Provenance and Tectonic Evolution of Bauxite Deposits in the Tethys: Perspective From Random Forest and Logistic Regression Analyses

• Published in: Geochemistry, geophysics, geosystems : G3 Vol. 24; no. 6
• Main Authors: Zhou, Jintao, Yu, Wenchao, Wei, Wei, Yang, Mingyu, Du, Yuansheng
• Format: Journal Article
• Language: English
• Published: Washington John Wiley & Sons, Inc 01.06.2023; Wiley
• Subjects: Bauxite; Carbonate rocks; Carbonates; Geochemistry; Geological time; Karst; machine learning; major elements; Metallogenesis; Modelling; Ores; Paleolatitude; Regression analysis; Subduction; tectonics; Tethyan bauxites; titanium; Titanium dioxide
• ISSN: 1525-2027; 1525-2027
• DOI: 10.1029/2022GC010745

Developments in Earth system research have led to widespread interest in multi‐sphere interactions that may have controlled bauxite formation. Although karst bauxite ores are abundant in the Tethyan Metallogenic Domain (TMD), their genesis and distribution are not yet fully understood. This study utilizes the Random Forest (RF) model to analyze geochemical data sets compiled from three bauxite provinces in the Tethys, including the Mediterranean, Central Asian, and Chinese provinces. Meaningful variables among the provenance, paleolatitude and tectonic setting that affected the formation of the bauxite deposits are identified. The Logistic Regression model is used to verify the importance of various factors. Our study shows that when RF model achieves the best classifier performance, TiO2 is the variable that has the greatest impact on the model. TiO2 content is the highest in the Central Asian Province. The high TiO2 content in bauxites indicates a provenance of magmatic origine associated with the Subduction Provenance Zone. Bauxites with TiO2 content similar to that of the continental basement are derived from siliciclastic and carbonate rocks, identified as the Orogenic Provenance Zone. This study illustrates a novel link between geochemical compositions (TiO2 content) of bauxites, provenance and regional tectonic setting. Plain Language Summary Abundant bauxite deposits formed in the Tethyan Metallogenic Domain are of great importance to bauxite research. However, previous studies generally focus on local bauxite occurrences formed over certain geologic time. A comprehensive understanding of the genesis and characteristics of bauxite deposits is hence essential. In this study, Machine learning methods (Random Forest and Logistic Regression) are applied to improve the understanding of Tethyan bauxite formation. Our research reveals that the TiO2 content in Tethyan bauxites is related to their provenance. The enrichment of TiO2 in bauxite deposits points to magmatic rock provenance that underwent subduction. Bauxites with a TiO2 content similar to that of continental basement are derived from siliciclastic and carbonate rocks exposed on orogenic belts. As such, this investigation elucidates a previously unexplored association between the geochemical characteristics (TiO2 content) of bauxite deposits, their provenance, and the regional tectonic context. Key Points Two Provenance Zones in the Tethys Domain are proposed: Subduction Provenance Zone and Orogenic Provenance Zone High TiO2 content in bauxites resulted from magmatic rocks of the Subduction Provenance Zone Ti in bauxites has the potential for tracing provenance and associated tectonic settings