Why closed seismic cycles matter for time-dependent seismic hazard: Lessons from global paleoearthquake records

The timing and size of the past large earthquakes that ruptured active faults are important to better understand seismic processes and time-dependent seismic hazards. A recent study highlights the rarity of ‘overdue’ earthquakes for New Zealand faults, a finding that directly contrasts observations from California, which indicate an unlikely long period of seismic quiescence. Here, we analyze paleoearthquake and historic records from 210 faults globally, including California, to test the international applicability of the findings for the New Zealand faults against a global active fault dataset. By comparing earthquake-elapsed and mean-recurrence data that derive from end-member fault systems, we explore the factors that control the shape of recurrence-interval distributions on different regions, and assess whether existing paleoearthquake and historical data can be used for estimating time-dependent seismic hazard. Our analysis: 1) demonstrates that the regions examined generally behave similarly for interevent and elapsed times, except for California which forms an outlier. This dissimilarity is important as faults in California have been commonly used to inform earthquake forecast models; 2) supports recurrence-interval distributions that are consistent with positively-skewed renewal models; and 3) proposes an improved approach for defining recurrence-interval distributions that involves the closed elapsed times constrained by historic ruptures and their penultimate events.