Singularity theories and methods for characterizing mineralization processes and mapping geo-anomalies for mineral deposit prediction

In this paper, we show that geo-anomalies can be delineated for mineral deposit prediction according to singularity theories developed to characterize nonlinear mineralization processes. Associ- ating singularity and geo-anomalies makes it possible to quantitatively study geo-anomalies with modern nonlinear theories and methods. This paper introduces a newly developed singularity analysis of nonlinear mineralization processes and nonlinear methods for characterizing and mapping geo-anomalies for mineral deposit prediction. Mineral deposits, as the products of singular mineralization processes caused by geo-anomalies, can be characterized by means of fractal or multifractal models. It has been shown that singularity can characterize the degree of geo-abnormality, and this has been demonstrated to be useful for mapping anomalies of undiscovered mineral deposits. The study of mineralization and mineral deposits from a nonlinear process point of view is a new but promising research direction. This study emphasizes the relationships between geo-anomalies and singularity, including singular processes resulting in singularity and geo-anomalies, the characterization of singularity and geo-anomalies and the identification of geo-anomalies for mineral deposit prediction. The concepts and methods are demon- strated using a case study of Sn mineral deposit prediction in the Gejiu mineral district in Yunnan, China.

Hydrocarbon Micro-Seepage Detection by Altered Minerals Mapping from Airborne Hyper-Spectral Data in Xifeng Oilfield,China

Hydrocarbon micro-seepage can cause oxidation reduction reactions and produce altered minerals in surface sediments and soft. The typical altered minerals mapping by their diagnostic spectral features on hyper-spectral images is an important tool for the petroleum exploration industry. In this study, the airborne hyper-spectral data were used to investigate the altered minerals induced by hydrocarbon micro-seepages by spectral feature fitting （SFF） in the loess coverage area of Xifeng Oflfield. The results re- veal that the distribution region of the altered minerals induced by hydrocarbon micro-seepage is larger than the known oilfield exploration area. The potential hydrocarbon micro-seepage region was also re- vealed by the distribution of altered minerals besides the known hydrocarbon area. A fast index was pro- posed by the absorption depths of clay and carbonate minerals for assessment of hydrocarbon micro- seepage. And it gave much clearer boundaries for the hydrocarbon micro-seepage in the loess coverage area than those by the altered mineral mapping. In addition, some field samples were analyzed by X-ray diffrac- tion （XRD） and atomic absorption spectrophotometer to validate the results. Within the extents of hydro- carbon micro-seepage, there are lower contents of ferric iron and higher contents of carbonate minerals in these samples. Therefore, it is satisfactory to have the airborne hyper-spectral data to outline the extents of hydrocarbon micro-seepage for further hydrocarbon exploration in the loess coverage area.

Mineral Prospectivity Prediction via Convolutional Neural Networks Based on Geological Big Data

Today’s era of big data is witnessing a gradual increase in the amount of data,more correlations between data,as well as growth in their spatial dimension.Conventional linear statistical models applied to mineral prospectivity mapping(MPM)perform poorly because of the random and nonlinear nature of metallogenic processes.To overcome this performance degradation,deep learning models have been introduced in 3 D MPM.In this study,taking the Huayuan sedimentary Mn deposit in Hunan Province as an example,we construct a 3 D digital model of this deposit based on the prospectivity model of the study area.In this approach,3 D predictor layers are converted from the conceptual model and employed in a 3 D convolutional neural network(3 D CNN).The characteristics of the spatial distribution are extracted by the 3 D CNN.Subsequently,we divide the 22 extracted ore-controlling variables into six groups for contrast experiments based on various combinations and further apply the 3 D CNN model and weight of evidence(WofE)method on each group.The predictive model is trained on the basis of the coupling correlation between the spatial distributions of the variables and the underground occurrence space of the Mn orebodies,and the correlation between different ore-controlling factors.The analysis of 12 factors indicates that the 3 D CNN model performs well in the 3 D MPM,achieving a promising accuracy of up to 100%and a loss value below 0.001.A comparison shows that the 3 D CNN model outperforms the WofE model in terms of predictive evaluation indexes,namely the success rate and ore-controlling rate.In particular,the 1–12 ore-controlling factors selected in experiment 5 provide a significantly better prediction effect than the other factors.Consequently,we conclude that the Mn deposit in the study area is not only related to the stratum and interlaminar anomalous bodies but also to the spatial distribution of the faults.The experimental results confirm that the proposed 3 D CNN is promising for 3 D MPM as it eliminates the interference factors.

Unsupervised ore/waste classification on open-cut mine faces using close-range hyperspectral data

The remote mapping of minerals and discrimination of ore and waste on surfaces are important tasks for geological applications such as those in mining.Such tasks have become possible using ground-based,close-range hyperspectral sensors which can remotely measure the reflectance properties of the environ-ment with high spatial and spectral resolution.However,autonomous mapping of mineral spectra mea-sured on an open-cut mine face remains a challenging problem due to the subtleness of differences in spectral absorption features between mineral and rock classes as well as variability in the illumination of the scene.An additional layer of difficulty arises when there is no annotated data available to train a supervised learning algorithm.A pipeline for unsupervised mapping of spectra on a mine face is pro-posed which draws from several recent advances in the hyperspectral machine learning literature.The proposed pipeline brings together unsupervised and self-supervised algorithms in a unified system to map minerals on a mine face without the need for human-annotated training data.The pipeline is eval-uated with a hyperspectral image dataset of an open-cut mine face comprising mineral ore martite and non-mineralised shale.The combined system is shown to produce a superior map to its constituent algo-rithms,and the consistency of its mapping capability is demonstrated using data acquired at two differ-ent times of day.

Spectral clustering eigenvector selection of hyperspectral image based on the coincidence degree of data distribution

Spectral clustering is a well-regarded subspace clustering algorithm that exhibits outstanding performance in hyperspectral image classification through eigenvalue decomposition of the Laplacian matrix.However,its classification accuracy is severely limited by the selected eigenvectors,and the commonly used eigenvectors not only fail to guarantee the inclusion of detailed discriminative information,but also have high computational complexity.To address these challenges,we proposed an intuitive eigenvector selection method based on the coincidence degree of data distribution(CDES).First,the clustering result of improved k-means,which can well reflect the spatial distribution of various types was used as the reference map.Then,the adjusted Rand index and adjusted mutual information were calculated to assess the data distribution consistency between each eigenvector and the reference map.Finally,the eigenvectors with high coincidence degrees were selected for clustering.A case study on hyperspectral mineral mapping demonstrated that the mapping accuracies of CDES are approximately 56.3%,15.5%,and 10.5%higher than those of the commonly used top,high entropy,and high relevance eigenvectors,and CDES can save more than 99%of the eigenvector selection time.Especially,due to the unsupervised nature of k-means,CDES provides a novel solution for autonomous feature selection of hyperspectral images.