@ ŒfÚŽGŽi˜ajF “ú–{’nkHŠwƒVƒ“ƒ|ƒWƒEƒ€˜_•¶W VolF 9-1Šª ”NF 1994”N •ÅF 553-558•Å ’˜ŽÒi˜ajF ãp L”VC@£”ö@˜a‘åC@²ŠÔ–ì@—²Œ› ƒ^ƒCƒgƒ‹i˜ajF ‹ß’n‘å’nk‚ð‘ÎÛ‚Æ‚µ‚½”¼ŒoŒ±“I‹­k“®—\‘ª‚ÌŽŽ‚Ý ´˜^i˜ajF
- ƒL[ƒ[ƒhi˜ajF - ŒfÚŽGŽi‰pjF PROCEEDINGS OF THE JAPAN EARTHQUAKE ENGINEERING SYMPOSIUM ’˜ŽÒi‰pjF Nobuyuki AGATA, Kazuoh SEO, Takanori SAMANO ƒ^ƒCƒgƒ‹i‰pjF A TRIAL OF SEMI-EMPIRICAL PREDICTION OF STRONG GROUND MOTIONS DURING A VERY NEAR AND LARGE EARTHQUAKE ´˜^i‰pjF
A semi-empirical method to evaluate strong ground motion at a certain site will be available when we already have an observed weak motion at the same site. But if we want to predict a strong motion for a very near and large earthquake, we face a question that "can an observed motion be an adequate Green's function for the prediction?". Then what should be improved in the procedure was examined in this paper. We pointed out two important things. One is the selection of the most adequate motion as an empirical Green's function among the observed weak motions. In such case, that long-period surface waves are expected in a predicted strong motion, a weak motion as an element must include such components. The other is the number of summation in the process. Excess number of summation will make the result unreliable. By taking such things into account, we tried the evaluation of strong ground motions during the 1923 Kanto earthquake. The results showed a good agreement when the 1990 Izu-Oshima-Kinkai earthquake was regarded as the element, although we need further improvement in the procedure. ƒL[ƒ[ƒhi‰pjF - ‹LŽ–‹æ•ªF - ‹æ•ª @@@ˆÏˆõ‰ï˜_•¶W