JELoWeb

Home Page

MinPins

Misc Photos

Jeep Events

2002 Jeep Wrangler

House Photos

Garage Addition

Masters Thesis

SouthWest Pacific Photos

Antarctic Trip

Analysis of geodetic measurements collected in the late summers of 1990 and 1992 in the South West Pacific have determined short term horizontal motion vectors for numerous islands there (Bevis et al., 1995; Calmant et al., 1995; Taylor et al., 1995). These include several islands in the Vanuatu Archipelago, New Caledonia, and the Loyalty Islands such that localized convergence rates across the New Hebrides Trench may be determined. Others have also calculated more generalized plate convergence rates for the region using geodetic data and modern global plate motion models (Larson and Freymueller, 1995).

Analysis of data collected on recent regional GPS campaigns has shown that, in the South West Pacific, plates are converging at some of the most rapid rates ever observed (Bevis et al., 1995). Over the years, several different estimates of the regional plate convergence between the Australian and Pacific plates have been used in investigations (10 cm/year Taylor et al., 1987; 10 cm/year Taylor et al., 1990; 8.5 cm/year Pacheco et al., 1993; 8.8 cm/year Taylor et al., 1994; 9 cm/year Liu et al., 1995; 7.9-8.6 cm/year Taylor et al., 1995). General estimates for plate convergence far from the trench based on the latest plate motion model, NNR-A, and recently collected geodetic data can be calculated from the work of Larson and Freymueller (1995). NNR-A is the No-Net-Rotation-NUVEL1A model - a modification by Argus and Gordon (1991) of DeMets' (1990) original NUVEL global plate motion model.

Using two sites occurring at nearly the same latitude but on opposite sides of the New Hebrides Trench and some distance away, an estimate for the far-field, or regional, convergence rate of the Australian and Pacific plates can be calculated. Sites TOWN, Townsville, Australia, which sits atop the Australian plate to the west, and PAMA, Pamatai, Tahiti, which sits atop the Pacific plate to the east, both occur near -20° latitude and are several hundred kilometers from the trench axis. Assuming a fixed Australian plate reference frame and performing vector arithmetic gives the equivalent convergence at the trench axis. This is shown in table 1 below. A corresponding estimate based on NNR-A global plate motion rates for these sites has also been calculated and is given in the table. Azimuth values represent angular directions east from north for a fixed Australian plate reference frame, i. e. the Australian plate is assumed to not move and other points are moving with respect to it.

Table 1. Regional plate convergence vectors.

The average direction and azimuth based on these values are: 10 cm/year of convergence at an azimuth of 257° . The above azimuths show good agreement with the average slip direction determined previously by Isacks et al. (1981) of 256° using selected thrust type focal mechanisms.

Analysis of geodetic measurements in the South West Pacific have determined short term horizontal motion vectors for numerous islands. The most detailed near trench rates of motion come from Taylor et al., (1995). Here, GPS data was processed to determine motion rates for individual island sites near the New Hebrides Trench. Publications by others (Calmant et al., 1995, 1997) also provide motion rates for some of these sites. Table 2 below summarizes the GPS determined motion rates for sites in Vanuatu. These are: SNTO, or Espiritu Santo, MAEW, or Maewo, PENT, or Pentecost, MLKL, or Malakula, EFAT, or Efate, and TANN, or Tanna. As before, azimuth values represent angular directions east from north for a fixed Australian plate reference frame.

Table 2. Local plate convergence vectors.

Figure 2 graphically depicts the magnitude and direction of each of the GPS determined measurements and model estimates at each of the sites in Vanuatu.

Data from the three sources show some variation but relatively good agreement in direction despite being derived using data collected with a selection of different GPS receivers and antennas, and processed under several different software packages. Local motion vectors for sites within the Vanuatu Archipelago differ slightly in direction from regional motion vectors for the Australian-Pacific plate convergence, tending to trend in a more southerly direction by a few degrees to as many as fifteen to twenty degrees. Measurements for the site on the island of Maewo, MAEW, further east from the trench axis, has the most anomalous direction reading with respect to the other local values. Magnitudes for the motion vectors for sites within different sections of the arc can show large variations from the regional plate convergence rate ranging from a portion of the regional plate convergence rate to slightly greater than the regional rate.

The GPS motion vectors for islands in central Vanuatu (Espiritu Santo, Malakula, Maewo, and Pentecost) are a portion of the regional convergence rate and show the greatest range of variation in direction of any of the measurements while being relatively closely spaced. When compared to the regional convergence vectors, magnitudes are less than half, ranging from 3.6 to 4.7 cm/year, and directions range from nearly coincident to 19.0° to 21.6° more southerly. Such complex motions in such a confined region likely indicate some unusual circumstances in this region. This could be due to distributed convergence or to time dependence of the convergence as might be found with a locked fault segment.

The GPS motion vectors for the southern islands show more reasonable agreement with regional convergence vectors. The Efate site, EFAT, has the closest match to the regional plate convergence vector. Very nearly the same in magnitude and only slightly more south in direction; by 8.0° to 16.6° . Tanna, TANN, further south, has a slightly more rapid convergence - by approximately 1.5 cm/year, but about the same difference in direction, 9.0° to 14.6° more southerly.

If we account for the effects of the suggested opening of the North Fiji Basin to the east (Louat and Pelletier, 1989), the true convergence along the New Hebrides Trench should be the vector sum of the Australian-Pacific plate convergence plus the spreading rate of the North Fiji Basin. Thus, if we attribute the difference, approximately 1.5 cm/year of additional convergence and -10° to -15° of direction, it would appear that only the site furthest south for which we have data, Tanna, could be moving at the appropriate rate while all others reflect lower motion rates to varying degrees.

Lastly, analyses of geodetic data as presented in Calmant et al. (1995) found, within errors, no evidence for convergence or divergence between sites on the Australian plate (sites in New Caledonia, the Loyalty Islands, and others). This implies that there is no deformation or strain accumulation within the subducting plate to the west of the trench axis for at least the duration between campaigns. Thus, variations in convergence between sites in central Vanuatu are likely to be taken up as deformation of some sort in the overriding plate east of the trench axis. However, this collection of some consistent and some contrasting local versus far-field convergence vectors does little to reveal the likely location or details of the deformation at the subduction interface. Analysis of seismic events can reveal information about source processes and their locations. Earthquake source processes within a given region when taken collectively are a manifestation of tectonic processes and plate motions can provide further insight. With the development of, and continued contributions to modern seismological catalogs, substantial information on earthquakes is now readily available. This, then, provides the foundation for further investigation.

© jlundy@wyoming.com 20 June 2003