第173回ジオダイナミクスセミナー
   Geodynamics Seminar

"Investigation of deep seismic structures and crustal waveguides in Japan by seismic tomography and waveform modeling"
     
講師:Mohamed F. Abdelwahed(GRC学術創成研究員)
   主催 : 愛媛大学地球深部ダイナミクス研究センター
   日時 : 2007年2月9日(金)17:00〜
   場所 : 愛媛大学理学部講義棟 101室

              

要 旨
  We determined 3-D P-wave velocity structure down to 700 km depth under the Japan Islands using a large number of arrival time data (~241000 arrivals) from local and teleseismic events simultaneously. Our tomographic model revealed some new features. The Philippine Sea slab is found to subduct down to 500 km depth under southwest Japan though the seismicity within the slab ends at 150-200 km depth. A new proposal for the Philippine Sea slab is presented. Significant low-velocity anomalies are found to exist in the deep portion of the mantle wedge above the Pacific slab, which may be caused by the deep dehydration process of the slab. Slow anomalies are detected in the mantle beneath the Pacific slab, which may be caused by a mantle plume or upwelling of hot mantle materials associated with the deep subduction of the Pacific slab and its collapsing down to the lower mantle.

A genetic waveform modeling approach is introduced. The Generalized Ray Theory (GRT) and the micro genetic algorithm, micro-GA (Carol 1996) are used to fulfill the optimum condition for solving the crustal waveguide more reliably. The GRT is used to construct the required synthetic seismograms. The genetic algorithm is used as the searching tool to optimize the best-fitting of observed-synthetic seismograms for the best velocity model. Sensitivity tests have been done to insure the efficiently of the technique. The high-frequency local waveforms are found well represented in synthetics and fitted well with observed data. The genetic conversion curves and the well-fitted waveforms show the reliability of the technique. The technique is applicable to solve the high-frequency waveform of local earthquakes in Japan. A good waveform matching is achieved and the resulted models are consistent with the average velocity model in Japan.

A new approach for later phase identification, Cross-Correlation Phase Identifier (CCPI) technique is also introduced in this study. Based on the above modeling technique and a multi-cross correlation scheme, the respected later phases arrival times are estimated. Application of the technique for local earthquake in Japan is presented.

Practical presentations of the computer codes used in this study are given.


        問い合わせ先:土屋 卓久  TEL   (089)927-8198
                       E-mail  takut@sci.ehime-u.ac.jp