The New Hebrides Trench in the South West Pacific is a nearly north-south trending bathymetric low located north of New Zealand and east of Australia that delineates a region where the Australian Plate is currently subducting beneath the western margin of the North Fiji Basin. Several large islands fall within a few hundred kilometers of the trench axis. These include New Caledonia and the Loyalty Islands to the west on the Australian Plate and the Vanuatu Archipelago, formerly known as New Hebrides, to the east. See figure 1 for a map of this region.
Analysis of data from recent Global Positioning System (GPS) campaigns has confirmed that portions of the South West Pacific exhibit some of the most rapid tectonic plate convergence rates ever observed (Bevis et al., 1995). However, in central Vanuatu, some of the measured rates of motion are not consistent with nearby rates (Bevis et al., 1995; Calmant et al., 1995; Taylor et al., 1995), with regional rates (Larson and Freymueller, 1995), or with predictions of modern global tectonic motion models (Larson and Freymueller, 1995). Some have suggested that the seismic coupling along the interface between the two plates varies along the arc and further point to large bathymetric features on the underthrusting Australian Plate as probable culprits affecting tectonic coupling (Collot et al., 1985; Fisher et al., 1991; Taylor et al., 1995). These features would be the d'Entrecasteaux Ridge, the Sabine Bank and the Bougainville Guyot just west of the islands of Espiritu Santo and Malakula. These features are tens of kilometers wide at the base parallel to the trench strike and
extend several thousand meters above the surrounding ocean floor (Collot et al., 1985; Fisher et al., 1991). The tectonic influence of large impinging features such as these might well explain the aberrant geodetic motion data (Calmant et al., 1995, 1997; Taylor et al., 1995), episodic vertical motions as evidenced in repeated coral emergences (Taylor et al., 1980, 1987, 1990) and in large multiple event earthquakes (Chatelain et al., 1986; Chinn and Isacks, 1983; Ebel, 1980) in this region.
Analysis of earthquake activity can provide some insight into active tectonic processes and can be further used to distinguish between neighboring regions with variations in tectonic processes. Indeed, others have given their interpretations of the active processes in the Vanuatu region by focusing on selected aspects and sources of seismicity information. These have included analyses of: shallow seismicity (Chinn and Isacks, 1983; Pacheco et al., 1993), deep seismicity (Chatelain et al., 1992), fore- and aftershocks (Chatelain et al., 1986), and earthquake sequences (Ebel, 1980; Prevot et al., 1994; Taylor et al., 1980; Vidale and Kanamori, 1983); and have utilized data from several sources: global catalogs such as the International Seismological Centre catalog (Chinn and Isacks, 1983; Prevot et al., 1994), the Harvard Seismology Centroid-Moment Tensor catalog (Liu et al., 1995; Pacheco et al., 1993), and local databases such as the ORSTOM/Cornel network (Chatelain et al., 1986, 1992, 1993; Prevot et al., 1994).
Isacks et al. (1981) postulated on the likelihood of strong interplate coupling along Espiritu Santo and Malakula in central Vanuatu. Lay and Kanamori (1981) in their analysis state the New Hebrides Trench has an uncertain percentage of seismic slip but estimate it to be 50%. Taylor et al. (1990), using selected large earthquakes, estimates the seismic coupling along the central New Hebrides Trench to be less than 10% and suggests evidence that coupling along sections further south is lower, while sections to the north may be slightly higher. In a comparative analysis of 19 subduction zones using selected earthquakes, Pacheco et al. (1993) estimated a similar value of 13% seismic coupling for the New Hebrides Trench as a whole. These estimates present a mixed interpretation of seismic coupling along the New Hebrides Trench. A more detailed analysis of seismic coupling along the plate interface would be useful to understand active tectonic processes there.
To tackle the issue of seismic coupling, magnitude or moment estimates for events available in seismological catalogs can be used to calculate fault model parameters for individual events. That is, estimates for fault rupture areas and average slips for the events can be determined using event energy release estimates and an assumed fault rupture model. Accumulating these model average slips over the fault rupture areas on the subduction interface and then dividing by the time duration covered, can then give an average annual seismic slip rate. This can be compared with far-field convergence rates and indicate the proportion of slip on the slab interface that occurs by sudden motion during an earthquake and, hence, the coupling along a given section.
For the investigation here, solutions contained in the Harvard Seismology Centroid-Moment Tensor Catalog (Harvard Seismology Centroid-Moment Tensor Catalog), and information in Abe's catalog (1981, 1982, 1984) and other selected references are used with a circular fault model to derive a detailed representation of the variation in seismic coupling along the New Hebrides Trench. The results are then compared with local and far-field convergence vectors from several GPS campaigns (Calmant et al., 1995, 1997; Larson and Freymueller, 1995; Taylor et al., 1995) and recent global plate tectonic motion models (Larson and Freymueller, 1995) to better understand the variation in tectonics within the New Hebrides subduction zone.
The organization of this investigation will be as follows. In section 2, the GPS and global plate motion data will be presented and motivations for the investigation discussed. In section 3 the two primary seismicity data sources used here will be introduced and the techniques used to select relevant events presented. This will be followed by a description of the slip modeling and accumulation method in section 4. Section 5 begins with a description of the major geographical regions within the arc and then a more detailed discussion of the relationship of the seismicity and motion data and their relevance to the tectonics of the region.