The AEM method measures regolith and rocks' bulk subsurface electrical conductivity, typically to a depth of several hundred meters. AEM survey data is widely used in Australia for mineral exploration (i.e. mapping undercover and detection of mineralisation), groundwater assessment (i.e. hydro-stratigraphy and water quality) and natural resource management (i.e. salinity assessment). Geoscience Australia (GA) has flown Large regional AEM surveys over Northern Australia, including Queensland, Northern Territory and Western Australia. The surveys were flown nominally at 20-kilometre line spacing, using the airborne electromagnetic systems that have signed technical deeds of staging with GA to ensure they can be modelled quantitatively. Geoscience Australia commissioned the survey as part of the Exploring for the Future (EFTF) program. The EFTF program is led by Geoscience Australia (GA), in collaboration with the Geological Surveys of the Northern Territory, Queensland, South Australia and Western Australia, and is investigating the potential mineral, energy and groundwater resources in northern Australia and South Australia. We have used a machine learning modelling approach that establishes predictive relationships between the inverted flight-line modelled conductivity with a suite of national environmental and geological covariates. These covariates include terrain derivatives, gamma-ray radiometric, geological maps, climate derived surfaces and satellite imagery. Conductivity-depth values were derived from a single model using GA's deterministic 1D smooth-30-layer layered-earth-inversion algorithm. (Brodie and Richardson 2015). Three conductivity depth interval predictions are generated to interpolate the actual modelled conductivity data, which is 20km apart. These depth slices include a 0-50cm, 9-11m and 22-27m depth prediction. Each depth interval was modelled and individually optimised using the gradient boosted tree algorithm. The training cross-validation step used label clusters or groups to minimise over-fitting. Many hundreds of conductivity models are generated (i.e. ensemble modelling). Here we use the median of the models as the conductivity prediction and the upper and lower percentiles (95th and 5th) to measure model uncertainty. Grids show conductivity (S/m) in log 10 units. Reported out-of-sample r-squares for each interval in order of increasing depth are 0.74, 0.64, and 0.67. A decline in model performance with increasing depth was expected due to the decrease in suitable covariates at greater depths. Modelled conductivities seem to be consistent with the geological, regolith, geomorphological, and climate processes in the study area. The conductivity grids are at the resolution of the covariates, which have a nominal pixel size of 85 meters. Datasets in this data package include; 1. 0-50cm depth interval 0_50cm_median.tif; 0_50_upper.tif; 0_50_lower.tif 2. 9-11m depth interval 9_11m_median.tif; 9_11m_upper.tif; 9_11m_lower.tif 3. 22-27m depth interval 22_27_median.tif; 22_27_upper.tif; 22_27_lower.tif 4. Covariate shift; Cov_shift.tif (higher values = great shift in covariates) Reference: Ross C Brodie & Murray Richardson (2015) Open Source Software for 1D Airborne Electromagnetic Inversion, ASEG Extended Abstracts, 2015:1, 1-3, DOI: 10.1071/ ASEG2015ab197

The regolith landform maps are drawn at various scales and illustrate the distribution of regolith materials and the landforms on which they occur. Regolith landforms are described using the regolith terrain mapping (RTMAP) scheme developed at Geoscience Australia or the Residual-Erosional-Depositional (RED) mapping scheme developed by the CSIRO Division of Exploration and Mining.

Weathering is an important process of the Earth’s surface that has a major influence on the chemical and physical properties of rock and soil. The intensity of this process largely controls the degree to which primary minerals are altered to secondary components, including clay and oxide minerals. The degree of surface weathering is particularly important in Australia, where variations in weathering intensity correspond to differences in the nature and distribution of regolith (weathered bedrock and sediments), which mantles approximately 80% of the Australian continent. Here, I use a random forest decision tree machine learning algorithm to first establish a relationship between field estimates of the degree of weathering and a comprehensive suite of covariate or predictive datasets. I then apply this relationship to generate an improved national model of surface to near-surface weathering intensity. Covariates include satellite imagery, terrain attributes, airborne radiometric imagery and mapped geology. The model performs very well, with an r-squared correlation of 0.85 based on 5 K-fold cross-validation on the mean and standard deviation of 300 random forest models. This new weathering intensity model has broad utility for mineral exploration in variably weathered landscapes, agricultural mapping of chemical and physical soil attributes, ecology, and advancing the understanding of weathering processes within the upper regolith. <b>Citation:</b> Wilford, J., 2020. Revised weathering intensity model of Australia. In: Czarnota, K., Roach, I., Abbott, S., Haynes, M., Kositcin, N., Ray, A. and Slatter, E. (eds.) Exploring for the Future: Extended Abstracts, Geoscience Australia, Canberra, 1–4.

Weathering intensity or the degree of weathering is an important characteristic of the earth’s surface that has a significant influence on the chemical and physical properties of surface materials. Weathering intensity largely controls the degree to which primary minerals are altered to secondary components including clay minerals and oxides. The degree of surface weathering is particularly important in Australia where variations in weathering intensity correspond to the nature and distribution of regolith (weathered bedrock and sediments) which mantles approximately 90% of the Australian continent. The weathering intensity prediction has been generated using the Random Forest decision tree machine learning algorithm. The algorithm is used to establish predictive relationships between field estimates of the degree of weathering and a comprehensive suite of covariate or predictive datasets. The covariates used to generate the model include satellite imagery, terrain attributes, airborne radiometric imagery and mapped geology. Correlations between the training dataset and the covariates were explored through the generation of 300 random tree models. An r-squared correlation of 0.85 is reported using 5 K-fold cross-validation. The mean of the 300 models is used for predicting the weathering intensity and the uncertainty in the weathering intensity is estimated at each location via the standard deviation in the 300 model values. The predictive weathering intensity model is an estimate of the degree of surface weathering only. The interpretation of the weathering intensity is different for in-situ or residual landscapes compared with transported materials within depositional landscapes. In residual landscapes, weathering process are operating locally whereas in depositional landscapes the model is reflecting the degree of weathering either prior to erosion and subsequent deposition, or weathering of sediments after being deposited. The weathering intensity model has broad utility in assisting mineral exploration in variably weathered geochemical landscapes across the Australian continent, mapping chemical and physical attributes of soils in agricultural landscapes and in understanding the nature and distribution of weathering processes occurring within the upper regolith.

Australia has a significant number of surface sediment geochemical surveys that have been undertaken by industry and government over the past 50 years. These surveys represent a vast investment and have up to now only been able to be used in isolation, independently from one another. The key to maximising the full potential of these data and the information they provide for mineral exploration, environmental management and agricultural purposes is using all the surveys together, seamlessly. These disparate geochemical surveys not only sampled various landscape elements and analysed a range of size fractions, but also used multiple analytical techniques, instrument types and laboratories. The geochemical data from these surveys require levelling to eliminate, as much as possible, non-geological variation. Using a variety of methodologies, including reanalysis of both international standards and small subsets of samples from previous surveys, we have created a seamless surface geochemical map for northern Australia, from nine surveys with 15,605 samples. We tested our approach using two surveys from the southern Thomson Orogen, which demonstrated the successful removal of inter-laboratory and other analytical variation. Creation of the new combined and levelled northern Australian dataset paves the way for the application of statistical and data analytics techniques, such as principal component analysis and machine learning, thereby maximising the value of these legacy data holdings. The methodology documented here can be applied to additional geochemical datasets as they become available.

Estimating the relative contributions of bedrock geology, mineralisation and anthropogenic contamination to the chemistry of samples collected at the Earth’s surface is critical in research and application fields as diverse as environmental impact studies and regional mineral exploration programs. The element lead (Pb) is a particularly useful tracer in this context, representing a toxin of environmental concern and associated with many other anthropogenic contaminants (e.g. mine wastes, waters, paints, aerosols), as well as with mineralisation. Although Pb concentration data are frequently collected in geochemical studies, isotopic analysis offers an important advantage, allowing discrimination between different sources of Pb. The Pb isotopic composition of regolith is likely to reflect contributions from underlying rock (including Pb-rich mineralisation), wind-blown dust and possibly anthropogenic sources (industry, transport, agriculture, residential, waste handling). Regolith samples collected at different depths may show distinct compositions; bedrock isotopic signatures are expected to dominate in deeper soils, whilst airborne dust and anthropogenic signatures are more important at the surface. Pb isotope ratios in the continental crust show large variations, which will be transferred to the regolith, providing a potentially unique bedrock signal that is easily measured. This research program examines if soil Pb isotope mapping can identify the underlying geology and metallogenic provinces, if different sampling and analytical approaches produce very different results, and how anthropogenic signals vary across the continent. Here, we present our results for the Northern Territory, where single regolith samples from many (not all) catchments define apparently consistent isotopic domains that can be interpreted in relation to the underlying geology (crystalline basement, basins) and mineral deposits. <b>Citation:</b> Desem, C.U., Maas, R., Woodhead, J., Carr, G. and de Caritat P., 2020. Towards a Pb isotope regolith map of the Australian continent: a Northern Territory perspective. In: Czarnota, K., Roach, I., Abbott, S., Haynes, M., Kositcin, N., Ray, A. and Slatter, E. (eds.) Exploring for the Future: Extended Abstracts, Geoscience Australia, Canberra, 1–4.

Preamble: The 'National Geochemical Survey of Australia: The Geochemical Atlas of Australia' was published in July 2011 along with a digital copy of the NGSA geochemical dataset (http://dx.doi.org/10.11636/Record.2011.020). The NGSA project is described here: www.ga.gov.au/ngsa. The present dataset contains additional geochemical data obtained on NGSA samples: the Lead Isotopes Dataset. Abstract: Over 1200 new lead (Pb) isotope analyses were obtained on catchment outlet sediment samples from the NGSA regolith archive to document the range and patterns of Pb isotope ratios in the surface regolith and their relationships to geology and anthropogenic activity. The selected samples included 1204 NGSA Top Outlet Sediment (TOS) samples taken from 0 to 10 cm depth and sieved to <2 mm (or ‘coarse’ fraction); three of these were analysed in duplicate for a total of 1207 Pb isotope analyses. Further, 12 Northern Australia Geochemical Survey (NAGS; http://dx.doi.org/10.11636/Record.2019.002) TOS samples from within a single NGSA catchment, also sieved to <2 mm, were analysed to provide an indication of smaller scale variability. Combined, we thus present 1219 new TOS coarse, internally comparable data points, which underpin new national regolith Pb isoscapes. Additionally, 16 NGSA Bottom Outlet Sediment (BOS; ~60 to 80 cm depth) samples, also sieved to <2 mm, and 16 NGSA TOS samples sieved to a finer grainsize (<75 um, or ‘fine’) fraction from selected NGSA catchments were also included to inform on Pb mobility and residence. Lead isotope analyses were performed by Candan Desem as part of her PhD research at the School of Geography, Earth and Atmospheric Sciences, University of Melbourne. After an initial ammonium acetate (AmAc) leach, the samples were digested in aqua regia (AR). Although a small number of samples were analysed after the AmAc leach, all samples were analysed after the second, AR digestion, preparation step. The analyses were performed without prior matrix removal using a Nu Instruments Attom single collector Sector Field-Inductively Coupled Plasma-Mass Spectrometer (SF-ICP-MS). The dried soil digests were redissolved in 2% HNO3 run solutions containing high-purity thallium (1 ppb Tl) and diluted to provide ~1 ppb Pb in solution. Admixture of natural, Pb-free Tl (with a nominal 205Tl/203Tl of 2.3871) allowed for correction of instrumental mass bias effects. Concentrations of matrix elements in the diluted AR digests are estimated to be in the range of 1–2 ppm. The SF-ICP-MS was operated in wet plasma mode using a Glass Expansion cyclonic spray chamber and glass nebuliser with an uptake rate of 0.33 mL/min. The instrument was tuned for maximum sensitivity and provided ~1 million counts per second/ppb Pb while maintaining flat-topped peaks. Each analysis, performed in the Attom’s ‘deflector peak jump’ mode, consists of 30 sets of 2000 sweeps of masses 202Hg, 203Tl, 204Pb, 205Tl, 206Pb, 207Pb and 208Pb, with dwell times of 500 μs and a total analysis time of 4.5 min. Each sample acquisition was preceded by a blank determination. All corrections for baseline, sample Hg interference (202Hg/204Pb ratios were always <0.043) and mass bias were performed online, producing typical in-run precisions (2 standard errors) of ±0.047 for 206Pb/204Pb, ±0.038 for 207Pb/204Pb, ±0.095 for 208Pb/204Pb, ±0.0012 for 207Pb/206Pb and ±0.0026 for 208Pb/206Pb. A small number of samples with low Pb concentrations exhibited very low signal sizes during analysis resulting in correspondingly high analytical uncertainties. Samples producing within-run uncertainties of >1% relative (measured on the 207Pb/204Pb ratio) were discarded as being insufficiently precise to contribute meaningfully to the dataset. Data quality was monitored using interspersed analyses of Tl-doped ~1 ppb solutions of the National Institute of Standards and Technology (NIST) SRM981 Pb standard, and several silicate reference materials: United States Geological Survey ‘BCR-2’ and ‘AGV-2’, Centre de Recherches Pétrographiques et Géochimiques ‘BR’ and Japan Geological Survey ‘JB-2’. In a typical session, up to 50 unknowns plus 15 standards were analysed using an ESI SC-2 DX autosampler. Although previous studies using the Attom SF-ICP-MS used sample-standard-bracketing techniques to correct for instrumental Pb mass bias, Tl doping was found to produce precise, accurate and reproducible results. Based upon the data for the BCR-2 and AGV-2 secondary reference materials, for which we have the most analyses, deviations from accepted values (accuracy) were typically <0.17%. Data for the remaining silicate standards appear slightly less accurate but these results may, to some extent, reflect uncertainty in the assigned literature values for these materials. Replicate runs of selected AR digests yielded similar reproducibility estimates. The results show a wide range of Pb isotope ratios in the NGSA (and NAGS) TOS <2 mm fraction samples across the continent (N = 1219): 206Pb/204Pb: Min = 15.558; Med ± Robust SD = 18.844 ± 0.454; Mean ± SD = 19.047 ± 1.073; Max = 30.635 207Pb/204Pb; Min = 14.358; Med ± Robust SD = 15.687 ± 0.091; Mean ± SD = 15.720 ± 0.221; Max = 18.012 208Pb/204Pb; Min = 33.558; Med ± Robust SD = 38.989 ± 0.586; Mean ± SD = 39.116 ± 1.094; Max = 48.873 207Pb/206Pb; Min = 0.5880; Med ± Robust SD = 0.8318 ± 0.0155; Mean ± SD = 0.8270 ± 0.0314; Max = 0.9847 208Pb/206Pb; Min = 1.4149; Med ± Robust SD = 2.0665 ± 0.0263; Mean ± SD = 2.0568 ± 0.0675; Max = 2.3002 These data allow the construction of the first continental-scale regolith Pb isotope maps (206Pb/204Pb, 207Pb/204Pb, 208Pb/204Pb, 207Pb/206Pb, and 208Pb/206Pb isoscapes) of Australia and can be used to understand contributions of Pb from underlying bedrock (including Pb-rich mineralisation), wind-blown dust and possibly from anthropogenic sources (industrial, transport, agriculture, residential, waste handling). The complete dataset is available to download as a comma separated values (CSV) file from Geoscience Australia's website (http://dx.doi.org/10.26186/5ea8f6fd3de64). Isoscape grids (inverse distance weighting interpolated grids with a power coefficient of 2 prepared in QGis using GDAL gridding tool based on nearest neighbours) are also provided for the five Pb isotope ratios (IDW2NN.TIF files in zipped folder). Alternatively, the new Pb isotope data can be downloaded from and viewed on the GA Portal (https://portal.ga.gov.au/).

The results of a pilot study into the application of an unsupervised clustering approach to the analysis of catchment-based National Geochemical Survey of Australia (NGSA) geochemical data combined with geophysical and geological data across northern Australia are documented. NGSA Mobile Metal Ion® (MMI) element concentrations and first and second order statistical summaries across catchments of geophysical data and geological data are integrated and analysed using Self-Organising Maps (SOM). Input features that contribute significantly to the separation of catchment clusters are objectively identified and assessed. A case study of the application of SOM for assessing the spatial relationships between Au mines and mineral occurrences in catchment clusters is presented. Catchments with high mean Au code-vector concentrations are found downstream of areas known to host Au mineralisation. This knowledge is used to identify upstream catchments exhibiting geophysical and geological features that indicate likely Au mineralisation. The approach documented here suggests that catchment-based geochemical data and summaries of geophysical and geological data can be combined to highlight areas that potentially host previously unrecognised Au mineralisation. <b>Citation:</b> M. J. Cracknell, P. de Caritat; Catchment-based gold prospectivity analysis combining geochemical, geophysical and geological data across northern Australia. <i>Geochemistry: Exploration, Environment</i>, Analysis 2017; 17 (3): 204–216. doi: https://doi.org/10.1144/geochem2016-012 This article appears in multiple journals (Lyell Collection & GeoScienceWorld)

<div>Bulk quantitative mineralogy of regolith is a useful indicator of lithological precursor (protolith), degree of weathering, and soil properties affecting various potential landuse decisions. To date, no empirical national-scale maps of regolith mineralogy are available in Australia. Satellite-derived mineralogical proxy products exist, however, they require on-the-ground validation. Catchment outlet sediments collected over 80% of the continent as part of the National Geochemical Survey of Australia (NGSA) afford a unique opportunity to rapidly and cost-effectively determine regolith mineralogy using the archived sample material. This report releases mineralogical data and metadata obtained as part of a study extending a previous pilot project for such a national regolith mineralogy database and atlas.</div><div>The area chosen for this study includes the part of South Australia not inside the pilot project, which focussed on the 2020-2024 Exploring for the Future (EFTF) Darling-Curnamona-Delamerian (DCD) region of southeastern Australia, as well as the EFTF Barkly-Isa-Georgetown (BIG) region of northern Australia. The South Australian part of the study was selected because the Geological Survey of South Australia indicated interest in expanding the pilot (DCD) project to the rest of the State. The BIG region was selected because it is a ‘deep-dive’ data acquisition and analysis area within the EFTF Australian Government initiative managed at Geoscience Australia. The whole study area essentially describes a continuous north-south trans-continental transect spanning South Australia (SA), Queensland (Qld) and the Northern Territory (NT), and is herein abbreviated as SA-Qld-NT.</div><div>Two hundred and sixty four NGSA sites from the SA-Qld-NT region were prepared for X-Ray Diffraction (XRD) analysis, which consisted of qualitative mineral identification of the bulk samples (i.e., ‘major’ minerals), qualitative clay mineral identification of the <2 µm grain-size fraction, and quantitative analysis of both major and clay minerals of the bulk sample. The identified mineral phases were quartz, kaolinite, plagioclase, K-feldspar, nosean (a sulfate bearing feldspathoid), calcite, dolomite, aragonite, ankerite, hornblende, gypsum, bassanite (a partially hydrated calcium sulfate), halite, hematite, goethite, magnetite, rutile, anatase, pyrite, interstratified or mixed-layer illite-smectite, smectite, muscovite, chlorite (group), talc, palygorskite, jarosite, alunite, and zeolite (group). Poorly diffracting material was also quantified and reported as ‘amorphous.’ Mineral identification relied on the EVA® software, whilst quantification was performed using Siroquant®. Resulting mineral abundances are reported with a Chi-squared goodness-of-fit between the actual diffractogram and a modelled diffractogram for each sample, as well as an estimated standard error (esd) measurement of uncertainty for each mineral phase quantified. Sensitivity down to 0.1 weight percent (wt%) was achieved, with any mineral detection below that threshold reported as ‘trace.’ </div><div>Although detailed interpretation of the mineralogical data is outside the remit of the present data release, preliminary observations of mineral abundance patterns suggest a strong link to geology, including proximity to fresh bedrock, weathering during sediment transport, and robust relationships between mineralogy and geochemistry. The mineralogical data generated by this study are downloadable as a .csv file from the Geoscience Australia website (https://dx.doi.org/10.26186/147990).&nbsp;</div>

The National Geochemical Survey of Australia (<a href="http://www.ga.gov.au/ngsa" title="NGSA website" target="_blank">NGSA</a>) is Australia’s only internally consistent, continental-scale <a href="http://dx.doi.org/10.11636/Record.2011.020" title="NGSA geochemical atlas and dataset" target="_blank">geochemical atlas and dataset</a>. The present dataset contains additional mineralogical data obtained on NGSA samples selected from the Darling-Curnamona-Delamerian (<a href="https://www.ga.gov.au/eftf/projects/darling-curnamona-delamerian" title="DCD website" target="_blank">DCD</a>) region of southeastern Australia for the first partial data release of the Heavy Mineral Map of Australia (HMMA) project. The HMMA, a collaborative project between Geoscience Australia and Curtin University underpinned by a pilot project establishing its feasibility, is part of the Australian Government-funded Exploring for the Future (<a href="https://www.ga.gov.au/eftf" title="EFTF website" target="_blank">EFTF</a>) program. The selected 223 NGSA sediment samples fall within the DCD polygon plus an approximately one-degree buffer. The samples were taken on average from 60 to 80 cm depth in floodplain landforms, dried and sieved to a 75-430 µm grainsize fraction, and the contained heavy minerals (HMs; i.e., those with a specific gravity >2.9 g/cm³) were separated by dense fluids and mounted on cylindrical epoxy mounts. After polishing and carbon-coating, the mounts were subjected to automated mineralogical analysis on a TESCAN® Integrated Mineral Analyzer (TIMA). Using scanning electron microscopy and backscatter electron imaging integrated with energy dispersive X-ray analysis, the TIMA identified over 140 different HMs in the DCD area. The dataset, consisting of over 29 million individual mineral grains identified, was quality controlled and validated by an expert team. The data released here can be visualised, explored and downloaded using an online, bespoke mineral network analysis tool (<a href="https://geoscienceaustralia.shinyapps.io/mna4hm/" title="MNA website" target="_blank">MNA</a>) built on a cloud-based platform. Accompanying this report are a data file of TIMA results and a mineralogy vocabulary file. When completed in 2023, it is hoped the HMMA project will positively impact mineral exploration and prospectivity modelling around Australia, as well as have other applications in earth and environmental sciences.