Interpretation of the Miyakejima 2000 eruption and dike emplacement using time animations of earthquakes

The seismic sequence of over 12,000 earthquakes accompanying the 2000 eruption of Miyakejima volcano has been studied by viewing time animations of the earthquakes beneath transparent topography. Seismic activity began on the evening of June 26 extending southwest from Miyakejima’s summit. A few hours later the seismicity abruptly shifted to the WNW and a submarine eruption occurred off the West Coast of Miyakejima on the morning of June 27. Phreatic eruptions at Miyakejima’s summit in July were accompanied by the formation of a new caldera. Following caldera formation explosive eruptions occurred in August. The eruption ended with minor explosions in September. The seismic activity that began with a low-magnitude swarm beneath Miyakejima grew to a major swarm with hundreds of events of M＞4 extending more than 40km WNW from Miyakejima. Lesser numbers of earthquakes occurred on two N-S trending lines extending south and north of the main seismic trend. The seismicity has been interpreted as evidence for emplacement of a massive dike on the main trend that triggered additional earthquakes on the two cross trends. Our interpretation involves more restricted dike emplacement west of Miyakejima, including the possibility of additional submarine eruption, following cracking of the Philippine Sea plate. The seismic activity associated with explosive eruptions in August helps to define Miyakejima’s magma plumbing. A shallow reservoir beneath the southwest slope is defined by concentrations of earthquakes at 4-6km depth, and a deeper source is suggested by a smaller number of earthquakes extending to 10km vertically beneath the shallow source. Seismic activity preceding and accompanying eruptions at Miyakejima’s summit are defined by seismic swarms extending from 4km depth to the surface along a path connecting the summit with the shallow reservoir. Away from Miyakejima shallow (＜1 km) earthquake swarms at minimum rates of 1 event per hour extending over several hours occur within restricted areas of diameter less than 3km and define possible additional sites of undersea eruption or intrusion. Beneath sites west of Miyakejima the seismicity at depths of less than 4km occurs earlier and toward Miyakejima, consistent with magma transport from Miyakejima’s shallow reservoir. Shallow swarms extending 15km to the WNW strongly suggest that additional intrusion and possibly eruption may have occurred on June 27-28. Between June 27 and July 12 along the main seismic trend, and beneath the shallow sites, progressively deeper earthquake swarms occur at progressively later times, a pattern inconsistent with magma transport and interpreted here as the Philippine Sea plate cracking downward. The initial shallow cracking guided magma to the June 27 undersea eruption site. Subsequent cracking to the west allowed very rapid lateral withdrawal of magma from the Miyakejima reservoir allowing a new caldera to form. The deep cracking of the plate may have triggered additional magma sources, including a deep source suggested by the modeling of regional ground deformation data