Understanding trace-element mobilization in a redox mixing zone of a river¿aquifer system: Insights from a multi-tracer Bayesian mixing framework
Academic Article in Scopus
Naturally occurring arsenic (As), uranium (U), and radium-226 (226Ra) are redox-sensitive groundwater constituents that threaten drinking-water supplies worldwide. We investigated a riparian aquifer along the Nakdong River, South Korea, a regulated river system where weir-controlled stage fluctuations produce highly variable river¿aquifer exchange and convergent groundwater flow. Unlike traditional physically based models that struggle to reproduce the hydraulic complexity of regulated rivers, we applied a Bayesian groundwater-age inversion constrained by multiple tracers (3H/3He, CFCs, radiogenic 4He, and 14C) to directly quantify residence-time mixing and visualize subsurface flow structures. Young, oxygenated groundwater contained measurable U (up to 11.9 ¿g/L), whereas strongly reducing pockets hosted the elevated As (¿11.5 ¿g/L) and 226Ra (¿0.17 pg/L); the Bayesian age-mixing inversion showed that the key redox transitions appear not in these endmembers but within the intermediate zones where young oxic recharge mixes with older reducing groundwater. In these mixed interfaces, U concentrations decline sharply while As and 226Ra become attenuated through Fe(II) oxidation and co-precipitation, revealing redox reversals that cannot be inferred from the apparent age alone. Microbial community signatures were consistent with these redox niches, with Geobacter enriched in reducing pockets that also showed elevated As and 226Ra; this pattern is consistent with microbial Fe(III) reduction contributing to the release of sorbed elements. Our findings demonstrate that trace-element mobility in regulated river¿aquifer systems is governed primarily by redox processes occurring at groundwater age-mixing interfaces rather than by residence time alone. Incorporating full groundwater-age distributions within a probabilistic framework therefore provides a critical basis for predicting redox-driven contaminant behavior and improving management of drinking-water resources in dynamic riparian environments.
