Looking for supersonic rupture velocities.

Fabrice Cotton
Institut de Protection et de Sûreté Nucléaire, Fontenay-aux-Roses, France

Bruno Hernandez
Institut de Protection et de Sûreté Nucléaire, Fontenay-aux-Roses, France
Laboratoire de Géophysique Interne et Tectonophysique, Université Joseph Fourier, Grenoble, France

Michel Campillo
Laboratoire de Géophysique Interne et Tectonophysique, Université Joseph Fourier, Grenoble, France

Abstract. The recent dynamic simulation of the 1992 Landers earthquake performed by Olsen et al. (Science, december 1997) has shown a str ong variation of the rupture front velocity from subsonic to supersonic values. Supersonic velocities were not expected from kinematic inversions of recent well studied earthquakes. Two reasons could explain why the evidence of such supersonic values are spar ce in past inversions. First there is a trade off between slip amplitude and rupture time that affects solutions based on seismic data alone. Therefor e it is necessary to constrain the slip amplitude distribution by independant data in order to recover the temporal details of rupture propagation. Second, a priori constraints introduced by the parameterisation used in the inversions usually constrain the maximum value of the rupture front velocity. To solve these problems, SAR interferometric data are used to obtain high res olution slip amplitude models. Then strong motion data are inverted through an inversion where the slip distribution and his error are constrained by the results of the geodetic inversion and the rupture front velocity varies freely without constrains to fit the seismic data.