Long-term source apportionment and assessment of metals in lake sediments contaminated by mining and smelting in South Korea

• Historical mining and smelting caused long-term metal contamination in Andong Lake sediments. • Sediment core profiles revealed dual contamination phases linked to different sources. • Cd and Zn showed distinct increases post-2005 due to smelting-related contributions. • Li-normalized metals were used to quantify contributions from mining and smelting sources. • This study provides a framework for tracing legacy pollution in lake sediments. Understanding the long-term evolution and sources of metal contamination in lacustrine environments is essential for assessing ecological risks and informing remediation. This study reconstructs the historical and spatial distribution of heavy metals in sediments of Andong Lake, an artificial reservoir in South Korea impacted by legacy mining and smelting activities. Eight sediment cores were collected along the lake’s main axis and analyzed for geochemical properties and concentrations of cadmium (Cd), zinc (Zn), copper (Cu), lead (Pb), arsenic (As), nickel (Ni), manganese (Mn), cobalt (Co), and lithium (Li). Sediment chronology was constrained using 210 Pb dating and stratigraphic boundaries corresponding to dam construction in 1976. Vertical metal profiles revealed two contamination phases: (1) substantial enrichment of Cd, Zn, Cu, Pb, Ni, Mn, and Co during 1976–1995, linked to discharges from upstream Pb–Zn mines (mining-derived materials, MDMs), and (2) renewed increases in Cd and Zn since 2005, associated with emissions from a Zn smelter (smelting-derived materials, SDMs). By integrating Li-normalized metal ratios and historical production data, the relative contributions of MDMs and SDMs to baseline sediments (1995–2005) were quantified. MDMs contributed up to 42.6 % of Cd and 50.0 % of Zn during the early post-dam phase (1976–1995), while SDMs accounted for approximately 42.8 % of Cd and 29.7 % of Zn in recent decades (2005–2020). This study establishes a transferable framework that integrates sediment chronology, normalized geochemical indicators, and industrial production records to trace long-term metal contamination in large artificial lakes—providing new insights for pollution source apportionment and reservoir management.