Flow resistance prediction in high-gradient streams
Date
2010
Authors
Yochum, Steven Edward, author
Bledsoe, Brian P., advisor
Rathburn, Sara L., 1962-, committee member
Watson, Chester C., committee member
Wohl, Ellen E., 1962-, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Flow resistance measurements were collected on high-gradient streams in the Fraser Experimental Forest, Colorado, for bankfull through low flows using Rhodamine WT dye tracing, ground-based LiDAR scans, and laser theodolite surveying of longitudinal profiles and below-water features. A dataset of 59 resistance measurements was collected on fifteen reaches with instream wood present in varying densities. Values of Manning's n ranged from 0.05 to 0.52, and Darcy-Weisbach ƒ varied from 0.28 to 56. All measurements indicated subcritical reach-average conditions, with Froude numbers ranging from 0.15 to 0.78. Relative grain submergence (R/D84) was a poor predictor of flow resistance while relative bedform submergence, defined as the ratio of depth or hydraulic radius to the standard deviation of the residuals of a bed profile regression (hm/σz, R σz), explained up to 76 and 80 percent of the variance of n and ƒ, respectively. Both clasts and instream wood contribute to bed variability; steps are heightened by wood lodging among the clast steps. Hence relative bedform submergence captures the combined influence of wood and clasts, which contribute both form and spill resistance. Relative bedform submergence is less effective for prediction in reaches with substantial non-step-forming instream wood and in steep channels. In the steepest reaches, with slopes over about 18 percent, the data indicate a shift towards a skimming regime with a partial submergence of bedforms and a threshold reduction in flow resistance. Three-dimensional measures of geometric variability were explored, to assess the correlation of flow resistance with higher-order spatial variation due to composite effects of bedforms, large clasts, and instream wood. With the exclusion of bank effects, a normalized variable (ha3/σz3) explained 77 and 81 percent of the variance of n and ƒ, respectively. Multivariate regression models with variables describing bedforms, bankforms, and instream wood explained 87 percent of the variance of n and ƒ. On average, flow resistance due to bedforms (form and spill) are the greatest contributor to overall flow resistance in these high-gradient streams, followed by form resistance generated by bankforms, and lastly, by form resistance induced by non-step instream wood.
Description
Zip file contains data spreadsheet.
Rights Access
Subject
velocity
Streamflow -- Colorado -- Fraser Experimental Forest
step pool
Stream measurements -- Colorado -- Fraser Experimental Forest
Manning's N
Channels (Hydraulic engineering) -- Colorado -- Fraser Experimental Forest
flow resistance
Frictional resistance (Hydrodynamics) -- Colorado -- Fraser Experimental Forest
Darcy-Weisbach ƒ
Cascades (Fluid dynamics) -- Colorado -- Fraser Experimental Forest