Inferring Interseismic Coupling Along The Lesser Antilles Arc: A Bayesian Approach

The Lesser Antilles subduction zone is a challenging region when it comes to unraveling its seismogenic behavior. Over the last century, the subduction megathrust has been seismically quiet, with no large thrust event recorded, which raises the question whether this subduction zone is able to produce large interplate earthquakes or not. However, two historical earthquakes in the 19th century, a M 7-8 in 1839 and M 7.5-8.5 in 1843, are proposed to have occurred along the subduction megathrust, although no direct evidence exists. Here we provide a new assessment of interseismic coupling for the Lesser Antilles subduction zone, based on updated Global Positioning System (GPS) velocities and the latest models of the slab geometry and elastic crustal structure. We use a Bayesian approach, allowing us to explore the entire range of plausible models and to provide realistic estimates of interseismic coupling and associated uncertainties. We find low to very low coupling along the entire plate interface, including in the proposed rupture areas of the 1839 and 1843 events, where the sensitivity of our model is high. While a further understanding of temporal variations in interseismic coupling needs to be addressed in future studies, our results indicate that the Lesser Antilles subduction zone is uncoupled, which challenges the idea that the 1839 and 1843 earthquakes were thrust events. The updated GPS velocities of this work now also reveal a small, but detectable amount of along-arc extension, consistent with geological observations of active normal faulting within the arc.Plain Language Summary The Lesser Antilles subduction zone forms the plate boundary where the North and South American plates sink underneath the overlying Caribbean plate. Such downgoing movement typically results in the buildup of stress along the frictional interface between the plates. When these stresses overcome the strength of the plate interface, they can be released through potentially devastating earthquakes. Using measurements from Global Navigation Satellite System stations on the islands of the Lesser Antilles, we model strain to determine how coupled the interface is. A high degree of coupling means that large "megathrust" earthquakes are likely, while low coupling means that big events are less likely and/or very rare. Two large earthquakes struck the Lesser Antilles in the 19th century, which have been interpreted to have occurred on the subduction interface. In this work we find a low to very low coupling along the plate interface, which implies that (1) these historical earthquakes are unlikely to have occurred on the subduction interface but rather deeper within the downgoing plate and (2) large, Tohoku-like, megathrust earthquakes are unlikely and/or must be very rare in the Lesser Antilles.