Suspended Sediment Moves 10 km Before Entering Storage: Re-Interpreting a 20th Century Industrial Mercury Release as a Tracer Experiment, South River, Virginia

Abstract:

Recent analyses suggest that the velocity of downstream transport of suspended sediment (averaged over long timescales that include periods of transport and storage in alluvial deposits) can be represented as the ratio L_s/T, where L_s is a distance particles move before entering storage and T is the waiting time particles spend in storage before being remobilized. Sediment budget analyses suggest that L_s is 1-100 km in the mid-Atlantic region, while T may be ~10³ years, such that particles move 3-5 orders of magnitude slower than the water in the channel. Given the well-known inaccuracy of sediment budgets, independent verification from a tracer study would be desirable. Here, an historic industrial release of mercury is interpreted as a decadal sediment tracer experiment, releasing sediment particles “tagged” with mercury that are deposited on floodplains. As expected, floodplain mercury inventories decrease exponentially downstream, with a characteristic decay length of 10 km (95% confidence interval: 5-25 km) that defines the distance suspended particles typically move downstream before entering storage. Floodplain mercury inventories are not significantly different above and below three colonial age mill dams (present at the time of mercury release but now breached), suggesting that these results reflect ongoing processes. Suspended sediment routing models that neglect long-term storage, and the watershed management plans based on them, may need revision.