Relocation of seismicity preceding the 1984 eruption of Mauna Loa Volcano, Hawaii: Delineation of a possible failed rift

Abstract

Waveform cross-correlation based repicking of P arrival times and subsequent relocation of 187 earthquakes that occurred near the summit of Mauna Loa, Hawaii, prior to the March 1984 eruption has illuminated a previously obscured structure beneath the northwestern flank. Simultaneous inversion for hypocenters and velocity model parameters using the refined arrival times resulted in well-constrained relative earthquake locations with very low arrival time misfits (average RMS 0.041 s). Pre-eruption seismicity from this time period occurred in two groups: a shallow group located near Mauna Loa’s summit, at depths of 1–3 km, and a deeper group (5–10 km) located 4–6 km northwest of the summit. After relocation, we found that most of the northwest flank earthquakes occurred along a 3-km-long planar feature striking about 60° West of North in a thin band about 1 km thick. There is a temporal migration of epicenters from the northwestern edge of the zone migrating towards the summit. Focal mechanisms in this zone reveal a change in faulting from strike-slip in the southeast to a mix of strike-slip and normal faulting in the northwest. This feature we interpret to be related to a rift zone that was stunted by the buttressing of the adjacent volcanoes Hualalai and Mauna Kea. Previous gravity and magnetic studies provide supporting evidence for the existence of a rift zone. There was only a modest increase in clustering of the summit events. After relocation, hypocenters that were previously located within a diffuse zone (4.5 km) were reduced to a 2.5-km-wide zone. They extend from a depth of 6 km to about sea level, trending toward the summit. The focal mechanisms from the summit events are mix of faulting types without a consistent pattern.