Search

This paper describes selected results from over a dozen collaborative projects led by Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia spanning a 30-year history of developments in satellite, airborne, field and drill core sensing technologies and how these can assist explorers to measure and map valuable mineral information. The exploration case histories are largely from Australian test sites and describe how spectral sensing technologies have progressed from early "niche creation" systems, such as the field PIMA-II (Portable Field Mineral Analyzer) and airborne Geoscan, HyMap™ and OARS-TIPS (Operational Airborne Remote Sensing - Thermal Infrared Profiling Spectrometer) systems and drill-core HyLogger™ systems, to the current expanding array of pubic and commercial mineral mapping sensors, including the ASTER (Advanced Spaceborne Thermal Emission and Reflectance Radiometer) satellite system which has acquired imagery spanning the entire Earth's land surface (<83°latitude). These sensors are delivering voluminous spectral data from different parts of the visible to the thermal infrared (400 to 14,000 nm) spectrum at different spectral, radiometric and spatial resolutions. Two critical exploration challenges are central to the case histories, namely: (i) can surface cover, such as vegetation, regolith or transported materials, be characterized and accounted for so that the target geology is accurately revealed; and (ii) does this revealed geology show evidence of alteration footprints to potential economic mineralization. Spectrally measurable minerals important to solving these challenges include white micas, kaolinite and garnets, with measurement of their respective physicochemistries being key. For example, kaolin disorder is useful for mapping transported versus weathered in situ materials, while the chemical substitution in white micas and garnets provide vectors to potential economic mineralization. Importantly, appropriate selection of the optimum sensor/data type for a given geological application depends primarily on the level of detail/accuracy of the mineral information required by the user. A major opportunity is to now harness the many sensor/data types and deliver to users consistent, accurate mineral information products, that is, creation of a number of valuable global mineral product standards. As part of this vision, CSIRO has been developing improved sensor/data calibration processes and information extraction methods that for example, unmix the target mineralogy from green and dry vegetation cover in remote sensing data sets. Emphasis to date has been on generating public spectral-mineral product standards, especially at ASTER's limited but geologically-valuable spectral resolution. The results are showing that scalable, global, three-dimensional (3D) mineral maps are achievable which will only improve our ability to more accurately characterize regolith and geological architecture, increase our understanding of formative processes and assist the discovery of new economic mineral systems. [ABSTRACT FROM AUTHOR]

This paper proposes an automated active fire detection framework using Sentinel-2 imagery. The framework is made up of three basic parts including data collection and preprocessing, deep-learning-based active fire detection, and final product generation modules. The active fire detection module is developed on a specifically designed dual-domain channel-position attention (DCPA)+HRNetV2 model and a dataset with semi-manually annotated active fire samples is constructed over wildfires that commenced on the east coast of Australia and the west coast of the United States in 2019–2020 for the training process. This dataset can be used as a benchmark for other deep-learning-based algorithms to improve active fire detection accuracy. The performance of active fire detection is evaluated regarding the detection accuracy of deep-learning-based models and the processing efficiency of the whole framework. Results indicate that the DCPA and HRNetV2 combination surpasses DeepLabV3 and HRNetV2 models for active fire detection. In addition, the automated framework can deliver active fire detection results of Sentinel-2 inputs with coverage of about 12,000 km2 (including data download) in less than 6 min, where average intersections over union (IoUs) of 70.4% and 71.9% were achieved in tests over Australia and the United States, respectively. Concepts in this framework can be further applied to other remote sensing sensors with data acquisitions in SWIR-NIR-Red ranges and can serve as a powerful tool to deal with large volumes of high-resolution data used in future fire monitoring systems and as a cost-efficient resource in support of governments and fire service agencies that need timely, optimized firefighting plans. [ABSTRACT FROM AUTHOR]