Rill network development and sediment budgets

Threshold conditions for rill incision are well known, but few studies have examined the effect of network geometry on water and sediment fluxes withi n an evolving rill system. This paper reports the first in a series of stud ies designed to identify the influence of soil properties on rill network a nd confluence geometry, and on water and sediment fluxes in rill systems. T he object of this study was to prepare detailed rill network water and sedi ment budgets identifying important sources and sinks. Simulated rainfall ex periments were carried out on a Canadian silt loam soil in a 7.1 m x 2.4 m flume on a 5 degrees slope. Rill networks of varying complexity developed, which were ultimately constrained by flume boundaries. Sediment and water f luxes and hydraulic conditions were measured within networks and at a termi nal weir. Networks evolved by initial knickpoint incision in the lower flume, as flow shear velocities reached critical levels of 4.5-5 cm s(-1) followed by hea dward migration and tributary development. Microtopography determined tribu tary location, but the timing and intensity of development were controlled by the incision and migration rate of the main channel, which changed local thalwegs and base levels, raising shear Velocities on side slopes above cr itical levels. Sediment discharge at the weir broadly reflected rill incisi on intensity and transport-limited conditions, but as active incision moved headward, the linkage became attenuated and identification of discrete ero sional incidents (e.g. local bank collapse) in the signal of the weir recor d became very difficult. Detailed water and sediment budgets showed much more complex patterns of lo calized incision and deposition within networks, strongly influenced by loc al changes in thalweg and in water discharge due to seepage or return flow, and by the effect of confluences on hydraulic conditions. Results indicate the value of detailed sediment budgets in interpreting weir water and sedi ment flux records, and the necessity of linking such measurements for erosi on plots and hillslope segments to rill network characterization. Experimen tal results are consistent with a simple model of rill system evolution bas ed on channel incision, headward migration, and the critical shear velociti es for rill initiation. Copyright (C) 2000 John Wiley & Sons, Ltd.