Seismic hazard from the Hispaniola subduction zone: Correction to “Historical perspective on seismic hazard to Hispaniola and the northeast Caribbean region”

[1] In the paper “Historical perspective on seismic hazard to Hispaniola and the northeast Caribbean region” by U. S. ten Brink et al. (Journal of Geophysical Research, 116, B12318, doi:10.1029/2011JB008497, 2011), we incorrectly inferred an expected average slip of 17 m on the subduction interface rather than 1.7 m (p. 11 of 15, paragraph 33). We used this incorrect seismic slip estimate and a slip accumulation rate of 14 mm/yr to infer an incorrect recurrence interval of 1200 years for this segment of the subduction zone. Eight M ≥ 7 earthquakes took place along this segment in the twentieth century, with two events dominating the seismic moment, the M7.7–7.9 1943 and the M7.8–8.1 4 August 1946 earthquakes [ten Brink et al., 2011, Table 3]. No other comparable great historical earthquakes were felt, even though the historical record of subduction interface earthquakes in Hispaniola is likely complete for the past 500 years [ten Brink et al., 2011]. Ten Brink et al. [2011] suggested that a 1200 year recurrence interval was consistent with these observations, but a 120 year recurrence interval is considerably shorter than the historical record. We hereafter consider the questions: Why have there been no other great earthquakes, like those in 1943 and 1946, on the subduction interface north of Hispaniola since Europeans settled in Hispaniola in 1494? What are the implications to seismic hazard from this segment of the subduction zone? [2] Reanalyzing the 1943–1953 earthquake sequence in the region, Dolan and Wald [1998] noted the progression of seismic activity from east to west throughout this period. They determined the rupture area of the 1943 and 1946 earthquakes on the basis of aftershocks as 100 × 65 km and 195 × 95 km, respectively. It is thus possible that these two events, with additional smaller events during the twentieth century, have ruptured the entire 415 km long segment, defined by these earthquakes (Figure 1). P wave tomographic images show that this segment is the only segment with a continuous slab extending from the surface to a depth of 500 km in the northern Caribbean [ten Brink and Villasenor, 2012, Figures 1c–1e]. These tomographic images were derived using the P wave global tomographic technique of Bijwaard et al. [1998], improved with additional arrival times from well-located earthquakes at both teleseismic and regional distances [Villasenor et al., 2003]. A horizontal high-velocity layer at depths <150 km underlies Haiti (Figure 2b), whereas north of Puerto Rico, the slab “is missing” at depths of 50–150 km (Figure 2f). Therefore, this 415 km long region likely defines a subduction segment. [3] Why have there been no other great earthquakes like the 1946 earthquake in the preceding historically documented 450 years? The expected average slip during an earthquake, d, is given by