Mapping spatial patterns of stream power and channel change along a gravel-bed river in northern Yellowstone

Abstract

Stream power represents the rate of energy expenditure along a river and can be calculated using topographic data acquired via remote sensing or field surveys. This study sought to quantitatively relate temporal changes in the form of Soda Butte Creek, a gravel-bed river in northeastern Yellowstone National Park, to stream power gradients along an 8-km reach. Aerial photographs from 1994 to 2012 and ground-based surveys were used to develop a locational probability map and morphologic sediment budget to assess lateral channel mobility and changes in net sediment flux. A drainage area-to-discharge relationship and DEM developed from LiDAR data were used to obtain the discharge and slope values needed to calculate stream power. Local and lagged relationships between mean stream power gradient at median peak discharge and volumes of erosion, deposition, and net sediment flux were quantified via spatial cross-correlation analyses. Similarly, autocorrelations of locational probabilities and sediment fluxes were used to examine spatial patterns of sediment sources and sinks. Energy expended above critical stream power was calculated for each time period to relate the magnitude and duration of peak flows to the total volumetric change in each time increment. Collectively, we refer to these methods as the stream power gradient (SPG) framework. The results of this study were compromised by methodological limitations of the SPG framework and revealed some complications likely to arise when applying this framework to small, wandering, gravel-bed rivers. Correlations between stream power gradients and sediment flux were generally weak, highlighting the inability of relatively simple statistical approaches to link sub-budget cell-scale sediment dynamics to larger-scale driving forces such as stream power gradients. Improving the moderate spatial resolution techniques used in this study and acquiring very-high resolution data from recently developed methods in fluvial remote sensing could help improve understanding of the spatial organization of stream power, sediment transport, and channel change in dynamic natural rivers.