Comparative analysis of bed resistance partitioning in high-gradient streams

Total flow resistance can be partitioned into its components of grain (ff(grain)), form (ff(step)), wood (ff(wood)), and spill (ff(spill)) resistance. Methods for partitioning flow resistance developed for low-gradient streams are commonly applied to high-gradient systems. We examined the most widely used methods for calculating each component of resistance, along with the limitations of these methods, using data gathered from 15 high-gradient (0.02 < S-0 < 0.195) step-pool, cascade, and plane-bed reaches in Fraser Experimental Forest. We calculated grain resistance using three equations that relate relative submergence (R/D-m) to ff(grain) as well as using an additive drag approach. The drag approach was also used for calculating ff(wood) and ff(step). The ff(grain) contributed the smallest amount toward all reaches at all flows, although the value varied with the method used. The Parker and Peterson (1980) equation using D-90 best represented ff(grain) at high flows, whereas the Keulegan (1938) equation using D-50 best characterized ff(grain) at base flows, giving a lower bound for grain resistance. This suggests that ff(grain) may be better represented if two grain sizes are used to calculate this component of resistance. The drag approach, which is used to calculate wood resistance, overestimated the significance of individual logs in the channel. The contribution of ff(spill) was reduced at higher flows when form drag around the step is accounted for at higher flows. We propose a method for evaluating the contribution of ff(step) that accounts for form drag around the steps once they are submerged at higher flows. We evaluated the potential sources of error for the estimation of each component of resistance. Determination of the drag coefficient was one of the major sources of error when calculating drag around wood, steps, or boulders.