Search

Sixty eruptions were recorded from a TV camera on the crater rim, and a 9 station seismic net and 2 infrasonic microphones on the mountain, to test a previous result that eruptions were being triggered by separate earthquakes of depth up to 4km. The recording period was 16 December 1986 to 7 January 1987. The seismic waveforms of similar large explosions were closely identical, and after stacking to improve the signal to noise ratio, plots of seismic arrival time versus distance from the eruption site showed that the seismic intercept time was 1.43 ±0.06s later than the TV explosion time, and the apparent velocity was 4060 ±92m/s. This velocity was much higher than that used for focal determinations (2.1km/s), and it appears that the errors in reading emergent onsets, plus an erroneously low velocity, were responsible for the previously published pipe-like distribution of explosion earthquake hypocenters extending to 4km depth. If so, the visible explosions were the source of the seismic waves. Explosions were occurring from areas of 2 to 10m across in the incandescent or convecting part of the lava lake, and were preceded by updoming for about 1s. All eruptions ejecting bombs caused earthquakes, but ash eruptions from vents outside the lava lake were almost aseismic. Ejection velocities of bombs calculated from flight times ranged from 10 to 76m/s. The fastest bombs followed an incandescent ash front expanding at up to 160m/s. The highest velocity of bombs ejected without ash was 35m/s. All bombs thrown out of the crater were highly vesicular. Relevelling after explosions took 3-8s the few times it was seen. More frequently, there was an upwelling at the site 8.8 ±1.6s later. This indicates a viscosity of ca.10^4 Pas. High enough for the lava foam itself to explode.