Cosmogenic 10Be as a tracer for hillslope and channel sediment dynamics in the Clearwater River, western Washington State

We use the Be-10 concentration in alluvial sand and gravel to document hillslope exposure history and trace the reduction in grain size of river alluvium in the Clearwater River basin, western Washington State. Clearwater tributary basins Wilson Creek (WC, 6 km(2)) and East Fork Miller Creek (EFMC, 12km(2)) were each sampled in headwater and downstream sites for Be-10 in alluvial sand (0.25-0.50 mm) and gravel (22.6-90 turn). Grain size distributions were determined at each sampling site as well as at several locations along the main stem of the Clearwater River. Channel and watershed geomorphology were quantified using long profile modeling and related morphometric analyses. We found that the Be-10 concentrations differ significantly between the two grain size fractions at all four sampling locations. At both headwater sites the gravel exhibits 25% lower Be-10 concentrations compared to the sand. Similarly, the downstream site of Lower WC exhibits 55% lower Be-10 concentration in the gravel compared to the sand. In contrast, the gravel from the downstream site on Lower EFMC exhibited 22% higher Be-10 concentration in the gravel compared to the sand. The disparity in Be-10 concentration at the WC sites is best explained by shielding of the coarser grain size fraction and its delivery to the channel by deep-seated landslide processes. More intense landsliding in the downstream WC site is consistent with the increased disparity of Be-10 between the sand and gravel fractions at that site. The inverse relationship between the sand and gravel Be-10 concentrations at the headwater EFMC site is best explained by a sediment provenance mechanism where the hillslope weathering rate exceeds the down-slope transport rate for this particular basin. The inverse grain size dependency (gravel>sand) observed in the downstream EFMC site requires a more complex interplay between hillslope and channel processes including cobble weathering and grain size reduction during fluvial transport, resulting in a dilution of the Be-10 signal in the sand fraction downstream. These results underscore the importance of geomorphic consideration and the grain size sampled in the correct interpretation of basin-average erosion rates. (c) 2007 Elsevier B.V. All rights reserved.