Prediction of sediment transport in step-pool channels

In mountainous drainage networks, sediment mobilized on hillslopes must first pass through steep streams before reaching lower-gradient channels. The bed of steep channels is typically composed of large, relatively immobile boulders and finer, more mobile gravel. Most sediment transport equations overpredict sediment flux in steep streams by several orders of magnitude because they do not account for the stress borne by immobile grains and the limited availability of the more mobile sediment. We previously developed and tested (in flume experiments) a sediment transport equation that accounts for these two effects. Here we modify the Parker (1990) bed load equation to include the resistance borne by steps and selective transport of the relatively mobile sediment using a range of hiding functions. We test a number of resistance equations and hiding functions, combined with our modified and the original Parker equations, against measured flow and sediment transport in three steep channels. Our modified sediment transport equation generally predicts the transported sediment volumes to within an order of magnitude of the measured values, whereas the unmodified equations do not. The most accurate sediment flux predictions were obtained from using our modified equation, combined with a hiding function that calculates highly selective transport of the relatively mobile sediment. Furthermore, this hiding function has a critical Shields stress that is similar to those reported for lower gradient channels. The effects of the immobile steps on flow and sediment transport are not adequately captured by simply increasing the critical Shields stress to values reported in steep streams.