S43B-2785
Strong Ground Motion Simulation for the Mw5.9 Central Chiba Earthquake of JULY 23, 2005 with Pseudo Point-Source Model

Thursday, 17 December 2015
Poster Hall (Moscone South)
Yosuke Nagasaka and Atsushi Nozu, Port and Airport Research Institute, Kanagawa, Japan
Abstract:
The pseudo point-source model is a source model for strong ground motion simulation proposed by Nozu(2012), originally for megathrust earthquakes. In this model, a simple source spectrum that follows the omega square model is given to each subevent without considering detailed spacio-temporal slip distributions; thus, the model is simpler compared to the characterized source model in which subevents are divided into subfaults. In the pseudo point-source model, the Fourier amplitude at a target station is evaluated by multiplying the path effect and the site amplification factor with the source spectrum explained above. Then, the waveform in the time domain is synthesized by inverse Fourier transform to the amplitude combined with phase characteristics properly selected from past records at the same station. The model has successfully been applied to several earthquakes in Japan including shallow crustal, subduction and intraslab earthquakes.

In this study, the pseudo point-source model was applied to the Mw5.9 Central Chiba earthquake of JULY 23, 2005 to check its validity for intraslab earthquakes occurring beneath the metropolitan area and to find possible error sources.

The overall result showed good agreement with the observations in terms of both Fourier spectrum and velocity time history (Figure); however, some unfavorable trends were found in the results. One was that the Fourier spectrum was underestimated at farther stations and at higher frequencies over 2 Hz. The discrepancy was significantly mitigated by revising the Q value from 100×f0.7 to 110×f1.0. The result indicates that the Q value has to be determined carefully when simulating ground motions in a wide area where the hypocentral distance exceeds 200 km. Another trend was the overestimation around 1 Hz at stations to the north of the epicenter. This could probably be attributed to the backward directivity effect or the variation in the site amplification factors and should be investigated further in the future study.