Solid-solution metal partitioning in the Humber rivers: application of WHAM and SCAMP

Abstract

The solid-solution partitioning of five trace metals (Co, Ni, Cu, Zn and Pb) in four LOIS rivers has been modelled using a predictive chemical speciation code (WHAM-SCAMP). Observed (log K D,obs) and predicted (log K D,pred ) log K D values were similar for Zn. For Co and Ni, the log K D,pred values were typically greater than the log K D,obs values, while for Cu and Pb the reverse was seen. Removing modelled competition by Ca and Mg for binding sites on particulate organic matter increased the log K D,pred values for all the metals except Pb, and gave better agreement between the observations and the predictions for Co, Ni and Zn. Modelling solution iron as iron oxide particles decreased the log K D,pred values for Pb. The Cu predictions were very sensitive to the metal binding strength of dissolved and particulate organic matter. Overall, the model shows promise for the prediction of solid-solution metal partitioning in aquatic systems. The modelling exercise has identified the following uncertainties: the extent of Ca and Mg competition for binding sites, the chemical nature of the measured particulate metal, the efficacy of the solid-solution separation method and the strength of copper binding to organic matter.