HR: 10:50h
AN: S21G-03    [PDF]
TI: Evidence for Self-Similar, Triangular Slip Distributions on Earthquake Faults, and Implications on Fault and Earthquake Mechanics
AU: * Manighetti, I
EM: manig@usc.edu
AF: University of Southern California, Department of Earth Sciences, Los Angeles, CA 90089 United States
AU: * Manighetti, I
EM: manig@usc.edu
AF: Institut de Physique du Globe de Paris, 4 Place Jussieu, Paris, 75005 France
AU: Sammis, C
EM: sammis@usc.edu
AF: University of Southern California, Department of Earth Sciences, Los Angeles, CA 90089 United States
AU: King, G
EM: king@ipgp.jussieu.fr
AF: Institut de Physique du Globe de Paris, 4 Place Jussieu, Paris, 75005 France
AU: Campillo, M
EM: michel.campillo@obs.ujf-grenoble.fr
AF: Laboratoire de Geophysique Interne et Tectonophysique, Universite Joseph Fourier, Grenoble, 38041 France
AB: The objective of our work is to characterize slip distributions for earthquake faults. We have compiled three types of data available in the literature: 1) Surface measurements of slip performed by geological mapping after earthquakes; 2) "total" earthquake slip distributions deduced from published finite-source rupture models (Mai, 2002); 3) Source time functions of large earthquakes known to have propagated unilaterally. Overall, the data sets include about 100 earthquakes of various magnitudes (5.5-8.1) and kinematics. Both surface and total slip profiles are broadly similar in overall shape (along both strike and dip), being roughly triangular and asymmetric regardless of magnitude and kinematics. This suggests that distribution of slip on an earthquake fault is not random, and that seismic slip distributions are self-similar overall. The observation of triangular, self-similar seismic slip distributions together with roughly constant stress drops during earthquakes, cannot be accounted for by the available mechanical models. Earthquakes with such an asymmetric triangular slip profile initiate close to one fault tip and mainly propagate unilaterally along-strike toward the other end of the fault. Most hypocenters locate at or close to where total slip is higher along-strike, but some initiate at or close to where total slip is minimum (fault taper). However, no earthquake is observed to initiate where total slip is minimum both along-strike and -dip. This demonstrates that earthquakes cannot initiate without any "stress concentrator" or barrier capable of sustaining high stresses and strain. As a matter of fact, earthquakes seem to accumulate less total slip (per length or width) along either strike or dip when their hypocenter is close to the fault taper along strike or dip, respectively. The triangular, self-similar seismic slip distributions resemble those recently found on long-term cumulative normal faults (0.1-60 km-long; 0.01-1 Ma-old) and rift systems (10-1000 km-long; 0.1-10 Ma-old). This further suggests that features created in tens of seconds and in tens of thousands years or millions years may share some common mechanical explanations. We suggest that the observation of triangular, self-similar slip profiles can be mechanically understood if some form of off-fault inelastic deformation is included. We use an elastic (static) modeling to show that large triangular zones of off-fault damage can explain the observed triangular, self-similar slip profiles provided damage is anisotropic and in the form of cracks and/or secondary faults sub-parallel or slightly oblique to the main fault. This suggests that on-fault processes may be insufficient to explain the dynamics of earthquake ruptures, with wide ranging implications for understanding earthquake processes.
DE: 7209 Earthquake dynamics and mechanics
DE: 7230 Seismicity and seismotectonics
DE: 8010 Fractures and faults
DE: 8020 Mechanics
SC: Seismology [S]
MN: 2003 Fall Meeting