Estimates of orientations of stress and strain tensors based on fault-plane solutions in the epicentral area of the great Hawaiian earthquake of 1868

Abstract

The 95 best-quality fault-plane solutions for earthquakes in a 30 by 10 km area near the epicenter of the 1868 Hawaiian earthquake ( M about 7.9) were used for estimating the orientation of the stress and strain tensors. Most of the earthquakes had magnitudes of ML = 3.5 ± 0.6 with nine events in the range 4.2 ≦ ML ≦ 5.5. Although this area, Hilea, is capable of great earthquakes, it seldom produces events in the range 5.5 ≦ ML ≦ 6.5. We found that the directions of the slip and the principal strains are much more uniform in Hilea than in the neighboring regions of Kaoiki and south Kona, not varying more than 15°. The directions of the stress tensor are poorly constrained because of the uniformity of fault-plane solutions. Therefore one cannot be certain that the stress tensor is uniform in this area, but we assume that it is, because the strain and slip directions are. The stress and strain tensor directions differ significantly from those in the Kaoiki and south Kona areas. The slip directions as well as the plane defined by e1 and e3 are oriented approximately perpendicular to the upper part of Mauna Loa's southwest rift, suggesting that magma intrusions there (at a distance of 20 km) probably provide the stress for the seismic activity in the Hilea area. We propose that a tectonic boundary exists near latitude 19° 19′ separating the Kaoiki and Hilea areas. The stress and strain tensor orientations are significantly different north and south of this boundary, and virtually no strike-slip earthquakes occur in the Hilea area, whereas they are the majority of solutions in the Kaoiki area. The fault-plane solutions in Hilea have a near-horizontal and a near-vertical nodal plane. In other areas of Hawaii, the evidence suggests that the near-horizontal plane is a zone of weakness provided by buried oceanic sediment, which allows slip of the upper crust away from the volcanic centers toward the coasts. Here, however, the stress inversion does not lead to a clear choice of preferred fault plane. Thus, it may be that both nodal planes do act as fault planes in this area and that the rupture process in large earthquakes of the 1868 type may be complex involving decollement, steeply dipping normal faulting, and even strike-slip faulting.