HR: 1340h
AN: S43A-1062    [Abstracts]
TI: Kinematic And Dynamic Simulation Of The 2000 Tottori Japan Earthquake
AU: * Francois-Holden, C
EM: holden@geologie.ens.fr
AF: Laboratoire de Geologie Ecole Normale Superieure, 24 rue Lhomond, Paris, 75005 France
AU: Di Carli, S
EM: sdicarli@geologie.ens.fr
AF: Laboratoire de Geologie Ecole Normale Superieure, 24 rue Lhomond, Paris, 75005 France
AU: Hubans, F
EM: hubans@geologie.ens.fr
AF: Ecole Normale Superieure, 46 all‚e d'Italie, Lyon, 69364 France
AU: Madariaga, R
EM: madariag@geologie.ens.fr
AF: Laboratoire de Geologie Ecole Normale Superieure, 24 rue Lhomond, Paris, 75005 France
AB: We study the kinematic and dynamic rupture propagation of the 2000 Tottori (Japan) earthquake. The earthquake region was well instrumented and provided a good strong motion data set that has been intensively studied. We develop a nonlinear kinematic inversion method based on the neighborhood algorithm which we applied to a set of 30 strong motion recordings located within 40 km of the epicenter. The main purpose of this first step of this study is to compare our results with those from linearised kinematic inversions by Yagi (2001), Sekiguchi (2002) and Semanne et al. (2005). We find several differences that are mainly related to the choice of hypocenter and rupture speed. In a second step, we use the non-linear kinematic source parameters to start a non-linear dynamic inversion. In order to satisfy the dynamic aspects of the rupture we define a rupture scenario where rupture propagation is controlled by the properties of the friction law on the fault. In this study a simple uniform slip-weakening friction law is used everywhere on the fault. From the kinematic slip distribution, we compute the initial stress field on the fault. We then force the beginning of the rupture into an asperity located around the hypocentre of Tottori. Then we determine the initial stress field that leads to a propagation of the rupture that is similar to the kinematics. Several aspects of our kinematic models need to be modified in order to satisfy the generalized Griffith criterion for rupture (from Madariaga and Olsen). The final step is to make a full dynamic inversion using a reduced number of dynamic parameters. For that purpose we project the initial stress field into a limited number of ellipsoidal patches that we parameterize with 5-6 parameters. Using a limited number of such patches we expect to simplify the non-linear inversion method proposed by Peyrat and Olsen.
DE: 7209 Earthquake dynamics (1242)
DE: 7215 Earthquake source observations (1240)
DE: 7290 Computational seismology
SC: Seismology [S]
MN: Fall Meeting 2005