ESTIMATE OF GRAVEL TRANSPORT RATES AT BANKFULL FLOW IN MOUNTAIN STREAMS
From a log-log scatterplot of worldwide data of qB,bf vs. A, a positive, slightly convex trend emerges when data are limited to central Rocky Mountain streams. The trend straightens when relating qB,bf to a modified unit stream power expression ω’ = ρ · qbf · S0.5 · %Dsub<8 that integrates the percentage of subsurface sediment < 8 mm (%Dsub<8). For most Rocky Mountain streams, measured data of qB,bf vs. ω’ plot within an envelope two orders of magnitude wide. A few outlier data in Colorado are explained by disturbances that caused overly large (after a log jam burst) and overly small (upstream gravel entrapment) transport rates. The plot shows geographical variations: small step-pool streams in the forested Pacific Northwest transport little gravel per unit width at bankfull flow. The upward extension of the data envelope encloses Alpine step-pool and plane-bed streams where large basin portions are above tree line. However, most mountain torrents (wide gravel-cobble beds with an incised step-pool low-flow channel) have larger bankfull transport rates than Rocky Mountain streams.
The plotted relation of qB,bf vs. ω’ facilitates an estimate of bankfull unit transport rates, especially if watershed sediment supply (e.g., hillslope-channel connection), active bank erosion, and downstream gravel conveyance potential (e.g., obstruction by beaver dams) are assessed from aerial photography. Stream type classification does not improve qB,bf estimates.