Internal erosion of soil pipes: Sediment rating curves for soil pipes

Authors: Wilson, Glenn; Ursic, Michael - Mick; FOX, GAREY - North Carolina State University; NIEBER, JOHN - University Of Minnesota

Submitted to: Earth Surface Processes and Landforms
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/15/2020
Publication Date: 10/13/2020
Citation: Wilson, G.V., Ursic, M.E., Fox, G.A., Nieber, J.L. 2020. Internal erosion of soil pipes: Sediment rating curves for soil pipes. Earth Surface Processes and Landforms. 45(15):3902-3916. https://doi.org/10.1002/esp.5009.
DOI: https://doi.org/10.1002/esp.5009

Interpretive Summary: Soil pipes are linear voids that develop in the soil as fast flowing water erodes the inside of large pores. As the internal erosion of the pipe continues, the pipe can collapse which can result in the sudden formation of a fully mature gully or the rapid advance of an existing gully. Very few studies have measured the rates of soil erosion and sediment transport through soil pipes in the field. Sediment rating curves are used in stream monitoring to determine the amount of small sediment particles that stay suspended that are lost from a watershed but no study to date has developed suspended sediment rating curves (SRC) for soil pipes. SRCs relate the suspended sediment to the flow rate by an offset parameter, a, and the slope, b. To fully assess sediment transport also requires measurements of the coarse sediment that bounces along the channel bed but no study has measured bed-sediment transport in soil pipes. The objectives of this work were to determine the suspended sediment concentrations in soil pipes along with the pipe flow rates to develop SRCs and measure the bedload transport in soil pipes. Flumes were installed in seven gully openings formed by pipe collapse and equipped with pressure sensors to measure pipe flow rates and instruments to collect suspended sediment samples. The typical response to flow through soil pipes was an initial flush of a high concentration of suspended sediment followed by a decrease as pipe flow increased. After pipe flow peaked and flow began decreasing, the pipe flow tended to have a relatively low suspended sediment concentration. Thus, soil pipe SRCs tended to depend upon the history of flow with different relationships for the rising and falling flow rates. Flow through soil pipes are often so rapidly changing that it was difficult to capture the initial flush of suspended sediment which resulted in weak relationships in the SRCs. The slope, b, parameter tended to decrease as the offset parameter, a, increases. Both SRC parameters (a and b) were correlated to the area upslope that contributes subsurface flow to that pipe location. Sediment transported along the pipe bed appeared to be an important contribution to overall sediment transport and generally increased as the size of the storm event increased. A significant portion (11 to 31%) of the bed material that was transported was aggregates (>2 mm diameter material). While this work was the first to determine SRCs for soil pipes, better sampling and measurement techniques are needed to fill this knowledge gap and produce more accurate SRCs.

Technical Abstract: The collapse of soil pipes due to internal erosion can result in sudden formations of fully mature gullies and rapid advancement of existing gullies. Few studies have measured the rates of sediment detachment and transport through soil pipes in situ. No study to date has developed sediment rating curves (SRC) for soil pipes or made measurements of bedload transport in soil pipes. The objectives of this work were to determine the suspended sediment concentrations (SSC) in soil pipes as a function of pipeflow rate to develop SRCs and measure the bedload transport as a function of cumulative flow per storm event. H-flumes were installed in seven discontinuous gullies formed by pipe collapse and equipped with pressure transducers and acoustic sensors for pipe discharge measurements and suspended sediment samplers. The typical response to pipeflow was an initial flush of high concentration of suspended sediment followed by a decrease as pipeflow increased (rising limb of hydrograph). Pipeflows were often so dynamic that it was difficult to consistently capture the initial flush of sediment, resulting in weak to non-existent SRCs. The falling limb of the hydrograph tended to have a relatively low SSC. Thus, soil pipe SRCs tended to be better represented by hysteretic SRCs although relationships between suspended sediment concentration and flow rate were poorly represented by SRCs. A power law equation given by SSC = aQb was adopted to represent the SRC relationships. Fitting this equation to data showed a correlation between the offset, a, and the slope, b, with the slope decreasing as the offset increases. Both SRC parameters (a and b) were correlated to the contributing area of the individual pipe. Bedload appeared to be an important contributor to sediment transport and generally increased as the size of the total flow (m3) of individual events increased. A significant portion (11 to 31%) of the bedload material was gravel and aggregates (>2 mm diameter material). While this work was the first to determine SRCs for soil pipes, refined sampling and measurement techniques are needed to fill this knowledge gap and produce more accurate SRCs.