Evidence Against the New Madrid Long-Lived Aftershock Hypothesis

Abstract:

It has been suggested that continuing seismicity in the New Madrid, central U.S. region is primarily composed of the continuing long-lived aftershock sequence of the 1811-1812 sequence, and thus cannot be taken as an indication of present-day strain accrual in the region. We examine historical and instrumental seismicity in the New Madrid region to determine if such a model is feasible given 1) the observed protracted nature of past New Madrid sequences, with multiple mainshocks with apparently similar magnitudes; 2) historical rates of M≥6 earthquakes after the initial activity in 1811-1812; and 3) the modern seismicity rate in the region. We use ETAS modeling to search for sub-critical sets of direct Omori parameters that are consistent with all of these datasets, given a realistic consideration of their uncertainties. High aftershock productivity is required both to match the observation of multiple mainshocks and to explain the modern level of activity as aftershocks; synthetic sequences consistent with these observations substantially overpredict the number of events of M≥6 that were observed in the past 200 years. Our results imply that ongoing background seismicity in the New Madrid region is driven by ongoing strain accrual processes and that, despite low deformation rates, seismic activity in the zone is not decaying with time.