From ambient vibration data analysis to 1D ground-motion prediction of the Mj 5.9 and the Mj 6.5 Kumamoto earthquakes in the Kumamoto alluvial plain, Japan

Abstract We present horizontal ground motion predictions at a soft site in the Kumamoto alluvial plain for Mj 5.9 and 6.5 earthquakes of April 2016, framework an international blind prediction exercise. Such were obtained by leveraging all available information which included: (i) analysis earthquake motions; (ii) processing ambient vibration data (AMV); (iii) 1D response analysis. Spectral ground-motion used to obtain empirical estimates amplification function, with evidence peak about 1.2 Hz. Horizontal-to-vertical spectral ratio AMV confirmed this resonance frequency pointed out also low-frequency around 0.3 Hz site. then processed cross-correlation, modified spatial autocorrelation high-resolution beamforming methods retrieve shear-wave velocity (Vs) structure joint inversion surface-wave dispersion ellipticity curves. The use low curve allowed reference Vs profile down few thousand meters depth was perform equivalent-linear simulations M event, both nonlinear event target Adopting quantitative goodness-of-fit metrics based on time–frequency representation signals, we fair-to-good agreement between observations poor earthquake. In terms acceleration spectra, while overpredictions simulated motions severely underestimate observations, especially those approach. Graphical