San Andreas Rupture Gates?

Date: 4/24/2019

Time: 04:30 PM

Room: Pike

Many earthquake hazard models employ fault-based source models, often stated in the form of stochastic event sets. Testing source models against actual earthquakes requires “association”, whereby each finite rupture can be matched with a hypothetical one with a preassigned rate or probability. In the old days, big earthquakes were assumed by some to occur in ruptures of prescribed fault segments or combinations thereof, with probabilities assigned for each combination. Segment boundaries played an important role in stopping rupture with a reasonably high probability. Now we recognize the need to assign uncertainties to segment geometry, or even abandon segmentation entirely. An alternative for linear strike slip faults is to identify polygonal sections with finite width and length, assigning “conditional stopping probabilities” (CSPs) describing the probability that rupture would end within a polygon if it gets in. Those with relatively high stopping probabilities are then referred to as “gates” or “sticky wickets.” This approach can be used for arbitrarily many gates: just a few, as in polygons about the segments in the 1988, 1990, 1995 etc. reports of the Working Group on California Earthquake Probabilities (WGCEP), or thousands, as in the Uniform California Earthquake Rupture Forecasts (UCERF) of 2013 and later. A rupture can then be described by two sections, one at each of its ends, and a likelihood can be calculated by which sections stop or don’t stop the rupture. An advantage of the approach is that it focuses on the resistance of the fault elements rather than the epicenters and size distribution of the earthquakes. Smaller events can help test the forecast model because each rupture has two end points. Another advantage is that simulated ruptures in a stochastic event set can be used to estimate the stopping probabilities and participation rates of the polygonal sections. In this presentation I’ll show how the method applies in the WGCEP and UCERF forecasts mentioned above.

Presenting Author: David D. Jackson

Authors