Paper No. 9
Presentation Time: 10:40 AM

PARTITIONING A STEADY STATE SEDIMENT BUDGET TO REPRESENT LONG TAILED DISTRIBUTIONS OF CONTAMINANT RESIDENCE TIMES: A MODELING APPROACH FOR ROUTING TRACERS THROUGH ALLUVIAL STORAGE RESERVOIRS


ACKERMAN, Tobias R., Department of Geological Sciences, University of Delaware, Newark, DE 19716 and PIZZUTO, James E., Department of Geological Sciences, University of Delaware, Newark, DE 19716-2544, tobiack@udel.edu

Mercury (Hg) was released into the South River, VA, from an industrial source from 1929-1950. Hg has a strong affinity for fine particles, so a budget quantifying the residence times, exchange rates, and storage mass for fine sediment can be used to model the trajectories of Hg through the alluvial valley (Malmon 2002). A 3 compartment model includes the floodplain (FP), hyporheic zone, and fine-grained channel margin (FGCM) deposits within the wetted perimeter, but it fails to fit Hg concentration histories in the FGCM and under predicts contemporary Hg loading to the channel from bank erosion. We think this is because the FP and FGCM deposits are insufficiently well-mixed for this modeling framework, and Hg is preferentially reworked in low elevation FP areas near the channel. Radiometric dates from FGCM deposits suggest that most sediments are reworked within a few years, but a portion remains in storage for decades. We therefore partition the FP and FGCM deposits into multiple reservoirs, each with a different residence time, sediment exchange flux, and storage mass. We divide the FGCM deposits into 2 sub-reservoirs based on observed age distributions, and the FP into 5 sub-reservoirs based on inundation frequency and sedimentation rates. Deposition can increase the elevation of an area on the FP into a less frequently inundated sub-reservoir, and this non alluvial mass flux must be added to the Malmon (2002) model. Deposition rates are the product of area and measured sedimentation rates. FP erosion rates are consistent with the sediment budget and conserve the mass in each sub-reservoir through time to satisfy the assumption of morphological steady-state. Our revised model successfully fits Hg concentration histories in the FGCM, reproduces contemporary loading of Hg from bank erosion, and provides reasonable estimates of Hg concentration in the water column. Thus, alluvial sediment storage reservoirs should not be assumed to be well-mixed, and sediment budgets cannot be used to route contaminants through river valleys without accounting for the long-tailed distribution of residence times that may occur in these environments.