Illuminating the upper mantle beneath the Newer Volcanics province, southeast Australia, using seismic body wave tomography

The Newer Volcanics province (NVP), located in the state of Victoria, southeast Australia, represents the youngest evidence of basaltic intraplate volcanism in the Australian continent, with the most recent eruptions dated less than 5ka. Although one of many Cenozoic eruption centers that populate the eastern edge of the Australian mainland, the NVP is unique in that it is not obviously part of a hot-spot chain. For example, the distribution of NVP eruption centres is elongated in the east-west direction, perpendicular to plate motion. Moreover, it appears that the NVP is the latest phase of an eruption cycle that has operated intermittently since the early Eocene when fast northern motion of the Australian continent commenced. Coupled with modest surface topographic response (~100 m) and a relatively low eruption volume (~20,000 km³) researchers have begun to suspect that the source of the NVP does not fit the mold of a traditional mantle plume model, but instead may be a phenomenon localized to the upper mantle. One possibility is that strong undulations in lithospheric thickness, as observed in surface and body wave tomography, together with a northward movement of the Australian plate relative to the subjacent mantle, may result in edge-driven convection, in which cells of warm, hydrous mantle periodically advect upwards and release melt, which then migrates up to the surface. In this study, we use teleseismic P-wave data recorded by the WOMBAT transportable array project in eastern Australia - the largest experiment of its type in the southern hemisphere - to image 3-D velocity perturbations beneath the NVP. Relative arrival times of global P-phases are inverted to constrain P-wavespeed anomalies in the uppermost mantle. Constraints from regional surface wave tomography are also incorporated into the results in order to account for the long-wavelength structures that are filtered out by the use of relative arrival time residuals. The final P-wave velocity model shows a clear zone of low velocity underlying the NVP (maximum perturbation of -4% relative to AK135). It clearly extends to a depth of just over 200 km, before terminating, with no evidence of reduced velocities down to approximately 300 km, the maximum depth resolution of the seismic data. Furthermore, nearer the surface (~100km depth), there appears to be three distinct regions of low velocity that are distributed E-W between central Victoria and Mt. Gambier near the South Australian border. The lack of evidence for a deep seated anomaly is consistent with the hypothesis that the source of the NVP is confined to the upper mantle, although an important caveat is that plumes are expected to be narrow as they rise through the mantle before broadening out as they encounter the base of the lithosphere; as such, it s possible that the limited spatial resolution of the data (approximately 50 km) is unable to detect narrow vertical structures at depth. However, combined with the observations discussed earlier, our results strengthen the argument for a localized upper mantle anomaly.