Lithospheric structures of the newer volcanics province, Western Victoria, Australia, from a long-period magnetotelluric method

In this thesis the magnetotelluric method is used to image the lithospheric structures of western Victoria and, particularly, to investigate previously identified teleseismic anomalies in the lithosphere and their possible relationship to magma genesis of the Newer Volcanic Province (NVP). An array of long-period MT data, consisting of 40 stations, has been collected in western Victoria over the same 270 km x 300 km grid of the previous teleseismic study. None of the traditional models of rifting, subduction and hot spot can adequately explain the distribution and age of volcanism. To constrain lithospheric temperatures, a 2D steady-state lithospheric thermal conduction model is developed based on the interpretation of a recent two-way-time deep seismic reflection survey conducted across the central and western Victorian Goldfields. This model predicts the temperature beneath the Stawell and Bendigo zones at the Moho (~40 km depth) is ~750°C and it increases to ~1200°C at ~80 km depth. This model shows that thermal refraction may perturb surface heat flow values across the Moyston Fault and the Heathcote Fault Zone. However, the paucity of heat flow data and poor understanding of geology at depth make the thermal conduction models highly sensitive and of limited use for constraining MT models. Forward modelling of MT data support the presence of a lithospheric conductor underneath the Central Highlands and the presence of an electrical conductivity contrast between the Delamerian and the Lachlan Orogens. In addition, forward models show the ocean has a strong effect on MT responses 50 km inland of the Victorian coastline and 130 km inland of the South Australian coastline, while the effect of the Otway basin is only significant within a distance of 30 km of the thickest sediment sequence. Results of 2D inversion modelling illustrate that there is a conductive zone (10-30 Ωm) underneath the Central Highlands at ~40-80 km depth that extends ~40-90 km NW-SE. The anomaly is consistent with the presence of 4-11% partial melting of lithospheric mantle in the presence of H2O (0.05-0.1%), supported by independent geochemistry studies on melting depths and degree of partial melting. Inversion results show the origin of the anomaly is not elevated temperature in a dry mantle due to a mantle plume beneath the NVP. Other possible causes considered include hydrous minerals, sulphides and high-Fe content, however, none produce conductivities of the anomaly on their own. The inversion models show that the melt is located within a “cavity” at the base of a younger thinner lithospheric mantle with oceanic origin neighbouring older thicker continental lithospheric mantle. The preferred model for magma genesis is by decompression melting caused by mantle upwelling of buoyant asthenosphere underneath this cavity. However, some water is required to reduce the mantle solidus.