Channel-forming discharge selection in river restoration design

The concept of channel-forming (Q(cf)) or dominant discharge is now a cornerstone of river channel restoration design. Three measures of channel-forming discharge are most commonly applied: effective discharge (Q(eff)), bankfull discharge Q(bf), and d discharge of a certain recurrence interval (Q(ri)), which theoretically are similar in geomorphically stable channels. The latter two measures have become particularly widely applied in some channel restoration design procedures, often to the exclusion of Q(eff) analyses, despite the additional utility of Q(eff) analysis for most channel design problems. We quantify the three measures of Q(cf) for four case studies and then follow this with a synthesis of previously published studies to illustrate sources of variability. This synthesis suggests that agreement among the three measures of Q(cf) is best for snowmelt-hydrology, nonincised channels with coarse substrate. Departures from these conditions result in greater discrepancy between the measures. Channel incision produces Q(bf) far greater than Q(eff), and flashy hydrology is associated with generally larger, briefer, and more frequent Q(eff). Regional mean or median values for the relative magnitudes of the three measures can be tightly constrained, but site to site variation is quite large. The construction of a cumulative sediment discharge curve and associated determination of Q(eff) allows quantification of the sediment budget of a channel for a given hydrologic regime, which provides process-based insight of drivers of current and future trajectories of channel stability, and is thus the recommended measure of channel-forming discharge. Reliance on only return-interval or bankfull discharge for channel design is not recommended for channel design activities.