The M8.1 Intraplate, Strike-Slip Macquarie Earthquake of 23 December 2004

The M8.1, strike-slip earthquake on 23 December 2004 was ~150 km from the Macquarie Ridge plate boundary. This large intraplate earthquake provides an opportunity to investigate both the source processes of oceanic strike-slip earthquakes, which are controversial and poorly known, and to constrain the tectonics of the complex region surrounding the Macquarie Ridge. We model teleseismic P and SH waves from 29 stations using a point source moment tensor inversion. Our preferred model has two subevents, both strike-slip with strike and dip within about 15 degrees of one another; the first is tightly constrained by the first motions. The sources overlap by 17 s and have a combined duration of ~50 s. The centroid depths are 30 -- 35 km. There is no evidence for directivity in the waveforms, so our result is consistent with bilateral rupture along a slightly curved or bent fault. The lack of directivity means that we cannot unambiguously identify the fault plane. We prefer the NNW-SSE plane as it aligns with the aftershocks and the fossil transform faults in the region. We are performing a slip inversion to constrain further the fault plane and slip distribution. The apparent stress of the earthquake is ~5 MPa, higher than continental strike-slip earthquakes, but similar to previous intraplate oceanic earthquakes, e.g. 1998 M8.1 Antarctica. The stress drop (based on the 200 km aftershock extent, and the depth of 35 km) is ~3 MPa, consistent with previous oceanic and also continental strike-slip earthquakes. The tectonics of the region around the Macquarie Ridge are poorly known. Cande and Stock (2004) proposed that the SE part of the Australian plate is actually a separate plate (the Macquarie plate) separated by a diffuse plate boundary. Their model predicts NNE-SSW extension in the region of the 2004 M8.1 earthquake, which is at the northerly extent of a group of previous, smaller, earthquakes. They all have T axis orientations that fit this plate model. The age of the lithosphere at the 2004 earthquake is ~35 Ma (Sutherland, 1995). This would imply that the seismic slip is focused in the deeper part of the seismic zone as noted for oceanic transform earthquakes (Abercrombie and Ekström, 2001).