Search

[Display omitted] • Employed a Variational Autoencoder (VAE) to detect anomalies in spectra. • Generate artificial spectra and calculate probability it came from the original spectra. • Spectrum, geochemistry, and latent space combined to analyse anomalous data. • Spectra, Geochemistry, with the VAE act as a powerful anomaly detection tool in unsupervised settings. Fourier Transform infrared spectroscopy (FTIR) is an emerging cost effective and rapid mineralogical characterization technique being applied in the geosciences. Detecting anomalous FTIR spectra is especially relevant to the geoscience domain, as it may indicate abrupt changes in geology or mineralogical composition of the rock sample being examined. Given a large volume of data, detecting anomalies that exhibit significant and abrupt spatial and compositional variability is a time-consuming and challenging task. This paper explores the use of an unsupervised variational autoencoder (VAE) for determining anomalies that may exist within a set of FTIR spectra collected from reverse circulation (RC) drill chip samples spanning several iron ore deposits from the Pilbara region in Western Australia. Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) were measured from 1,579 two-metre composite samples. Our results showed that the VAE was effective in separating anomalous spectra from spectra typical of unmineralized banded iron formation by leveraging the probabilistic latent representation of the spectra in as few as two latent dimensions. To validate our results, detected anomalous samples were compared with their respective geochemical assays to analyse their mineralogical differences, which may have led to the anomalous spectra. In the iron ore sample data used in this study, the observed spectral anomalies were shown to have elevated concentrations of Al 2 O 3 and TiO 2 wt.% while being several standard deviations below the mean Fe 2 O 3 wt.% indicating mineralogies rich in shale as opposed to iron oxide rich mineralogies. While the paper demonstrates the efficacy of the VAE in anomaly detection, it can also be effective in assuring the quality of the FTIR data as a pre-processing step, which is critically important for machine learning applications. [ABSTRACT FROM AUTHOR]

The paper presents new geological and geochemical data on the Perseverance komatiite complex (PK) that hosts the Perseverance Ni-sulphide deposit and the volcanic and sedimentary rocks forming the footwall of the PK. Using the 3D reconstruction of the PK and its' footwall volcano-sedimentary succession the heterogeneity of the komatiite channel has been characterised on the basis of spatial distribution of the geochemical petrogenetic indicators. The study shows that the komatiite facies proximal to the mineralised bodies and the main pathway are expressed in the 3D models as the segments with anomalously high MgO concentrations coupled with low contents of Al 2 O 3 , TiO 2 and CaO (Fig - abstract). The geochemical indicators of komatiite fertility, Ni/Cr and Ni/Ti ratios, are commonly in the range of 1 to 3 for Ni/Cr and 2 to 5 for Ni/Ti within the main pathway. The Ni/Cr exceeds 5 and Ni/Ti 20 in the mineralised parts of the channel. In the peripheral zone komatiites these values commonly drop to less than 1, corresponding to barren komatiites. Similarly large variations are exhibited by the ratios of mantle normalized Pt and Pd to Al and Ti over short distances from 1 to 25. Closely spaced high and low values of these ratios hinder their use for discriminating fertile from barren komatiites. However, within the komatiite main lava pathway they can be used as geochemical pathfinders for magmatic sulphides. Reconstruction of the structurally complex komatiite sequences can be facilitated by using lithogeochemical correlation of the host volcano-sedimentary successions and particularly tracing geochemically distinctive marker units. For these purposes a special reference in this study was made to the geochemical systematics of rocks that make up footwall succession of the PK complex. It was shown that amphibolites of the host sequence can be geochemically subdivided into different types, including mafic volcaniclastics (mafic-1) and intrusive gabbro (mafic-2). The mafic volcaniclastics are strongly enriched in LREE, possibly due to presence of sedimentary epiclastic material or reflecting magma contamination by the crustal material during ascent through the crust. Additionally, they have notably high-Th contents, allowing them to be classified as having high-Th geochemical affinities. The second group (mafic-2), interpreted as coherent in origin, are characterised by flat REE patterns and lower overall abundances of the rare earths and Zr corresponding to a tholeiitic composition. Contents of Th and TiO 2 of these rocks are clustered in the two narrow intervals of compositions, corresponding to low-Th and intermediate-Th basalts, which form different amphibolite bodies. Mafic volcanics intercalate with the felsic volcanics, and quartz-feldspar-biotite-amphibole (QFBA) schists. It is shown that QFBA schist composition corresponds to metamorphosed clastic sedimentary and volcaniclastic rocks of mixed composition, which contain both clastic material derived from an Archaean provenance and volcaniclastics material derived from mafic and felsic tuffs. In conclusion, abundances and ratios of the trace elements in composition of the mafic rocks are matching to that of the different parts of the Agnew Wiluna greenstone belt and can be used for overall stratigraphic correlations. [Display omitted] • The paper presents new geological and geochemical data on the Perseverance komatiite complex (PK) that hosts the Perseverance Ni-sulphide deposit and the volcano-sedimentary succession forming the footwall of the PK. • The 3D reconstruction of the PK and its' footwall volcano-sedimentary succession has been undertaken that has allowed to enhance the heterogeneity of the komatiite channel on the basis of spatial distribution of the geochemical petrogenetic indicators. • The study shows that the komatiite facies proximal to the mineralised bodies and the main pathway are expressed in the geochemical 3D models as the segments with anomalously high MgO concentrations coupled with the anomalous values of the geochemical indicators of komatiite fertility, including Ni/Cr, Ni/Ti ratios and the ratios of mantle normalized Pt and Pd to Al and Ti. • Reconstruction of the structurally complex komatiite sequences can be facilitated by using lithogeochemical correlation of the host volcano-sedimentary successions and particularly tracing geochemically distinctive marker units. It was shown that amphibolites of the host sequence can be geochemically subdivided into different types, including mafic volcaniclastics (mafic-1) and intrusive gabbro (mafic-2). • Mafic volcanics intercalate with the felsic volcanics, and quartz-feldspar-biotite-amphibole (QFBA) schists. It is shown that QFBA schist composition corresponds to metamorphosed clastic sedimentary and volcaniclastic rocks of mixed composition, which contain both clastic material derived from an Archaean provenance and volcaniclastics material derived from mafic and felsic tuffs. The paper presents new geological and geochemical data on the Perseverance komatiite complex (PKC) that hosts the Perseverance Ni-sulphide deposit and the volcanic and sedimentary rocks forming the footwall of the PKC. Using the 3D reconstruction of the PKC and its' footwall volcano-sedimentary succession the heterogeneity of the komatiite channel has been characterised on the basis of spatial distribution of the geochemical petrogenetic indicators. The study shows that the komatiite facies proximal to the mineralised bodies and the main pathway are expressed in the 3D models as the segments with anomalously high MgO concentrations coupled with low contents of Al 2 O 3 , TiO 2 and CaO. The geochemical indicators of komatiite fertility, Ni/Cr and Ni/Ti ratios, commonly are in the range of 1 to 3 for Ni/Cr and 2 to 5 for Ni/Ti within the main pathway, The Ni/Cr exceeds 5 and Ni/Ti 20 in the mineralised parts of the channel. In the peripheral zone komatiites these values commonly drop to less than 1, corresponding to barren komatiites. Similarly large variations are exhibited by the ratios of mantle normalized Pt and Pd to Al and Ti over short distances from 1 to 25. Closely spaced high and low values of these ratios hinder their use for discriminating fertile from barren komatiites. However, within the komatiite main lava pathway they can be used as geochemical pathfinders for magmatic sulphides. Reconstruction of the structurally complex komatiite sequences can be facilitated by using lithogeochemical correlation of the host volcano-sedimentary successions and particularly tracing geochemically distinctive marker units. For these purposes a special reference in this study was made to the geochemical systematics of rocks that make up footwall succession of the PKC complex. It was shown that amphibolites of the host sequence can be geochemically subdivided into different types, including mafic volcaniclastic rocks (mafic-1) and intrusive mafic rocks, apparently metagabbro sills (mafic-2). The mafic volcaniclastic rocks are strongly enriched in LREE, possibly due to presence of sedimentary epiclastic material or reflecting magma contamination by the crustal material during ascent through the crust. Additionally, they have notably high-Th contents, allowing them to be classified as having high-Th geochemical affinities. The second group (mafic-2), interpreted as coherent in origin, are characterised by flat REE patterns and lower overall abundances of the rare earths and Zr corresponding to a tholeiitic composition. Contents of Th and TiO 2 of these rocks are clustered in the two narrow intervals of compositions, corresponding to low-Th and intermediate-Th basalts, which form different amphibolite bodies. Mafic volcanics intercalate with the felsic volcanics, and quartz-feldspar-biotite-amphibole (QFBA) schists. It is shown that QFBA schist composition corresponds to metamorphosed clastic sedimentary and volcaniclastic rocks of mixed composition, which contain both clastic materials derived from an Archaean provenance and volcaniclastics material derived from mafic and felsic tuffs. The different geochemical affinities of the mafic rocks distributed in the footwall succession of the PKC apparently suggests different magma sources and, possibly, can also reflect the position of the rocks within the mantle plume structure. In conclusion, abundances, and ratios of the trace elements in composition of the mafic rocks matches that of the different parts of the Agnew Wiluna greenstone belt and can be used for overall stratigraphic correlations. [ABSTRACT FROM AUTHOR]

• Analysis of land cover changes of Darling Range. • Fourteen time series images between 1988 and 2014 were used. • Compares the performances of three machine learning algorithms. • The analysis outputs of random forest algorithm were validated. The Darling Range in Western Australia is a major bauxite producing region. Clearing, excavation and rehabilitation activities related to bauxite mining have influenced land cover within this region since mining commenced in the 1960s. This paper presents a study that used machine learning and time series visualisation to analyse the land cover changes of the Darling Range using time-lapse multispectral images, with the aim of understanding the impact of the mining activities and land rehabilitation patterns of the region. Land cover changes were analysed using 14 Landsat Thematic Mapper (TM) images between 1988 and 2014. The spatial distribution of land cover was classified automatically using machine learning algorithms, and their temporal changes were visualized for analysis. Supervised classification was carried out based on the six spectral features of the images using three machine learning algorithms, namely support vector machine (SVM), random forest (RF) and Naïve Bayes (NB). The results showed that the RF algorithm achieved overall accuracy, (ratio of correctly classified samples divided by the total number of samples), greater than 95% for all years. The temporal changes of land cover distribution over the study period were visualized using a change map. These changes were compared with land clearing and rehabilitation records of a mining company operating in that region. A close correlation was observed between the automated analysis outputs and the company's records. This work demonstrates the potential use of machine analysis in improving the accuracy of spatial data related to land clearing; and in monitoring vegetation recovery of closed and rehabilitated mines. [ABSTRACT FROM AUTHOR]

Graphical abstract Highlights • Geochemical simulations of precipitation of uranyl-vanadate minerals from groundwater. • Hydrogeochemistry of uranium leaching, transportation and precipitation in palaeochannels. • Discussion of critical factors controlling calcrete uranium deposits. • Conceptual models of calcrete-hosted uranium systems. Abstract This three-part paper reports the results of geochemical modeling, prospectivity modeling and quantitative resources assessments of calcrete-hosted surficial uranium deposits in the palaeochannels of geologically, physiographically, and climatologically permissive part of Western Australia. In this Part 1, geochemical dynamics of uranium mobilization and precipitation in near-surface oxidized groundwater systems are reviewed in order to understand the processes responsible for precipitation of uranyl vanadate minerals in valley/lacustrine calcrete within palaeochannels in arid and semi-arid desertic regions. The review indicates that uranium precipitation is essentially a function of concurrent changes in three mutually interdependent parameters of groundwaters, namely, ∑ CO 2 (i.e. activities of all the carbonate bearing species), Eh-pH, and the activities of other ionic species. Geochemical modeling of groundwater data from northern part of the Yilgarn craton in Western Australia indicates that opening of the groundwater aquifer system to the atmosphere and consequent evaporation is likely the key process leading to precipitation of uranyl vanadate minerals in valley calcrete and playa lake sediments. Fluid mixing could also induce the precipitation below the water table, provided that the mixing groundwaters have contrasting geochemistry. These processes may operate in diverse geomorphic traps in palaeochannels. A generalized conceptual model of calcrete-hosted uranium systems is presented and regional-scale targeting criteria and their spatial proxies are identified, which, in turn, are used in Part 2 to develop a targeting model for calcrete-hosted uranium deposits in Western Australia. [ABSTRACT FROM AUTHOR]

Vast parts of the Australian continent are prospective for precious and base metal mineralisation, but exploration is hindered by extensive cover of often deeply reaching regolith. New operational exploration methods are required that can help to characterise the cover and provide information about bedrock signatures. This paper shows how mineral mapping information from a combination of satellite multispectral Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) imagery and drill core hyperspectral profiling data (HyLogging™) can be used to unravel the regolith stratigraphy and to describe regional variations of regolith landforms, delivering important information for mineral exploration. The case study is located in the Neale tenements in the northeastern Albany–Fraser Orogen (Western Australia), which is prospective for Tropicana-style gold mineralisation. By interpretation of indicator minerals from hyperspectral drill hole logging data the regolith stratigraphy atop a metamorphic basement, comprising saprock, ferrugineous saprolite, kaolinitic saprolite, silcrete and transported cover, is recorded in cm-detail. Important mineralogical parameters extracted from the hyperspectral subsurface data and validated by XRD and FTIR, are 1) the abundance and type of iron oxides, 2) the abundance and crystallinity of kaolinite, 3) the abundance and composition of primary minerals, such as white mica, and 4) the abundance of quartz. The HyLogging™ data served as ground control points for mineral mapping information provided by CSIRO's ASTER Geoscience Products, which are a collection of mineral maps that highlight variations in the abundance, type or chemistry of selected mineral groups. Key ASTER Geoscience Products for regolith characterisation were the Ferric Oxide and AlOH abundance and composition images. The comparison of the surface with the subsurface data suggests three major different regolith landforms, including erosional, depositional and relict areas, which were used to generate a map showing transported versus relict and erosional areas. Erosional domains were mapped out in great detail, providing important information for exploration in saprolite dominated areas. Furthermore, source areas of transported material could be identified, which may help to understand the distribution of geochemical signatures collected during, for example, geochemical soil sampling projects. [ABSTRACT FROM AUTHOR]

• The Teutonic Bore volcanics are broadly coeval to the Penzance granite. • The age of the Penzance granite is ca. 2682 Ma. • The Jaguar volcanics and the ore at the Teutonic Bore camp are ≤ca. 2693 Ma. • The Penzance granite possibly supplied heat and metals to the mineralisation. • Exploration in the EGS should focus on fluid pathways of HFSE-enriched granites. The Teutonic Bore Camp, comprised of the Teutonic Bore, Jaguar and Bentley deposits, is one of the most significant volcanic-hosted massive sulphide (VHMS) camps in Western Australia. Despite being extensively studied, only recently there have been advances in the understanding of the mechanism that drove the formation of mineralisation. It has been recognized by recent studies that the volcanic-hosted deposits from the Teutonic Bore Camp represent replacement-type VHMS systems, with significant input of fluids and metals from a magmatic source. This paper tests the existing hypothesis that the nearby Penzance granite acted as the metals source and/or thermal engine driving the development of these ore deposits. New age constraints on the formation of the host volcanic sequence at the Bentley deposit and the crystallization of the Penzance granite allows for the construction of a 4D evolutionary model for the ore system. A new U-Pb SHRIMP monazite age of 2681.9 ± 4.5 Ma indicates that the Penzance granite post-dates the host stratigraphy at Bentley (ca. 2693 Ma) and is probably coeval with mineralisation. All zircons (Penzance, Bentley units I and III) have very similar ƐHf (i) , with most values between −1 and + 6, slightly higher than the ƐHf (i) of zircons from other granites and volcanics within the Kurnalpi Terrain, and indicative of juvenile sources. The mean Th/U ratios are ~0.7 and ~0.6 for the Penzance and Bentley zircons, respectively. All zircons have similar Ce/Nd (CN) ratios. The chemical similarities between the zircons from the granite and the volcanic rocks at Bentley support a shared magmatic source between the Penzance and the Teutonic Bore Camp sequence. The Penzance granite is the likely source of heat, and potentially metals, which drove the VHMS mineralisation at the Teutonic Bore Camp. [ABSTRACT FROM AUTHOR]