Postrifting seismicity and a model for the 1993 diking event on the Coaxial segment, Juan de Fuca Ridge

Abstract

Microearthquake data from two deployments of ocean bottom seismographs made shortly after the June/July 1993 rifting episode on the CoAxial segment of the Juan de Fuca Ridge constrain the propagation path of the magmatic dike and provide insight into the effect of the dike injection on the axial stress field. The first deployment, consisting of five instruments in a small array (aperture of ∼2 km) at the eruption (Flow) site (46°30′N), was made 3 months after the episode. The second deployment, consisting of two larger arrays (apertures of ∼9 km) of five instruments each, one at the eruption site and one at a site of vigorous venting of bacterial flocculent material (Floe) site (46°18′N), was made 11 months after the eruption. Epicenters from nine posteruption earthquakes near the Floe site at the south end of the segment are confined to a narrow zone on the rift axis. Epicenters from 61 earthquakes observed near the Flow site are located throughout the rift valley and include an event cluster just south of the observed lava flows. We combine these posteruption earthquakes with T wave data recorded by SOSUS hydrophones during the rifting episode to develop a model for the diking event. In this model, magma propagated at depth along the axial rift zone until reaching the eruption site near 46°31′N. During the next few days the dike swelled as magma accumulated behind the arrested tip, elevating local stress levels and generating a swarm of seismic activity. About 3 days after the magma reached the north end of the segment, the top of the dike penetrated the seafloor surface, resulting in eruptions along the rift zone at the Flow site.