INTERACTIONS BETWEEN STREAM CHANNEL SLOPE AND PARTICLE ROUGHNESS IN A CELLULAR AUTOMATA

Slope may have an important effect on both particle and bedform roughness in steep, gravel-to-boulder-bed streams. Published analytical, field and flume studies indicate that particle roughness may depend on partial mobility effects; may adjust with slope; may modify or suppress bedform development; and that coincident changes in slope and bedform type, from step-pool to pool-riffle, may occur. Cellular automata (CA) are models defined on a grid of cells, with changes in the cell state depending solely on the current state of nearby cells. Previous CA models, although not in the context of stream channels, have been used to investigate roughening of surfaces formed by particle deposition and erosion. Results from those models suggest that particulate surfaces evolve to a maximum roughness, as hypothesized for mountain river channels, and that the surface growth may depend on slope. In the work presented here, cellular automata are used to investigate how slope and particle mobility affect particle size and topographic form in steep stream channels. The rate of surface roughening, and its slope dependence, are also evaluated. Roughness in the model is described by local height differences, by textural patches and their particle size, and by patterns in surface height evolved from an initial random spatial distribution. In one model variant, the surface evolves characteristic particle associations over time, as a result of size selection for transport that is influenced by the size of nearby particles. These model-generated features may be analogues of textural patches in stream channels. Preliminary results of the model variants, similar to results of previous CA surface models with similar controls, indicate that the surface evolves to a maximum roughness on average, but may have wide variations about the average. This CA surface modeling approach is one way of investigating how slope and particle size effects can modify streambed roughness. Such modeling can lead to a better understanding of particle-particle interactions controlled by local patterns in slope and particle size, and by partial mobility of sediment. Understanding interactions of slope and particle patterns can clarify how the channel boundary responds to change, and how it influences formation of pool sequences.