Poster #169, Fault and Rupture Mechanics (FARM)

Dynamic Rupture Simulations of Coseismic Interactions on Orthogonal Strike-Slip Faults

Julian C. Lozos
Poster Image: 

Poster Presentation

2021 SCEC Annual Meeting, Poster #169, SCEC Contribution #11594 VIEW PDF
Strike-slip fault systems are typically classified by the strike direction and sense of slip of their primary faults. However, orthogonal faults with the opposite sense of slip can also exist in such systems, particularly in transpressional or transtensional tectonic settings. At a basic level, there are three possible rupture behaviors at an orthogonal fault junction: a rupture begins on one fault and propagates onto the second fault (as in the 2012 M8.6 Off-Sumatra earthquake), a rupture occurs on one fault and then a second rupture occurs later on the other fault (as in the 2019 Ridgecrest, California sequence), or a rupture occurs on one fault and the second remains uninvolved (as in cou...ntless other events). In other words, the orthogonal fault question is an earthquake gate question: what allows rupture to propagate through the junction and through both component faults versus confining rupture to one fault or the other, and is this a persistent behavior or does it change with time? Understanding the physics behind these different behaviors is important for evaluating hazard near orthogonal fault systems.

Here, I use the 3D finite element method to simulate dynamic rupture on simplified orthogonal strike-slip fault systems, to determine which geometrical and stress conditions produce coseismic rupture on both component faults. In models with uniform initial tractions on both faults, which isolate the effects of fault geometry, the position of the nucleation zone relative to the second fault is key: if the initial part of the rupture unclamps/reduces normal stress on the orthogonal fault, it will rupture, and if the initial rupture clamps/increases normal stress on the orthogonal fault, it experiences no more than a small patch of triggered slip at the junction. If the initial rupture nucleates on a fault segment that ends at cross fault, the stopping phase from rupture reaching the end of the fault will cause the cross fault to rupture, regardless of clamping or unclamping. In models where I resolve a uniform regional stress field on the faults, which leads to different initial tractions on each fault, I find that only a narrow range of stress orientations allow multi-fault rupture. Within that range, however, stopping phases and nucleation location control rupture patterns the same way as in the uniform traction models.