Resolving simulated sequences of earthquakes and fault interactions

Valere R. Lambert, & Nadia Lapusta

Submitted August 12, 2021, SCEC Contribution #11312, 2021 SCEC Annual Meeting Poster #140

There is growing interest in using numerical earthquake models to directly determine quantities of interest for seismic hazard, such as the probability of an earthquake rupture jumping from one fault segment to another. Here, we investigate the sensitivity of numerical simulations of long-term sequences of earthquakes and aseismic slip (SEAS), including the interaction of fault segments, to choices in numerical discretization and treatment of inertial, wave-mediated effects. We study SEAS in a simplified 2-D model of a crustal fault with two co-planar segments separated by a creeping barrier.

We find that the resolution that achieves numerical convergence of individual earthquake ruptures and short earthquake sequences does not guarantee longer-term convergence of SEAS in cases when the fault is long compared to the earthquake nucleation size: even when earthquake sequences are virtually indistinguishable for an initial simulation period, the sequences of slip events eventually begin to diverge due to the subtle accumulation of numerical differences interacting with the highly nonlinear nature of the problem. Some simulated properties, such as average static stress drop and the history of average shear stress on the fault, are similar among adequately resolved simulations even if there is no long-term convergence. Other properties, such as the frequency of two-segment ruptures, are highly sensitive to numerical resolution. In general, simulations with different numerical discretization (including insufficient discretization), different treatments of inertial effects, and minor changes in the fault properties can produce very different probabilities of multi-segment ruptures. The significant sensitivity of rupture jump probabilities to even relatively minor changes in numerical models implies that this hazard parameter may also be sensitive to physical perturbations on natural faults, suggesting that it may be impractical to estimate in a reliable manner. This motivates further exploration of meaningful metrics for constraining long-term fault behavior and assessing seismic hazard.

Citation
Lambert, V. R., & Lapusta, N. (2021, 08). Resolving simulated sequences of earthquakes and fault interactions . Poster Presentation at 2021 SCEC Annual Meeting.


Related Projects & Working Groups
Fault and Rupture Mechanics (FARM)