN, JAPAN .1. APPLICATION OF THE SURFACE-WAVE GAUSSIAN-BEAM METHOD, Bulletin of the Seismological Society of America, 83(6), 1993, pp. 1676-1699
Citations number
41
Categorie Soggetti
Geosciences, Interdisciplinary
ISSN journal
00371106
Volume
83
Issue
6
Year of publication
1993
Pages
1676 - 1699
Database
ISI
SICI code
0037-1106(1993)83:6<1676:3SOSPI>2.0.ZU;2-8
Abstract
The goal of this study is to simulate the displacement waveforms of 8 sec-period Love waves observed at Tokyo and Yokohama stations in the K anto sedimentary basin during the Izu-hanto-toho-oki Earthquake of 29 June 1980 (M(o) = 7 x 10(25) dyne . cm). The surface wave Gaussian bea m method is applied for this purpose. On the basis of the 3-D seismic velocity and density structure of the Kanto basin and assuming that th e earth medium is laterally homogeneous outside the Kanto basin, wavef orms of Love waves are synthesized. The synthesized seismograms undere stimate the observed peak amplitudes at Yokohama station. This is prim arily because the station is located in the direction of the nodal pla ne of the Love-wave radiation. As indicated by Yamanaka ef al. (1992), a Quaternary basin exists in the Sagami Bay between the source locati on and the Kanto basin. We include the Sagami basin in our model by th e approximation of a circular low-velocity region. Excellent agreement between observed and synthesized waveforms was achieved not only for amplitude but also for phase for the early parts of the wave trains at both stations. We conclude that the low velocity Quaternary basin in the Sagami Bay acts like a lens to focus surface wave energy resulting in high amplitudes. The later arriving waves, in particular the long duration observed at Tokyo station, however, cannot be adequately expl ained by this method. One possible reason for the failure of simulatin g the later phases is that this method disregards the secondary Love w aves converted from S-waves and/or surface waves at a laterally discon tinuous boundary. Although the surface wave Gaussian beam method canno t adequately predict the duration of observed seismograms, it provides us with a satisfactory prediction of amplitudes and phases for early arrivals in laterally slowly-varying media at drastically lower comput ation costs and less memory requirements than does other methods.