The impact of storm events on a riverbed system and its hydraulic conductivity at a site of induced infiltration

Abstract

The spatial and temporal variability of riverbed vertical hydraulic conductivity (K(v)) was investigated at a site of induced infiltration, associated with a municipal well field, to assess the impact of high-stage events on scour and subsequently the riverbed K(v). Such impacts are important when considering the potential loss of riverbank filtration capacity due to storm events. The study site, in and along the Great Miami River in southwest Ohio, overlaid a highly productive glacial-outwash aquifer. A three-layer model for this system was conceptualized: a top layer of transient sediment, a second layer comprising large sediment resistant to scour, but clogged with finer sediment (the armor/colmation layer), and a third layer that was transitional to the underlying higher-K(v) aquifer. One location was studied in detail to confirm and quantify the conceptual model. Methods included seepage meters, heat-flow modeling, grain-size analyses, laboratory permeameter tests, slug tests and the use of scour chains and pressure-load cells to directly measure the amount of sediment scour and re-deposition. Seepage meter measured riverbed K(v) ranged from 0.017 to 1.7 m/d with a geometric mean of 0.19 m/d. Heat-transport model-calibrated estimates were even lower, ranging from 0.0061 to 0.046 m/d with a mean of 0.017 m/d. The relatively low K(v) was indicative of the clogged armor layer. In contrast, slug tests in the underlying riverbed sediment yielded K(v) values an order of magnitude greater. There was a linear relationship between scour chain measured scour and event intensity with a maximum scour of only 0.098 m. Load-cell pressure sensor data over a 7-month period indicated a total sediment-height fluctuation of 0.42 m and a maximum storm-event scour of 0.28 m. Scour data indicated that the assumed armor/colmation layer almost always remained intact. Based on measured layer conductivities and thicknesses, the overall K(v) of this conceptualized system was 1.6 m/d. Sensitivity analyses indicated that even complete scour of the armor/colmation layer would likely increase the overall K(v) only by a factor of 1.5. Most scour events observed removed only the transient sediment, having very little effect on the entire system indicating low risk of losing filtration capacity during storms. The research, however, focused on the point bar, depositional side of the river. More research of the entire river profile is necessary.