Aftershock sequences in the mid-ocean ridge environment: an analysis using hydroacoustic data

Hydroacoustic data from autonomous arrays and the U.S. Navy's Sound Surveillance System (SOSUS) provide an opportunity to examine the temporal and spatial properties of seismicity along portions of the slow-spreading Mid-Atlantic Ridge (MAR), intermediate-spreading Juan de Fuca Ridge (JdFR) and fast-spreading East Pacific Rise (EPR). Aftershock and foreshock events are selected from the hydroacoustic earthquake catalog using single-link cluster (SLC) analysis, with a combined space-time metric. In the regions examined, hydroacoustic data improve the completeness level of the earthquake catalog by similar to1.5-2.0 orders of magnitude, allowing the decay constant,p, of the modified Omori law (MOL) to be determined for individual sequences. A non-parametric goodness-of-fit test indicates six of the seven sequences examined are described well by a MOL model. The p-values obtained for individual ridge and transform sequences using hydroacoustic data are larger than that previously estimated from the analysis of a stacked sequence generated from teleseismic data. For three sequences along the Siqueiros, Discovery and western Blanco Transforms, p-values are estimated to be similar to0.94-1.29. The spatial distribution of aftershocks suggests that the mainshock rupture is constrained by intra-transform spreading centers at these locations. An aftershock sequence following a 7.1M(s) thrust event near the northern edge of the Easter Microplate exhibits p=1.02 +/- 0.11. Within the sequence, aftershocks are located to the north of a large topographic ridge, which may represent the surface expression of the shallow-dipping fault that ruptured during the mainshock. Two aftershock sequences near 24degrees25'N and 16degrees35'N on the MAR exhibit higher p-values, 1.74 +/- 0.23 and 2.37 +/- 1.65, although the latter estimate is not well constrained because of the small number of aftershocks. Larger p-values along the ridge crest might reflect a hotter thermal regime in this setting. Additional monitoring, however, will be needed to determine if p-value differences between the ridge and transform sequences are robust. A 1999 sequence on the Endeavour segment of the JdFR, which has been correlated with changes in the hydrothermal system, is described poorly by the MOL model. The failure of the MOL model, the anomalously large number of earthquakes within the sequence and absence of a clearly dominant mainshock are inconsistent with aftershock activity and the simple tectonic origin that has been proposed previously for this sequence. (C) 2002 Elsevier Science B.V. All rights reserved.