Effective diffusivity and mass flux across the sediment-water interface in streams

The exchange of water between a stream and its hyporheic zone (defined as the sediment beneath and immediately adjacent to a stream) underpins many ecological and hydrological functions in turbulent streams. Hyporheic exchange can be parameterized in terms of an effective diffusion coefficient D-eff and considerable effort has gone into developing process-based models and empirical correlations for predicting the value of this transport parameter. In this paper we demonstrate previous laboratory estimates for D-eff can be biased by as much as a factor of 10, due to errors in the equations and/or ambiguities in the variables used to reduce data from transient tracer experiments in flow-through and recirculating flumes. After correcting these problems, an analysis of 93 previously published flume experiments reveals D-eff depends on properties of the tracer (molecular diffusivity), flow field (shear velocity, kinematic viscosity), and sediment bed (permeability and depth). The shear velocity depends implicitly on the Darcy-Weisbach friction factor, which captures the influence of bed roughness and bed forms on hyporheic exchange in both laboratory and field studies. The dependence of D-eff on sediment bed depth is consistent with the hypothesis that coherent turbulence in the water column drives mass transport across the sediment-water interface. Furthermore, the dependence of D-eff on sediment bed depth raises the possibility that hyporheic exchange rates measured in the laboratory are not representative of hyporheic exchange rates in the field.