The Lattice Solid Model: Towards a realistic simulation model for earthquake micro-physics and the development of a virtual laboratory for the earthquake cycle

1999 
A realistic numerical simulation model for all physical processes underlying the earthquake phenomenon on HPC's would provide a powerful tool to study fault behavior and earthquake nucleation. The microphysical particle-based Lattice Solid Model (LSM) currently being developed at QUAKES provides a basis on which to construct such a model. Presently, the model simulates stress transfer, seismic waves, fracture, friction, heat and gouge dynamics. Simulations show numerous features compatible with laboratory and field studies including shear localization, low-strength faults compatible with the Heat-Flow Constraint, slip pulses on faults, Gutenburg-Richter power law statistics and cycles of seismic activity exhibiting accelerating energy release prior to large events. Ultimately, when fully developed, it is envisaged that the LSM will be capable of simulating all physical processes underlying earthquakes including lubrication and dynamics of fluids, phase transformations, and chemical effects as well as all observable signals including strain, seismic, electric and magnetic. Increased computational capacity, a model refinement process involving feedback with laboratory and field observations, and integration with macroscopic simulation models would provide the means to study the earthquake cycle, and hence, to develop earthquake hazard quantification and forecasting methodology that best uses the incomplete recorded and incoming data. Recent LSM simulation results of patterns of accelerating energy release prior to large events suggest that earthquake statistics can evolve in a predictable way. These results demonstrate the potential utility of realistic numerical simulation models as a means to probe the earthquake cycle, and provide encouragement that earthquake forecasting is feasible, at least under certain conditions.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    10
    References
    1
    Citations
    NaN
    KQI
    []