SCALING ANALYSIS OF EXPERIMENTAL CHANNEL MIGRATION RATES IN RESPONSE TO VARYING SEDIMENT FLUX, BASELEVEL, AND VEGETATION DENSITY

Scale experiments were analyzed to investigate the changes in river channel migration rates under three forcings: changes in vegetation density, changes in baselevel, and changes in sediment flux. Migration rates were analyzed through image analysis. Time-lapse images of these experiments were thresholded and separated into “wet” and “dry” pixels. By comparing the locations of the “wet” and “dry” pixels between a reference timestep and successive timesteps, it is possible to show changes in channel positions over time. The correlation between the original pattern of wet and dry pixels and the pattern of these pixels at a subsequent time declines logarithmically with increasing time from the original reference state. The slopes of these curves and the times at which they reach maximum decorrelation provide quantitative measures with which to compare decorrelation rates under varying systemic conditions. The vegetated river system decorrelated at between 1/10 and 1/20 the rate of the unvegetated system. No change in channel migration rates were found during baselevel changes that were slow with respect to channel migration rate. However, the system was found to be more channel-dominated during baselevel fall and more sheet-flow dominated during baselevel rise. Though the analysis is not yet complete, sediment flux is believed to cause an increase in channel migration rate. Further analytical work is planned in order to increase the model precision from one-dimensional correlation to two-dimensional deformation of the river plan form. In addition, comparing the experimental scaling parameters to field data would be invaluable for applying these migration timescales to real landscapes.