Artificial intelligence-based anomaly detection of the Assen iron deposit in South Africa using remote sensing data from the Landsat-8 Operational Land Imager

Most known mineral deposits were discovered by accident using expensive,time-consuming,and knowledgebased methods such as stream sediment geochemical data,diamond drilling,reconnaissance geochemical and geophysical surveys,and/or remote sensing.Recent years have seen a decrease in the number of newly discovered mineral deposits and a rise in demand for critical raw materials,prompting exploration geologists to seek more efficient and inventive ways for processing various data types at different phases of mineral exploration.Remote sensing is one of the most sought-after tools for early-phase mineral prospecting because of its broad coverage and low cost.Remote sensing images from satellites are publicly available and can be utilised for lithological mapping and mineral exploitation.In this study,we extend an artificial intelligence-based,unsupervised anomaly detection method to identify iron deposit occurrence using Landsat-8 Operational Land Imager(OLI)satellite imagery and machine learning.The novelty in our method includes:(1)knowledge-guided and unsupervised anomaly detection that does not assume any specific anomaly signatures;(2)detection of anomalies occurs only in the variable domain;and(3)a choice of a range of machine learning algorithms to balance between explain-ability and performance.Our new unsupervised method detects anomalies through three successive stages,namely(a)stage Ⅰ–acquisition of satellite imagery,data processing and selection of bands,(b)stage Ⅱ–predictive modelling and anomaly detection,and(c)stage Ⅲ–construction of anomaly maps and analysis.In this study,the new method was tested over the Assen iron deposit in the Transvaal Supergroup(South Africa).It detected both the known areas of the Assen iron deposit and additional deposit occurrence features around the Assen iron mine that were not known.To summarise the anomalies in the area,principal component analysis was used on the reconstruction errors across all modelled bands.Our method enhanced the Assen deposit as an anomaly and attenuated the background,including anthropogenic structural anomalies,which resulted in substantially improved visual contrast and delineation of the iron deposit relative to the background.The results demonstrate the robustness of the proposed unsupervised anomaly detection method,and it could be useful for the delineation of mineral exploration targets.In particular,the method will be useful in areas where no data labels exist regarding the existence or specific spectral signatures of anomalies,such as mineral deposits under greenfield exploration.

Single-element Anomaly Mapping from Stream Sediment Geochemical Landscapes Aided by Digital Terrain Analysis

The identification of anomalies within stream sediment geochemical data is one of the fastest developing areas in mineral exploration.The various means used to achieve this objective make use of either continuous or discrete field models of stream sediment geochemical data.To map anomalies in a discrete field model of such data,two corrections are required:background correction and downstream dilution correction.Topography and geomorphology are important factors in variations of element content in stream sediments.However,few studies have considered,through the use of digital terrain analysis,the influence of geomorphic features in downstream dilution correction of stream sediment geochemical data.This study proposes and demonstrates an improvement to the traditional downstream dilution correction equation,based on the use of digital terrain analysis to map single-element anomalies in stream sediment geochemical landscapes.Moreover,this study compares the results of analyses using discrete and continuous field models of stream sediment geochemical data from the Xincang area,Tibet.The efficiency of the proposed methodology was validated against known mineral occurrences.The results indicate that catchment-based analysis outperforms interpolation-based analysis of stream sediment geochemical data for anomaly mapping.Meanwhile,the proposed modified downstream dilution correction equation proved more effective than the original equation.However,further testing of this modified downstream dilution correction is needed in other areas,in order to investigate its efficiency further.

Deriving big geochemical data from high-resolution remote sensing data via machine learning:Application to a tailing storage facility in the Witwatersrand goldfields

Remote sensing data is a cheap form of surficial geoscientific data,and in terms of veracity,velocity and volume,can sometimes be considered big data.Its spatial and spectral resolution continues to improve over time,and some modern satellites,such as the Copernicus Programme’s Sentinel-2 remote sensing satellites,offer a spatial resolution of 10 m across many of their spectral bands.The abundance and quality of remote sensing data combined with accumulated primary geochemical data has provided an unprecedented opportunity to inferentially invert remote sensing data into geochemical data.The ability to derive geochemical data from remote sensing data would provide a form of secondary big geochemical data,which can be used for numerous downstream activities,particularly where data timeliness,volume and velocity are important.Major benefactors of secondary geochemical data would be environmental monitoring and applications of artificial intelligence and machine learning in geochemistry,which currently entirely relies on manually derived data that is primarily guided by scientific reduction.Furthermore,it permits the usage of well-established data analysis techniques from geochemistry to remote sensing that allows useable insights to be extracted beyond those typically associated with strictly remote sensing data analysis.Currently,no generally applicable and systematic method to derive chemical elemental concentrations from large-scale remote sensing data have been documented in geosciences.In this paper,we demonstrate that fusing geostatistically-augmented geochemical and remote sensing data produces an abundance of data that enables a more generalized machine learning-based geochemical data generation.We use gold grade data from a South African tailing storage facility(TSF)and data from both the Landsat-8 and Sentinel remote sensing satellites.We show that various machine learning algorithms can be used given the abundance of training data.Consequently,we are able to produce a high resolution(10 m grid size)gold concentration map of the TSF,which demonstrates the potential of our method to be used to guide extraction planning,online resource exploration,environmental monitoring and resource estimation.

Big geochemical data through remote sensing for dynamic mineral resource monitoring in tailing storage facilities

Evolution in geoscientific data provides the mineral industry with new opportunities.A direction of geochemical data generation evolution is towards big data to meet the demands of data-driven usage scenarios that rely on data velocity.This direction is more significant where traditional geochemical data are not ideal,which is the case for evaluating unconventional resources,such as tailing storage facilities(TSFs),because they are not static due to sedimentation,compaction and changes associated with hydrospheric and lithospheric processes(e.g.,erosion,saltation and mobility of chemical constituents).In this paper,we generate big secondary geochemical data derived from Sentinel-2 satellite-remote sensing data to showcase the benefits of big geochemical data using TSFs from the Witwatersrand Basin(South Africa).Using spatially fused remote sensing and legacy geochemical data on the Dump 20 TSF,we trained a machine learning model to predict in-situ gold grades.Subsequently,we deployed the model to the Lindum TSF,which is 3 km away,over a period of a few years(2015-2019).We were able to visualize and analyze the temporal variation in the spatial distributions of the gold grade of the Lindum TSF.Additionally,we were able to infer extraction sequencing(to the resolution of the data),acid mine drainage formation and seasonal migration.These findings suggest that dynamic mineral resource models and live geochemical monitoring(e.g.,of elemental mobility and structural changes)are possible without additional physical sampling.

A Fractal Measure of Spatial Association between Landslides and Conditioning Factors

Measuring the relative importance and assigning weights to conditioning factors of land- slides occurrence are significant for landslide prevention and/or mitigation. In this contribution, a fractal method is introduced for measuring the spatial relationships between landslides and condition- ing factors （such as faults, rivers, geological boundaries, and roads）, and for assigning weights to condi- tioning factors for mapping of landslide susceptibility. This method can be expressed as p=C~, where d is the fractal dimension, and C is a constant. This relationship indicates a fractal relation between land- slide density （p） and distances to conditioning factors （6）. The case of d〉0 suggests a significant spatial correlation between landslides and conditioning factors. The larger the d （〉0） value, the stronger the spatial correlation is between landslides and a specific conditioning factor. Two case studies in South China were examined to demonstrate the usefulness of this novel method.

Petrophysical Texture Heterogeneity of Vesicles in Andesite Reservoir on Micro-Scales

It is of great significance to study the spatial distribution patterns and petrophysical complexity of volcanic vesicles which determine whether the reservoir spaces of the volcanic rocks can accumulate oil and gas and enrich high yields or not.In this paper,the digital images of three different textures of vesicular andesite samples,including spherical vesicular andesite,shear deformation vesicular andesite,and secondary filling vesicular andesite,are obtained by microscopic morphology X-CT imaging technology.The spatial micro-vesicle heterogeneity of vesicular andesite samples with different textures is quantitatively analyzed by fractal and multifractal methods such as box-counting dimension and the moment method.It is found that the shear stress weakens the spatial homogeneity since vesicles rupture are accelerated,elongated directionally,and connected with one another under the strain;the secondary filling breaks the vesicles,which significantly enhances the spatial heterogeneity.In addition,shear stress and secondary filling increase the complexity of vesicle microstructures characterized by different fractal and multifractal parameters.These conclusions will provide important theoretical and practical insights into understanding the degassing of volcanic rocks and prediction of high-quality volcanic reservoirs.

Fuzzy Modeling of Surficial Uranium Prospectivity in British Columbia(Canada)with a Weighted Fuzzy Algebraic Sum Operator

This paper demonstrates knowledge-guided fuzzy logic modeling of regional-scale surficial uranium(U)prospectivity in British Columbia(Canada).The deposits/occurrences of surficial U in this region vary from those in Western Australia and Namibia;thus,requiring innovative and carefully-thought techniques of spatial evidence generation and integration.As novelty,this papers introduces a new weighted fuzzy algebraic sum operator to combine certain spatial evidence layers.The analysis trialed several layers of spatial evidence based on conceptual mineral system model of surficial U in British Columbia(Canada)as well as tested various models of evidence integration.Non-linear weighted functions of(a)spatial closeness to U-enriched felsic igneous rocks was employed as U-source spatial evidence,(b)spatial closeness to paleochannels as fluid pathways spatial evidence,and(c)surface water U content as chemical trap spatial evidence.The best models of prospectivity created by integrating the layers of spatial evidence for U-source,pathways and traps predicted at least 85%of the known surficial U deposits/occurrences in>10%of the study region with the highest prospectivity fuzzy scores.The results of analyses demonstrate that,employing the known deposits/occurrences of surficial U for scrutinizing the spatial evidence layers and the final models of prospectivity can pinpoint the most suitable critical processes and models of data integration to reduce bias in the analysis of mineral prospectivity.