Seismic wavefield imaging of long-period ground motion in the Tokyo metropolitan area, Japan

摘要: Long-period ground motions due to large earthquakes can cause devastating disasters, especially in urbanized areas located on sedimentary basins. To assess and mitigate such damage, it is essential rapidly evaluate seismic hazards for infrastructures, which be simulated by response analyses that use waveforms at the base of each infrastructure as an input motion. The present study reconstructs wavefield Tokyo metropolitan area Kanto basin, Japan, from seismograms Metropolitan Seismic Observation network (MeSO-net). obtained fully explains observed frequency band 0.10–0.20 Hz. This attributed imaging technique proposed Kano et al. [2017], implements replica exchange Monte Carlo method simultaneously estimate model parameters related subsurface structure source information. Further investigation shows reconstructed lower than 0.30 Hz high quality terms variance reduction, quantifies a misfit waveforms, but reduction worsens higher frequencies. Meanwhile, velocity spectra show good agreement with observations up 0.90 combined goodness-of-fit, measure spectra. Inputting into analyses, we overall damage infrastructures immediately after earthquake.