Recurrence Implications of California Paleo-event Hiatus

Abstract:

Estimated rates and recurrence properties of pre-historic earthquakes depend strongly on recorded dates of paleo-events. California paleo-event data, published in the UCERF-3 reports in 2014 and 2015, reveal a curious property: The paleo-event rate for the whole collection of 32 sites exceeds about 4 per century, yet no events at any of the sites has been recorded in about 100 years. Possible explanation include extreme chance (probability about 1% or less) , large-scale clustering with ensemble cv greater than one, or overestimation of paleo-event rates in the pre-instrumental period. Any of these explanations has implications for recurrence. If the hiatus results from extreme luck, any interpretation of the data must be subject to chance occurrence as well. If clustering, single-site recurrence misses the most interesting features of earthquake behavior. So far, no physics-based model of California faulting explains the hiatus. If over-estimation, then some reported events are not earthquakes, and “recurrence” has dubious meaning.