Proterozoic crust-mantle evolution in Central Australia : geochemical and isotopic constraints
This thesis is divided into three major integrated parts centring around the theme of Proterozoic crust-mantle evolution in central Australia. They are (1) geochemical, geochronological and isotopic studies of Proterozoic granitoids and mafic amphiblites in the eastern Arunta Inlier, central Australia; (2) geochemical and isotopic investigations of late Proterozoic mafic dyke swarms in central-southern Australia and their implications for crust-mantle evolution and origin of the intracratonic...[Show more] sedimentary basins in the region; and (3) Sm-Nd, U-Pb zircon and REE investigations of sediments from the Amadeus Basin, central Australia, with emphasis on the origin and provenance of the sediments. Part I: SHRIMP U-Pb zircon ages determined for ten major granite suites in the eastern Arunta Inlier, combined with those previously reported, allow a systematic geochronological framework to be established in the region. This study suggests (1)major granitic activities in the Arunta Inlier can be grouped into at least six episodes: 1850-1880 Ma, 1820 Ma, 1750-1780 Ma, 1710-1730 Ma, 1650-1660 Ma and 1600-1615 Ma; (2) the dominant orogenic event occurred during 1750-1780 Ma, when the majority of syntectonic granitoids were intruded, in contrast with the significantly older 1850-1880 Ma Barramundi Orogeny considered to be predominant further north in central Australia. Geochemical and Sm-Nd isotopic studies of major granite suites in central Australia allow three major geochemical groups to be identified, which are a Calcalkaline-trondhjemitic Group (CAT), a Normal Group and a High-heat-production Group (HHP). They can be further subdivided into different age and geochemical subgroups, respectively. The CAT Group, which occurs only in the southern margin of the Arunta Inlier, is characterised by highest Na20, Na/K, Sr, K/Rb and Sr/Y, and lowest K20, Rb, Rb/Sr, Th, U, REE, Nb and Y if compared with other groups. It can be subdivided into a gabbro-diorite-tonalite-trondhjemite (GDTT) subgroup and a tonalite-trondhjemite-granodiorite (TTG) subgroup, the former analogous to modem calc alkaline suites occurring in convergent plate margins, and the latter more like the ubiquitous Archean TTG suites. The Normal Group, the dominant group occurring throughout the Arunta Inlier, is mainly K-rich, and geochemically intermediate between the CAT and HHP Groups. It can be subdivided into four age subgroups, 1820 Ma, 1750-1780 Ma, 1650 Ma and 1600-1615 Ma, respectively. By comparison with the Palaeozoic I-type granites, the Normal Group is overly enriched in K20, Rb, Th, U, Zr, Y and LREE and depleted in MgO, CaO, Sr, Ni and Cr. The HHP Group, which occurs mainly in the inland area of the Arunta Inlier and is spatially associated with preexisting suites of the Normal Group, is characterised by highest K, Rb, Th, U, Rb/Sr and Rb/Zr, and lowest Sr, Ba, Na/K, K/Rb, Ba/Rb, MgO, Cr and Ni among the three. However, on the spiderdiagrams, all the three groups show similar negative Nb anomalies. Despite of the large geochemical diversity in the Arunta granites, no correlations between the Nd isotopic and geochemical signatures are observed. The ranges of initial ENd values and Nd depleted mantle model ages T Nd, DM for the three groups of T Nd DM granites considerably overlap each other. Overall, there are two groups of t ages with the dominant one ranging from 1.96 to 2.33 Ga, and the other from 1.72 to 1.83 Ga. At a given crystallization age, initial ENd values show large variations, whilst the initial ENd values generally increase with decreasing crystallization ages. In conjunction with studies of the Arunta granites, geochemical and Sm-Nd isotopic investigations were also undertaken on mafic amphibolites in the Alice Springs area, southern Arunta Inlier, which led to the recognition of two groups of amphibolites, one characterised by flat to LREE-depleted patterns and positive initial ENd values (+4.2 to+5.1), and one typical ofLREE-enriched patterns and negative ENd values (-1.0 to -2.8). Both groups show geochemical affinities to island arc tholeiites. All the above igneous rocks occurring in the Arunta Inlier can be interpreted in a plate-tectonic scenario. An integrated igneous petrogenetic model for the origin of the igneous rocks has been formulated, in which the igneous precursors of the arc-type amphibolites were derived from slab-component metasomatised mantle wedge, the CAT Group by fractionation and/or partial melting of the arc-type underplates or intrusions, the Normal Group by partial melting of fractionated arc-type intrusions or underplate and the HHP Group by remelting of preexisting granitic sources such as the Normal Group granites. Mixing model, which involves a newly derived mantle component and an older crustal component probably in the form of subducted sediments or preexisting lower crust, is preferred for the interpretation of Nd isotopic compositions in granitoids and mafic rocks of the Arunta Inlier. The older crustal component may have become progressively younger with time as a result of increasing proportions of newly accreted island-arc materials incorporated in the subducted sediments. A derivation of the Nd isotopic signatures from an isotopically uniform 2.1-2.3 Ga mafic underplate is rejected. However, it is considered that the formation ofNd isotopic signatures in these rocks must have involved complex processes and neither simple mixing nor simple two-stage protolith model can satisfactorily explain the observations. In addition, considering the large uncertainties involved in constraining the different components and magma generation processes, caution must be taken in using Nd isotopic data for modelling the growth rates of the continental crust. Tectonically, the Arunta Inlier was located on the southern margin of the Northern Australian Orogenic Province, where subduction-related plate tectonics may have been in operation during the period of 1.9 to 1.7 Ga. The Arunta Inlier itself may represent amalgamation of a series of island-arc, back-arc basin accretionary complexes developed along the margin of the 1.88 Ga continental crust further north. Evidence which supports the plate tectonic concept includes: (1) zircon U-Pb and Nd model age constraints suggesting the Arunta Inlier represents a crustal terrain boundary bordered to the south by the significantly younger Musgrave Inlier; (2) the presence of arc- and MORB-type metavolcanics and subduction-related calcalkaline suites and exclusive absence of mafic volcanics with within-plate signatures;(3) general occurrence of negative Nb anomalies in all igneous rocks suggesting they or their sources or the sources of their sources were subduction-related; (4) geochemical polarity across the Arunta Inlier; (5) Nd isotopic constraints suggesting mixing between a mantle-derived component and an older continental component through subduction of sediments; and (6) the unique geochemical and Nd isotopic constraints from post-tectonic Stuart Dyke Swarm (see Part II). The alternate ensialic rifting tectonics of Etheridge et al. (1987) fails to explain many observations in the Arunta Inlier although it may be applicable for the tectonic and crustal evolution further north in central Australia. A number of problems with the ensialic model have been outlined in the thesis. Part II: Part II of this thesis is aimed at geochemical and Sm-Nd isotopic investigations of Late Proterozoic mafic dyke swarms in central-southern Australia. Through Sm-Nd mineral isochron dating of mafic dykes, two episodes of mafic magmatism in central southern Australia (1075-1090 Ma, 790-870 Ma) have been delineated. Such a study also provides an improved method for relatively reliable and precise dating of mafic igneous rocks. The study of the 1080 Ma mafic dyke swarms in central Australia demonstrates: (1) the continental lithospheric mantle (CLM) can be formed as a result of continental crust formation at convergent plate margins through oceanic crust subduction; and (2) partial melting of subduction-modified CLM during post-tectonic events is possible and some post-tectonic dyke swarms may have resulted from this process. The study of the -800 Ma mafic dyke swarms and flood basalts in central southern Australia suggests they were derived by decompression melting of a large-scale mantle plume impinging upon the base of the continental lithosphere in the region. Large scale crustal extension followed by thermal subsidence as a result of the plume activity may have been responsible for the formation of the broad sedimentary basins in central southern Australia. Part III: In Part III of this thesis, combined Nd and detrital zircon U-Pb constraints have been obtained which suggest that the sediments of the Amadeus Basin have an at least two-component-mixing provenance. They were probably derived by mixing of materials derived from the Arunta and Musgrave Inliers or their equivalents with those from the Musgrave Inlier increasing in stratigraphically higher sequences. The study of the sediment provenance reveals a rare case of REE fractionation during sedimentation as a result of sorting, syn-deposition chemical reaction or post depositional diagenesis. A theoretical model for REE fractionation and a refined way for constraining the sediment provenance ages have been developed.
|Collections||Open Access Theses|
|Zhao J Thesis 1992.pdf||15.18 MB||Adobe PDF|
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.