Sudholz, Zachary
Description
The integration of petro-chemical datasets for mantle rocks with magnetotellurics and seismic tomography datasets is crucial in advancing our understanding the architecture, lithology and chemical composition of the sub-continental lithospheric mantle (SCLM) underlying the Australian continent. Progress towards establishing a set of uniform models that integrate both up-to-date geophysical and geochemical datasets has been hindered by (1) the absence of suitable geothermobarometers that are...[Show more] required to determine the equilibration pressure (P) and temperature (T) of mantle xenoliths and single-grain xenocrysts, and (2) a lack of data for mantle-derived samples from across the Australian continent, particularly from cratonic regions (i.e., Pilbara Craton, Yilgarn Craton, Gawler Craton, Kimberley Craton). We address the first of these issues by revising current geothermobarometry methods for mantle rocks. This is achieved through the use of a large experimental dataset comprised of new and published experiments completed between 10 to 70 kbar and 850 to 1820 C in a variety of natural and synthetic, fertile and refractory experimental starting compositions. Our experimental dataset is used to recalibrate the (a) chromium (Cr) in clinopyroxene single-grain geobarometer, (b) nickel (Ni) in garnet single-grain geothermometer, (c) garnet-clinopyroxene Fe-Mg exchange geothermometer, and (d) garnet-orthopyroxene Fe-Mg exchange geothermometer. For each of our updated calibrations we demonstrate substantial improvements in precision and accuracy through application to independent experimental datasets and well-equilibrated kimberlite and lamproite-hosted xenolith suites.
To build on recent geophysical observation we apply our updated single-grain geothermobarometers to a large suite of heavy mineral concentrates and mantle xenoliths from kimberlites and basalts emplaced across the South Australian Craton and Phanerozoic Tasmanides of eastern Australia. Pressure and temperature estimates on Cr diopside are used to construct FITPLOT paleogeothermal gradients which are used to constrain lithospheric thicknesses and mantle heat flow for over 25 intrusions. Single-grain PT estimates on Cr diopside and pyrope garnet are coupled with major, minor and trace element compositions to vertically map the lithology and chemistry of the SCLM beneath each pipe. For the South Australian Craton, our results define two litho-chemical domains within the shallow (ca. 60-130 km) and deep (>160 km) SCLM that are separated by a geochemically and geophysically distinct mid-lithosphere discontinuity (ca. 130-160 km). The shallow SCLM is comprised of re-fertilised- previously refractory low- to moderate Cr2O3 lherzolite and high-CaO wehrlite. The mid-lithosphere is made up of extremely fertile, low modal pyrope garnet and low modal Cr diopside litho-types, such as orthopyroxenite or dunite. The deep SCLM is made up of less-depleted silicate melt metasomatized- high to moderate-Cr2O3 lherzolite that contain pyrope garnet with elevated concentrations of TiO2 and heavy- REE enriched types. The lithology and chemical composition of the SCLM reflects a distinct layered-cake stratigraphy that suggests top-down growth through plume impingement. For the Phanerozoic Tasmanides of eastern Australia, our results indicate that the mantle is anomalously thin and warm with respect to the neighbouring South Australian Craton. Similarities' in heat flow and lithospheric thickness between several closely spaced Mesozoic and Cenozoic aged pipes suggest the architecture of eastern Australia has not changed markedly over the past 150 Myr. The mantle beneath the Tasmanides is dominated by fertile spinel and garnet-spinel lherzolite, with occasional pockets of garnet and garnet-spinel pyroxenite. The abundance of low-olivine websterite within xenoliths suites likely reflect remnants of earlier melting episodes that have been collected during recent volcanic eruptions.
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.