Bacterial and sediment transport in an artificial sand bed stream during unsteady flow

Storms cause a substantial increase in the fecal indicator bacteria (FIB) concentrations in stream water. Causes for this concentration increase include FIB-laden stormwater runoff and the release of bacteria from stream-bottom sediments. Studies were carried out to quantify this second cause of water quality impairment. Coliform bacteria are a group of FIB, indicate the presence of pathogenic microorganisms, and have been shown to be transported in streams far from their source areas, especially during storms. FIB transport is dependent on several factors including water velocity, size and transport rates of bed sediments, and the FIB concentration in bed sediment. The objective of this work is to determine the effect of varying flow rates on the transport of bed-dwelling coliforms in sand bed streams.;Artificial stream experiments were run with field-induced storm hydrographs to determine resuspension rates of coliforms and their association with the movement of bed material. Six artificial storm events were created by varying discharge in a 0.63-m wide stainless steel flume with 2.5 m of the 8-m-long channel bed covered with 0.26 to 0.375-mm sand. Peak flow rates varied from 0.04 to 0.074 m3/s, peak velocities were 0.3 to 0.65 m/s, and event durations were either 60 min or 90 min. Water samples were collected upstream and downstream of the sand bed at selected intervals throughout each event, and bed sediment samples were collected before and after each event. Average FIB concentrations in water up- and downstream of the sand bed were statistically similar for events with peak velocities less than 0.58 m/s. Peak concentrations of total coliform and suspended sediments occurred in the downstream samples during the rising limb of hydrographs, due to resuspension of coliforms and sand from the bed. However, after the initial surge, there was no noticeable effect of the water flows on resuspension. To simulate downstream concentrations due to resuspension, we modeled the system using a 1-D advection-dispersion equation with added source/sink terms to account for sediment settling and resuspension. This experiment and its analysis may provide a methodology for determining coliform and suspended sediment inputs into total maximum daily load (TMDL) calculations.