Introduction to the thematic collection: applications of innovations in geochemical data analysis

Thematic collection: This article is part of the Applications of Innovations in Geochemical Data Analysis collection available at: https://www.lyellcollection.org/topic/collections/applications-of-innovations-in-geochemical-data-analysis

Geochemical data record various descriptive and indicative clues to mineralization and have long been considered as an important information carrier in support of mineral exploration (Grunsky and Caritat 2020). Especially in geographic information system (GIS)-based mineral exploration targeting, geochemical data can be manipulated easily and delicately analysed by geostatistical methodology (Bonham-Carter 1994; Carranza 2009). Analysis of geochemical data plays a significant role in narrowing down target areas and increasing the possibility of discovering mineral deposits. As mineral exploration goes further, exposed and/or shallowly buried deposits might no longer be adequate to maintain the fast development of the social economy. Seeking new discoveries in covered areas and deep Earth puts new requirements on innovation of both geochemical data and its analytical methodology (Cheng 2012). Geochemical signatures of deeply buried mineral resources become more difficult to characterize by conventional strategies. Within this context, Professor Qiuming Cheng and his colleagues have contributed many efforts on developing geochemical data analytical methods. Nonlinear models including concentration–area (C–A), perimeter–area (P–A) (Cheng et al. 1994), spectrum–area (S–A) (Xu and Cheng 2001), multifractal singular value decomposition (MSVD) (Li and Cheng 2004), and local singularity analysis (LSA) (Cheng 2007) have been proposed and applied to identify and map mineralization-related geo-anomalies. In 2020, the Association of Applied Geochemists (AAG) awarded the Gold Medal to Professor Qiuming Cheng in recognition of his innovative and original contributions in geochemical data analysis and applications (https://www.appliedgeochemists.org/2020-quiming-cheng).

Geochemistry: Exploration, Environment, Analysis (GEEA), as the co-owned scientific journal of the Geological Society of London and the AAG, focuses on the use of geochemistry in mineral exploration/resource development. Since the first issue of GEEA was published in 2001, there have been hundreds of papers regarding methodologies for geochemical data analysis. In 2017, a thematic collection: analysis of exploration geochemical data for mapping of anomalies (Carranza and Zuo 2017) was published on the development of methods and techniques for exploration geochemical data analysis. As a successor, this current thematic collection introduces some new progress and advancements in geochemical data analysis, e.g. nonlinear, fractal/multifractal, multi-statistical, machine learning methods and their applications in mineral exploration. Twelve papers from Professor Qiuming Cheng's colleagues in China and Canada are organized in this thematic collection in honour of his Gold Medal Award.

Wang and Zhu (2022) review local singularity analysis (LSA) from its original proposition to recent progress. From the perspective of limitations in current singularity index estimation algorithms, analytical window parameters in U-statistics and algorithms in local singularity are integrated, establishing an irregular window-based singularity analysis. Practical application in the Duolong mineral district, northern Tibet, China, well-demonstrates its efficiency for weak anomaly identification compared with that of the original algorithm. Meanwhile, the multiscale irregular windows supplement delicate patterns descriptive of anisotropic natures associated with mineralization.

Zuo and Yin (2022) introduce some new progress of Google Earth in geochemical data visualization and interpretation. Within the easily accessible and web-based Google Earth platform, geochemical data-related sampling strategies, collection, management, visualization and public sharing are demonstrated. Integration with other exploratory datasets is discussed. Furthermore, some potentially useful add-on functions of Google Earth like Keyhole Markup Language are discussed using geochemical data from the Daqiao district, Gansu Province, China.

Xie et al. (2022) establish evidence layers based on identified geo-anomalies from stream sediment geochemical data, and further utilize back-propagation neural network and fuzzy weights of evidence to quantitatively predict potentials of Pb–Zn mineralization in Guangxi, China. Moreover, advantages of multi-information fusion techniques in eliminating influence of weathering-related secondary accumulation of elements for geo-anomaly recognition are demonstrated and discussed. According to the case study, a new idea is proposed for better manipulating geochemical anomaly delineation, evaluation and interpretation not only in Guangxi with a karst landform but also in other complex geomorphologic environments.

Y. Zhang et al. (2022a) apply multifractal methods to drillhole geochemical data in the Songnan Low Uplift of the Qiongdongnan Basin deepwater area, China. Geochemical behaviour of 13 oil and gas indices and their principal components (PC) are characterized by multifractal methods. Indices with a right-skewed arc are distinguished from the other indices with weak or single-fractal characteristics. By comparison among multifractal properties of individual and integrated indices, methane, ethane, propane and heavy hydrocarbons are determined as important indices for hydrocarbon prospecting in the Songnan Low Uplift. This study demonstrates a new application of multifractal methods in index selection and comprehensive information extraction for oil and gas exploration.

Y. Zhang et al. (2022b) utilize conventional multivariate statistical and multifractal methods to investigate mineralization-related geochemical distributions in the Central Kunlun Mountains, Xinjiang, China. Ore-forming strengths of elements are characterized by moment method-based multifractal parameters. Jointly interpreted with correlations among 12 elements explored by R-mode cluster, Pearson correlation and PC analyses, element groups associated with regional geological background and mineralization with high ore-forming strengths and potentials are determined. This study indicates that multi-parameter discrimination diagrams constructed by multifractal and multivariate statistical methods can provide new avenues for quantitative evaluation of the ore-forming potential of elements.

Li et al. (2022) apply a classic prospecting modelling method of Professor Qiuming Cheng's group consisting of PC analysis and spectrum–area (S–A) model to identify geochemical anomalies indicative of black shale-type vanadium mineralization in the Jiujiang region of the northeastern Jiangxi Province. Their results demonstrate that S–A fractal analysis decomposes effectively mixed patterns preserved in geochemical anomalies and is able to identify significant V anomalies from complex background. In addition, the highly anomalous areas are validated by discovered V deposits, based on which the northeastern parts of the Jiangxi Province that potentially host undiscovered V deposits have been proposed.

Yang et al. (2022) analyse in-situ a soil profile overlying mineralized veins cutting through the bedrock in the Dalaimiao district, Inner Mongolia, China with a portable X-ray fluorescence spectrometer. Spatial patterns of mineralization-related elements are identified by joint LSA and PC analysis. Principal components of singularity patterns well-demonstrate migration of ore elements from underlying bedrock through the soil. Moreover, higher correlation among singularity patterns of elements are stable at/in different depths and media that reflect the effects of LSA in geochemical anomaly identification, especially jointly utilized with PC analysis.

Z. Zhang et al. (2022) introduce a Monte Carlo-based framework consisting of random forest and hierarchical clustering analysis to evaluate the elemental inheritance and migration characters in the Daliangshan area, China. The results show that the main controlling factor on element distributions in C horizons during rock-weathering is geochemical compositions of parent rocks, while that in B horizons during pedogenesis is geochemical compositions of C horizons. In addition, this framework can be transferred to other local places and applied to the assessment of mineral resources, environment, and agriculture using the elemental inheritance categories.

Wang et al. (2022) introduce a hybrid modelling method consisting of sequential indicator simulation and concentration–area fractal model to investigate uncertainty in mapping of Ag geochemical anomalies in the NE of Dong Ujimqin Banner district of Inner Mongolia, North China. Analytical results demonstrate that Ag anomalies delineated by local probability and spatial joint probability are more acceptable and reliable. In comparison with joint probability statistics, which are stricter than local uncertainty, this hybrid modelling provides a valuable way to delineate anomalous areas with consideration on uncertainty of geochemical distributions.

Liu (2022), focusing on compositional balance analysis (CoBA) and its practical application to geochemical data, well-illustrate how to choose the optimal balances from the perspective of CoBA and multivariate statistical analysis. Thus, 14 compositional balances and three principal factors associated with different geochemical patterns in the western Tianshan region, China are selected for comparison. Through systematic investigation, especially when CoBA is used to recognize and map anomalous geochemical patterns, this study contributes helpful and valuable experiences on the determination of the optimal balance from a series of balances.

McCurdy et al. (2022) discuss the potential of magnetite in stream sediments as a vector to porphyry Cu mineralization through a detailed study at the Casino porphyry Cu–Au–Mo deposit, Yukon, Canada. According to linear discriminant analysis of trace element composition and Ti v. Ni/Cr diagrams, geochemical signatures of magnetite recovered from stream sediments around the deposit are compared with igneous and magmatic-hydrothermal magnetite recovered from mineralized and unmineralized host rocks. Taking the occurrence of copper in various types of magnetite into consideration, the discovery of similar composition of magnetite in stream and alteration zones strongly supports that composition of magnetite recovered from stream sediments and other detrital sediments can be a practically useful indicator mineral in locating porphyry systems.

Liang et al. (2022) use major and trace element analysis data to determine the original type and genesis of the ore-bearing gneiss and surrounding rocks in the sedimentary–metamorphic graphite belt of the North China Craton (NCC). Descriptive evidence of the evolution of graphite ore bodies is discovered and indicates that the ore deposit received and accumulated weathering products of surrounding rocks during the sedimentary period. Combined with the regional geology, potential areas for graphite mineralization are predicted. This study establishes a sedimentary-metamorophic model that improves understandings on geological origins of graphite deposits in the NCC.