DETERMINING HYPOCENTER POSITIONS OF SEISMIC SWARM EVENTS ALONG THE LUCKY STRIKE SEGMENT OF THE MID-ATLANTIC RIDGE, 2001 MARCH 16 AND 17

In February 1999, a group of U.S. investigators deployed a 6-element hydrophone array for long term monitoring of the seismicity along the Mid-Atlantic Ridge between 15„a and 35„a N. During the 16th and 17th of March 2001, a large seismic swarm was detected by this array on the Lucky Strike segment of the Mid-Atlantic Ridge south of the Azores hotspot (~37„a N) [Dziak 2003]. The highest frequency of events occurred at the beginning of the swarm, with activity becoming less frequent 1.5 hours afterwards [Dziak 2003]. This style of seismicity seems to be indicative of a diking event in the segment. The first determinations of the earthquake positions using T-wave records had a large degree of error and thus it is difficult to correlate the events spatially to determine the mechanism of the swarm. Here we report the results of a study to better define these locations using P-waves recorded by seismic stations in the Azores islands in addition to the T-waves. We acquired earthquake hydrophone logs indicating T-wave arrival times and P-wave arrival times from land-based seismogram logs. Each log was comprised of a list of earthquakes that occurred during the time frame mentioned. Events were then chosen from all listed events of the swarm according to the following criteria: number of stations that received the event, the magnitude of the event, and relative spacing between preceding and subsequent events. We then filtered out unnecessary frequencies and computed estimated arrival times as a guide to manually pick the T and P-wave arrival times. These arrival times were then analyzed using an iterative grid-search program that defined smaller and smaller grids to yield precise positions of the events selected for evaluation. These positions were compared to the positions determined previously. Preliminary results show that the addition of P waves has not significantly reduced the uncertainty of earthquake locations due to a poor signal to noise ratio and relatively large errors in P wave arrival times. It is thought that using the surface wave instead of the P-wave may improve the results.