Effects of sediment transport and seepage direction on hydraulic properties at the sediment-water interface of hyporheic settings

Abstract

Relations between seepage flux and hydraulic properties are difficult to quantify in fluvial settings because of the difficulty in measuring these variables in situ. Tests conducted in a 1.5-m diameter by 1.5-m tall sediment-filled tank indicate that hydraulic gradient increased and hydraulic conductivity (K) decreased following the onset of downward seepage but both parameters were little changed following the onset of upward seepage. Reductions in K during downward seepage were more pronounced when surface-water current was sufficient to mobilize sediment on the bed. Averaged ratios of K determined during upward seepage to K determined during downward seepage (K up /K down ) through a sand-and-gravel bed increased from 1.4 to 1.7 with increasing surface-water velocity, and decreased to slightly greater than 1 when the sediment bed became fully mobile. K up /K down for tests conducted with a silt veneer on the bed surface was greater than 2 for all but the fastest surface-water velocities. Substantial reductions in K also were associated with a silt floc that formed on the bed surface during and following test runs. Although the silt floc was typically less than 0.5 mm in thickness, most of the hydraulic gradient was distributed across this thin layer. K of the thin silt floc was reduced by two to three orders of magnitude relative to the underlying sediment. Directional bias in K and relation between K and surface-water velocity require the presence or absence of a layer of lower-K sediment at or near the bed surface, without which no reduction in K and corresponding increase in hydraulic gradient can occur at the bed surface. The lack of prior observation of the consistent bias in K associated with seepage direction is somewhat surprising given the numerous studies where K has been measured in fluvial settings, but may be explained by the small value of the bias relative to the typical uncertainty associated with field determinations of K. If shown to exist in field settings, this bias and its relation to fluvial processes will be relevant to many studies conducted in hyporheic settings that require determination of fluxes across the sediment-water interface.