River sediments have the effect of aggregating geochemical environmental information, such as that related to geological and artificial pollution resulting from mine closure. This information comprises high-dimensional data and is related to the distribution and quantities of elements in river sediments. However, accessing and interpreting this geochemical information can be difficult. This study employed a data-driven analysis that can be mathematically and statistically reduced in dimension. Using high-dimensional geochemical and environmental information on river sediments, this study evaluated the environmental impact of closed mines. Sample for analysis were collected from three rivers. There are differences in the existence of mines and mine wastewater treatment methods in this river. A total of 33 elements were measured in river sediments. Frequency distribution analysis and Principal component analysis revealed that the elements had unique distribution and frequency characteristics in each river catchment. Four environmental factors could be extracted from the relationship of elements due to lower dimension. PC1 was influenced by the land use in the river area. PC2 captured the geological background. PC3 captured the mixing-diluting effect that occurs in rivers. PC4 effectively captured the effects of domestic wastewater and the effects of closed mines. The effects of the closed mines could be confirmed using the PC4 score for the Okawa River and the Akagawa River. By examining the elemental relationships obtained using these mathematical methods, it is possible to infer the effect of geological features and mines on sediment physiochemistry using existing data on river sediments.
Keywords: Closed mines; Data-driven analysis; Geochemical information; Principal component analysis (PCA); River sediments.
© 2021 Published by Elsevier Ltd.