Estimation of strong ground motion in broad-frequency band based on a seismic source scaling model and an empirical Green's function technique

Abstract

We introduce a generalized method for simulating strong ground motion from large earthquakes by summing subevent records to follow the ?2 law. The original idea of the method is based on a constant stress parameter between the target event and the subevent. It is applicable to a case where both events have a different stress drop after some manipulation. However, the simulation for a very large earthquake from a small event with this method has inevitably some deficiencies of spectral amplitudes in the intermediate frequency range deviating f`rom the ?2 model, although the high and low frequency motions match the scaling. We improve the simulation algorithm so as not to make spectral sags, introducing self-similar distribution of subfaults with different sizes in the fault plane, so-called fractal composite faulting model. We show successful simulations for intermediate-sized earthquakes (MJMA = 5.0, 6.0 and 6.1), the large aftershocks of the 1983 Akita-Oki earthquake. using the records of smaller aftershocks (MJMA = 3.9 and 5.0) as an empirical Green's function. Further, we attempted to estimate strong ground motion for the 1946 Nankai earthquake with Mw 8.2, using the records of a MJMA 5.1 earthquake occurring near the source region of the mainshock. We found that strong ground motions simulated for the fractal composite faulting model with two asperities radiating significantly high frequency motions matched well the observed data such as the near-field displacement record, the source spectrum estimated from the teleseismic record, and the seismic intensity distribution during the 1946 Nankai earthquake.