Microearthquakes recorded on mid-ocean ridge spreading centers using seafloor ocean-bottom seismometer (OBS) networks provide valuable constraints on the mechanisms of tectonic faulting and magmatic deformation.  From 2003-2006, an eight-station OBS network comprising one broadband and seven short-period seismometers was deployed along a 10-km section of the central portion of the Endeavour Segment of the Juan de Fuca Ridge.  Remotely operated vehicles (ROVs) were used to deploy the sensors below the seafloor to ensure a good coupling to the ground.  The ROVs also enabled us to accurately locate the instruments on the seafloor and measure sensor orientations.  The short-period instruments consisted of three orthogonal Mark Products L-28B geophones that have a flat frequency response from 1 – 90 Hz that were deployed in horizontal coreholes drilled into basalts or concrete caissons that were partially buried in sediments.  The broadband instrument consisted of a three-component Guralp CMG-1T sensor that has a flat frequency response from 2.8 mHz – 50 Hz and was buried with glass beads within a caisson that was excavated into the sediments.  

We have developed a Matlab-based algorithm to automatically pick P- and S-wave arrival times and calculate earthquake locations.  Since our network contains extensive recordings of fin whale vocalizations, the algorithm distinguishes earthquakes from whale calls based on signal frequency.  We present a 38-month catalog of automatically determined hypocentral parameters for over 37,000 earthquakes recorded along the Endeavour Segment between August 2003 and October 2006.  The catalog includes two substantial and complex earthquake swarms recorded in January and February 2005 that appear to mark the termination of a multi-year non-eruptive spreading event that began with another regional swarm in 1999 that was detected using the SOSUS array.  The spatial distribution of epicenters and pattern of moment release observed during the smaller January swarm are consistent with magmatic extension along the southern portion of the West Valley Segment.  In contrast, seismicity associated with the larger February swarm and the pressure response in regional boreholes indicates that magmatic extension was concentrated on the northern portion of the Endeavour Segment.  Vent-field seismicity levels increased within 2 days of the onset of each swarm, but the spatial distribution of seismicity was different for the two swarms.  The time delay is consistent with along-axis diffusion of a hydrologic pressure transient within the upper crustal aquifer.  Following the swarms, daily seismic activity levels decreased ~80% compared to pre-swarm levels, while seismicity ceased almost entirely at the segment ends.  We posit that the post-swarm pattern of seismicity is a result of decrease in crustal stresses.