Fine Sediment Infiltration

Bed mobility (the movement of sediment on the bed of the river) can be increased or decreased by fine sediment infiltration. Particles on the bed that protrude into the flow have more force acting on them than particles that don't protrude, but when smaller particles fill in the voids between particles, the protruding particles are less mobile (Kirchner 1990).

If particles do not protrude into the flow, the water can move faster because there is less turbulence (Ikeda 1984). When the particles disturb the flow, the direction of the water is not straight and may cause turbulent eddies (areas of flow in the opposite direction). When sediment is not protruding into the flow there is less turbulence.

Theoretical modeling and flume studies provide a framework (a way to think about a process) for fine sediment infiltration where the extent of infiltration is largely dictated by the relative sizes of the grains and supply rate of such grains (Figure 1). If the infiltrating grains are small, they can fit anywhere. If the grains on the bed are large, it is easier to accommodate all types of infiltrating grains in the void spaces.

The rate at which sediment is delivered controls in part the efficiency and depth of infiltration. Using shopping carts as an example, if a lot of people are rushing to one spot everything gets jumbled and people are not lined up in the most efficient way. If people slowly move into place a few at a time, the line is more orderly and efficient. In a simple system, fine sediment is fairly well understood. But how can we test this framework in a real river?

To conduct a test similar to a flume at a larger scale, a large amount of sediment needs to be released into a river system. Dams on rivers stop sediment from moving downstream and end up storing a lot of the sediment in the reservoir behind the dam. When dams are removed, people are concerned about how the sediment moves downstream. In this way, a dam removal is comparable to a flume experiment but incorporates more of the complexity of a large channel such as eddies, islands, and a larger scale. In more complex channels, can we predict what will happen?

The best way to test the validity of the framework is to use a tracer in a complex system, ideally naturally occurring. One example would be the Milltown Dam in Missoula, Montana that is part of a Superfund site. The Milltown Dam was built in 1908 at the confluence of the Blackfoot and Clark Fork rivers near Missoula, Montana (Figure 2). In 1909, a 100-year recurrence interval flood occurred filling in much of the reservoir with tailings from mining operations upstream in the Butte-Anaconda Mining District. In 1983, elevated levels of trace metals (As, Cu, Pb and Zn) in the water and sediment found within the reservoir led to its designation as an extended National Priority Superfund Site. The removal of the Milltown Dam was part of the remediation effort. The most contaminated sediments from the reservoir were removed and a new channel constructed. During and after the dam removal, the channel adjusted and pulses of sediment fluxed downstream.

To examine fine sediment infiltration, we used the metal signature from the reservoir and the influx of sediment to examine fine sediment infiltration in the bed of the river downstream from the dam, where the river has a couple of different channels (multi-thread). To do this we used freeze cores (Figure 3), sediment infiltration bags and bulk sampling to collect sediment for metal analysis. We hypothesized that high metal concentrations would reflect infiltration of metal-contaminated sediment from the reservoir release.

Although sediment with elevated metal concentrations was found on the banks and floodplains in the field site, metal concentrations in the bed of the river were relatively lower. The metal concentrations on the banks and floodplain demonstrate that contaminated sediment moved through the system even though evidence of fine sediment infiltration was not found in the bed. The lack of high metal concentrations in the bed suggests limited fine sediment infiltration and suggests that a conceptual framework for fine sediment infiltration needs to incorporate larger scale processes when examining fine sediment infiltration in natural systems. 


Bed mobility and sediment routing are important ways fine sediment infiltration is impacted by larger scale processes that are not fully addressed in lab experiments. For fine sediment infiltration to occur, larger grains must be present on the bed of the river. Bed mobility is important because if the large grains are moved either before or after infiltration occurs, the fine sediment is reworked and transported further downstream.

Within the Clark Fork system, the bed mobility was so high that the bed of the river was largely reworked during high flows. The results of our study also reflect how the most highly contaminated sediment was deposited downstream. The smallest particles tend to have the highest surface area per volume and therefore the higher metal concentrations. The finest particles are also preferentially deposited on the banks and floodplain, and less frequently in-channel. Testing geomorphic ideas in a large, natural system added to our understanding of fine sediment infiltration.

• Cui, Y., Wooster, J.K., Baker, P.F., Dusterhoff, S.R., Sklar, L.S., and Dietrich, W.E., 2008, Theory of Fine Sediment Infiltration into Immobile Gravel Bed: Journal of Hydraulic Engineering, v. 134, p. 1421.
• Ikeda, H., 1984, Flume experiments on the transport of sand-gravel mixtures: Bulletin of the Environmental Research Center, University of Tsukuba, v. 8, p. 1–15.
• Kleinhans, M.G., 2010, Sorting out river channel patterns: Progress in Physical Geography, v. 34, no. 3, p. 287.
• Lisle, T.E., 1989, Sediment transport and resulting deposition in spawning gravels, north coastal California: Water Resources Research, v. 25, no. 6, p. 1303–1319.