EVALUATION OF LOCAL SITE EFFECTS AND THEIR REMOVAL FROM BOREHOLE RECORDS OBSERVED IN THE SENDAI REGION, JAPAN

We present results from a borehole observation network composed of 12 borehole sites deployed in a 20 X 20 km area in the Sendai region, Jap an. The subsurface Quaternary layers vary with each site, but a Plioce ne layer is commonly underlying at a depth of 0 to 80 m throughout the Sendai region. We define this Pliocene layer, whose S-wave velocity i s larger than 500 m/sec and whose N value of the standard penetration test is greater than 50, as an engineering bedrock in the region. Once we characterize ground motions at the engineering bedrock, we can eit her use these motions directly in the response analysis of whole soil- building systems or use them in the nonlinear analysis of soft soil la yers as an input. The purpose of this study is to evaluate the local s ite effects due to surface layers overlying the engineering bedrock an d to remove them by using one-dimensional(1D) soil models whose proper ties are determined by weak-motion records of 18 events. First we iden tify S-wave velocities and frequency-dependent quality factors 12 from amplification factors between surface records and borehole records ob served at 10 sites at depths of several tens of meters. The identified S-wave velocities are very close to S-wave logging values. The identi fied Q values show strong frequency dependence, proportional to f((0.4 6 to 1.16)), where f is frequency. We confirm that the observed amplif ication factors in the frequency range from about 0.1 to 20 Hz can be explained well by the theoretical ones based on the ID wave propagatio n theory with the identified S-wave velocities and Q. Then we estimate the so-called engineering bedrock waves, which are supposed to be obs erved on the outcrop of the engineering bedrock, from borehole records by using the ID models with these identified soil constants. We confi rm that local site effects due to surface layers overlying the enginee ring bedrock are properly removed so that we have similar characterist ics in the resultant engineering bedrock waves from records at differe nt sites.