J. Denis Newbold, Stroud Water Research Center, 970 Spencer Road, Avondale, PA 19311
Runkel (2007, L&O Methods) showed that longitudinal dispersion can affect the measured uptake length, SW, of solutes, and so may compromise the utility of SW in estimating uptake mass-transfer coefficients. I applied Runkel’s analysis to 57 measurements of SW in 17 streams in southeastern New York State encompassing flows of 0.02 to 7.7 m3/s. In 70% of the cases and for all flows >1 m3/s, the dispersion-induced error was <5%. In smaller streams the median error was 4%, but ranged up to 16%. Among all streams, the error was well described as a decreasing function of the dimensionless Péclet number (Pe), given by v SW /D, in which v is water velocity, and D is longitudinal dispersion. The error exceeded 4% for Pe <25 and was <4% for Pe>25. Combining predictive formulae for SW and D suggests that Pe varies as v d2/w, in which d is depth and w is width. Thus Pe generally increases with flow but, for a given flow, wider, shallower channels should be more error prone. These results suggest that dispersion is rarely a serious problem, particularly when compared to the much larger errors that arise from the saturating effects of experimental nutrient additions.
Web Page:
spiraling, nutrient uptake, longitudinal dispersion