In probabilistic seismic hazard analysis, severer ground motions than those used in usual design is very important to estimate a damage frequency of structures and components. An artificial ground motion of such a severer earthquake is required to have nonlinear response spectral properties similar to those of real motions. In this paper we propose a new simple non-parametric method to generate an artificial earthquake motion which has a target response spectrum. In this method, the artificial earthquake motion is generated in the form of the orthogonal wavelet expansion, in which the wavelet coefficients of the artificial earthquake motion are given by weighted sums of wavelet coefficients of observed earthquake motions in each frequency range. The weights of observed earthquakes are adjusted to reduce the difference between the response spectrum of the artificial earthquake motion and the target response spectrum, to a tolerance level. We apply this method to generate an artificial earthquake motion subject to the Ohsaki spectrum, from randomly selected 20 observed motions in a given range of hypocentral distance. Each of these 20 observed motions is found to contribute appreciably to the generated motion in all frequency ranges. The average and standard deviation of group delay time of the generated motion are, respectively, nearly at the central values of those for the observed motions. These results show that the artificial earthquake motions have average time-frequency characteristics of those for the observed motions.

View full abstract

The purpose of this paper is to generate a nationwide map of average shear-wave velocity (AVS) in the upper 30m. The integrated database containing the AVS data obtained from borehole logs at 1, 785 sites and the corresponding geomorphological data retrieved from the Digital National Land Information is constructed. Using the database, the relations between the AVS and the geomorphological land classification are examined. A statistical hypothesis test for the relations shows that the AVS in the central part of Japan tends to be lower than those in the northeastern and southwestern parts on certain geomorphological units. Such differences in the AVS are incorporated into a method for estimating the AVS in terms of the geomorphological data. Results indicate that the method considering the regional difference would predict the AVS and site amplification factor in all the parts of Japan more accurately.

View full abstract