Multiscale Grain-Size Mapping Along the Upper Sandy River

Abstract

Grain-size remote sensing presents an opportunity to address our poor understanding of the relationships among the patterns and processes of sediment sorting across spatial scales and advance the accessibility to riverscape mapping approaches. This thesis confronts the barriers to applying image-based grain-size mapping techniques by presenting a methodological framework that addresses the principal components of photosieving and their implications for the resulting grain-size dataset. This framework is leveraged to develop a protocol for mapping the spatial variability in coarse sediment within gravel bars and throughout a 12 km segment of the upper Sandy River, Oregon, USA. I analyze the grain-size variability at the river segment and bar scales in relation to active channel width. The results illustrate that bar scale variability in some cases is nearly equivalent to the grain-size variability observed within the 12 km reach. Additionally, the results show that grain-size is inversely correlated with channel width, but that this relationship is strongest for patches located at bar heads. These finding indicate that systematic patterns are modulated by local processes such as hydraulics or sediment supply, and that the application of downstream fining models may have limited application on the Upper Sandy River at scales finer than a few kilometers.