4:30 PM - 4:45 PM
[SSS11-05] Toward Development of Spectral Inversion Method Considering with Rupture Directivity
Keywords:Spectral inversion, Rupture directivity, Source spectrum
Source, propagation path, and site effects were separated from strong motion records observed during the 2016 Kumamoto earthquake sequence using the spectral inversion method (Somei et al., 2019). Somei et al. (2020) have obtained a source spectrum for a specific station from an observed Fourier amplitude spectrum by removing the propagation path and site effects, in order to clarify the factor of variation in residual of spectral inversion results. For several events (Mw: 3.6-5.3) in the 2016 Kumamoto earthquake sequence, the azimuthal dependence of both the corner frequencies and fall-off rate was found in the displacement source spectra at stations. They thought this azimuthal dependence was caused by rupture directivity. In this study, we estimated the rupture directivity coefficient based on the variation in source spectra among stations toward developing a new spectral inversion method considering with the rupture directivity.
The rupture directivity coefficient assuming the bilateral line source model (e.g., McGuire, 2004) was estimated from the variation in the fall-off rate of the displacement source spectra among stations. The S-wave velocity was assumed to be 3.4 km/s in the rupture directivity coefficient. The take-off angle between the source and station was calculated using the 1D velocity structure model of JMA2001 (Ueno et al., 2002). The model parameters in rupture directivity coefficient are horizontal and vertical rupture directions, rupture velocity, and directivity ratio (i.e., percentage of unilateral rupture). These parameters were determined by grid searching to minimize the misfit function (Boatwright, 2007). In this misfit function, Boatwright (2007) originally used the ratio of the peak ground motion (PGA and PGV) for each station to that expected from the ground motion prediction equation, in order to estimate the directivity coefficient. We used the ratio of the fall-off rate for each station to -2 (i.e., the fall-off rate of the ω-2 source spectral model).
For an example of an Mw5.3 event at 17:52 on 19 April 2016 (JST) occurring in southwest of aftershock area, the estimated rupture directivity coefficient showed that the horizontal and vertical rupture directions are N214° E and 135° from vertically down, respectively, with a rupture velocity of 2.4 km/s. The best directivity ratio was 1.0, which means the rupture propagates unilaterally. On the other hand, the strong motion generation area for this event modeled by simulating the observed ground motions using the empirical Green’s function method showed the rupture propagated southwest and shallow direction from the hypocenter (Somei et al., 2020). These results indicate that rupture directivity coefficient estimated from the variation in the fall-off rate was reasonable though the rupture directivity coefficient was upon the assumption of the line source model. Based on the rupture directivity coefficient, we will develop a new spectral inversion method considering with the rupture directivity.
Acknowledgements: The strong motion data were provided from the K-NET, KiK-net, and F-net operated by the National Research Institute for Earth Science and Disaster Resilience (NIED), Japan Meteorological Agency (JMA), Kumamoto Prefectural Government, Kagoshima Prefectural Government, and Nagasaki Prefectural Government. This study was supported by JSPS KAKENHI Grant Number JP20K05044. and by achievements of the collaborative research program (2020H-01) of the Disaster Prevention Research Institute of Kyoto University.
The rupture directivity coefficient assuming the bilateral line source model (e.g., McGuire, 2004) was estimated from the variation in the fall-off rate of the displacement source spectra among stations. The S-wave velocity was assumed to be 3.4 km/s in the rupture directivity coefficient. The take-off angle between the source and station was calculated using the 1D velocity structure model of JMA2001 (Ueno et al., 2002). The model parameters in rupture directivity coefficient are horizontal and vertical rupture directions, rupture velocity, and directivity ratio (i.e., percentage of unilateral rupture). These parameters were determined by grid searching to minimize the misfit function (Boatwright, 2007). In this misfit function, Boatwright (2007) originally used the ratio of the peak ground motion (PGA and PGV) for each station to that expected from the ground motion prediction equation, in order to estimate the directivity coefficient. We used the ratio of the fall-off rate for each station to -2 (i.e., the fall-off rate of the ω-2 source spectral model).
For an example of an Mw5.3 event at 17:52 on 19 April 2016 (JST) occurring in southwest of aftershock area, the estimated rupture directivity coefficient showed that the horizontal and vertical rupture directions are N214° E and 135° from vertically down, respectively, with a rupture velocity of 2.4 km/s. The best directivity ratio was 1.0, which means the rupture propagates unilaterally. On the other hand, the strong motion generation area for this event modeled by simulating the observed ground motions using the empirical Green’s function method showed the rupture propagated southwest and shallow direction from the hypocenter (Somei et al., 2020). These results indicate that rupture directivity coefficient estimated from the variation in the fall-off rate was reasonable though the rupture directivity coefficient was upon the assumption of the line source model. Based on the rupture directivity coefficient, we will develop a new spectral inversion method considering with the rupture directivity.
Acknowledgements: The strong motion data were provided from the K-NET, KiK-net, and F-net operated by the National Research Institute for Earth Science and Disaster Resilience (NIED), Japan Meteorological Agency (JMA), Kumamoto Prefectural Government, Kagoshima Prefectural Government, and Nagasaki Prefectural Government. This study was supported by JSPS KAKENHI Grant Number JP20K05044. and by achievements of the collaborative research program (2020H-01) of the Disaster Prevention Research Institute of Kyoto University.