Event Date
The 1906 San Francisco Earthquake was the last ‘Big One’ in California, producing a 475 km-long surface rupture — one of largest continental strike-slip earthquakes ever observed. Since then, it’s been unusually quiet on California’s active faults. Is a repeat of a great 1906-type event what we should be most worried about in California? Answering this question relies on long-term earthquake earthquake forecasts calibrated from paleoseismic data. In this presentation, I show that long paleoseismic records (n ≥ 10 events) from plate-boundary strike slip faults in California and New Zealand are well fit by a Weibull distribution of inter-event times, defining hazard increasing as a power-law of time (∼t^k ). Two modes of earthquake-cycle behavior emerge from this analysis, which I hypothesize to result from whether faults produce fewer, large, system-spanning ruptures (like 1906) versus more frequent, smaller, partial ruptures. The Alpine fault in New Zealand exhibits a quasi-periodic, system-spanning rupture mode, with strongly convex hazard functions (k≥2). The San Andreas and San Jacinto faults in California exhibit a higher rate of activity and more irregular recurrence behavior that I associate with a partial-rupture mode (k≤1). This suggests 1906 may have been a rare event,and that the next ‘Big One’ will likely be a smaller, partial rupture. Despite claims that a ‘Big One’ on the southern San Andreas is overdue, the timing of past events shows that this is not the case. Rather the likelihood of a large, surface rupturing earthquake increases steadily (get’s more due) with each passing year. From a hazard standpoint, the time-dependent earthquake likelihood estimated from Weibull-distributed paleoseismic return times for the southern San Andreas fault exceeds the level anticipated by formal earthquake forecasts.