There is presently no well-tested, consistently accurate equation for calculating the resistance coefficients of steep channels (gradient > 0.002). The estimation of total resistance and the indirect estimation of discharge are complicated by steep gradients, poorly-sorted beds, coarse particles, localized flow transitions, and stage-dependent forms of resistance. The numerous uncertainties in estimating roughness coefficient and discharge for steep channels indicate the need for a systematically-designed program of data collection from various types of steep channels during a range of discharges. The primary objectives of the proposed research are to directly measure (i) the channel boundary configuration as this influences hydraulic roughness, and (ii) the hydraulic variables of velocity, hydraulic radius, and water-surface slope in various steep channel types during a range of discharges. We will focus on the steepest subset of high-gradient channels, which have cascade, step-pool, or plane-bed channel morphology. Secondary research objectives are to (i) calculate reach-scale total resistance using the Darcy-Weisbach equation to test for consistent differences between channel types, and evaluate rate of change in total resistance with stage as a function of channel type, (ii) evaluate magnitude and variability of different resistance components as a function of channel type, (iii) test the possibility that reach-scale total resistance can be more accurately assessed using parameters other than reach-averaged values of R and S, and (iv) use pressure sensors and dataloggers to develop magnitude-frequency-duration records of flow that can be used to evaluate parameters such as the temporal distribution of stream power in relation to hydroclimatic regime at the field sites. We will use 90 sets of field data from 30 study reaches (10 in each channel type) with lengths several times the average channel width. Field sites will be located along East St. Louis Creek and North St. Vrain Creek, both in Colorado, and along the Lookout Creek drainage basin in Oregon. East St. Louis and North St. Vrain Creeks are snowmelt-dominated streams, whereas the middle and lower reaches of Lookout Creek are dominated by winter rain-on-snow floods that produce flashier hydrographs. Analyses of field data will focus on testing hypotheses, and on using linear regression and multiple regression techniques to explore relations among control variables (w, h, R/D84, S, other measures of grain size such as D84 or D90) and the response variables of f and fg. We will also employ dimensional analysis in which f depends on dimensionless groups including alternative measures of R/D and w/ksbank, as well as water-surface slope, ksbed/ksbank, sorting of bed material, Reynolds number, Froude number, flow blockage, and wood drag per channel area. The robustness of competing regression models will be assessed using Mallow's Cp, AIC, cross-validation techniques, and process-based interpretation. Earth scientists, stream biologists, and land and resource management agencies increasingly attempt to estimate flow, sediment yields, and channel stability in steep channels in relation to the design of culverts and bridges, fish passage, and stream stabilization or rehabilitation measures. The ability to indirectly estimate flow resistance, velocity and/or discharge is critical to understanding and managing steep channels. Results from this work will be of great use to the water-resources management, engineering, and habitat management communities.
