HR: 0830h
AN: NG41C-0072 INVITED     [PDF]
TI: On the role of disorder upon the effective dimensionality and dynamic complexity of earthquake faults
AU: * Weatherley, D K
EM: dion@quakes.uq.edu.au
AF: QUAKES, Earth Systems Science Computational Centre, University of Queensland, St. Lucia, Qld 4072 Australia
AU: * Weatherley, D K
EM: dion@quakes.uq.edu.au
AF: Australian Computational Earth Systems Simulator MNRF, University of Queensland, St. Lucia, Qld 4072 Australia
AU: Anghel, M
EM: manghel@lanl.gov
AF: Computer and Computational Sciences Divison, Los Alamos National Laboratory, Los Alamos, NM 87545 United States
AB: We measure the role of quenched disorder (representing failure strength) upon the effective dimensionality of a driven, dissipative system describing the dynamics of an earthquake fault. The system consists of a discrete 2D cellular fault zone embedded within a 3D elastic solid (Y. Ben-Zion, J. Geophys. Res. 101, 5677, 1996) and is defined by a set of parameters that describe the dynamics, rheology, property disorder, and fault geometry. Depending upon the location in the system parameter space, the coarse dynamics of the fault can be confined to an attractor whose dimension is significantly smaller than the space in which the dynamics takes place. The dynamics of the fault system is probed by recording the surface deformations that indirectly reflect the brittle processes of the fault (which are observable by InSAR and GPS techniques). The asymptotic attractors of the system are studied by identifying coherent structures (or dominant modes) present in the surface deformation fields and projecting the system dynamics onto the principal directions defined by these coherent structures. We estimate the effective dimensionality by computing the number of modes needed to explain $95%$ of the statistical variation of the surface deformation fields and by probing the geometry of the attractor using an analysis of its correlation dimension. A sharp transition has been detected in the number of effective degrees of freedom as the dynamic weakening of failure strengths is varied (M. Anghel, Chaos, Solitons, and Fractals 19, 399, 2004). This transition is associated with a separation of time and length scales in the system dynamics. We extend these results by studying the impact of varying the statistical properties of the failure strength distribution upon the effective dimensionality of the fault dynamics. A demonstration of the robustness of the low dimensional coarse dynamics of the system to changes in quenched disorder, implies that simplified fault models may be employed for forecasting the seismicity of active fault zones. Simplified fault models need not capture precisely the detailed structure of the fault zone in order to be useful for development of data driven forecasting models based on statistical learning techniques (M. Anghel, Y. Ben-Zion, and R. Rico-Martinez, PAGEOPH, in press).
DE: 3210 Modeling
DE: 3220 Nonlinear dynamics
DE: 7209 Earthquake dynamics and mechanics
DE: 7223 Seismic hazard assessment and prediction
SC: Nonlinear Geophysics [NG]
MN: 2003 Fall Meeting