Magma migrations and caldera formation revealed by tilt observation during the 2000 Miyakejima volcano eruption, Japan: (1) dike intrusions and (2) tilt-steps with 50-s pulse waves

Miyakejima volcano, central Japan, has been experiencing an enormous seismic swarm, crustal deformation and emission of volcanic gases since June 26, 2000. The NIED (National Research Institute for Earth Science and Disaster Prevention) Miyakejima volcano observation network has continuously and successfully monitored the volcanic activities, mainly by the borehole-type seismometers and tiltmeters, and broadband seismometers. The 2000 Miyakejima volcano eruption is characterized by four stages in the first two months: the first stage: the dike intrusion and migration from the southeastern to the western part of the volcano (June 26 - June 27 1200 LT), the second stage: the shrinkage of the volcanic body (June 27 1200 LT - July 7), the third stage: episodic summit eruption and summit crater collapse (July 8 - August 18) and the fourth stage: episodic summit eruption and volcanic-gas emission (August 18 - ). In this presentation, we focus on two significant phenomena among many suggestive volcanic activities, that is (1) dike migration during the first stage and (2) tilt-steps with 50-s pulse wave, associated with caldera formation at the summit area during the third stage. (1) dike intrusions and migrations were revealed by the analyses of tilt-data observed at the five stations in the island. We applied the tensile crack-opening models by Okada [1985] to recognize temporal and spatial tilting changes. The results suggested that the magma began to climb up toward the summit with the dip angle about 60- 70 degrees, reaching almost the ground surface, but the magma did not erupted. After that, the magma intruded vertically toward the western frank of the volcano and furthermore, moved toward western-off the island, causing the huge seismic swarms, those were never recorded. (2) the tilt-steps with 50-s pulse waves were observed during the third stage. This stage began after the first summit eruption at July 8 1841LT, interpreted to be due to the first summit collapse. The tilt-steps with 50-s pulse waves occurred intermittently once or twice per day on average, but none of their events were synchronized with the summit eruption except the first tilt-step on July 8 1841 LT. Analyses of tilt and broadband seismic data show that the tilt-steps with the 50-s pulse waves were caused by the opening of sill-like tensile cracks oriented in a SE-NW direction with a dip-angle of 20 degrees, at a depth of 7-8 km. Each volumetric change ranges 10⁶-10⁷ m³. This source can be interpreted to be thermal two-phase flow instabilities and the caldera formation may be due to the result of this perturbation of stress-field of volcanic body.