2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 14
Presentation Time: 8:00 AM-12:00 PM

SEDIMENT STORAGE IN A HEADWATER TRIBUTARY OF THE OREGON COAST RANGE: A METHOD FOR DETERMINING RESIDENCE TIME


CASEBEER, Nathan E., Geosciences, Oregon State Univ, 104 Wilkinson Hall, Corvallis, OR 97331-5506 and LANCASTER, Stephen T., Dept. Geosciences, Oregon State Univ, Corvallis, OR 97331, casebeen@geo.orst.edu

This study is part of the effort to quantify sediment budgets and understand the geomorphic evolution of steep mountains where debris flows are the dominant agent of upland erosion. Observations indicate that headwater tributary basins in the Oregon Coast Range (OCR) can store a relatively large amount of sediment, mostly from debris flows, in unusually wide valley floors. Bear Creek, a 2.2 km2 tributary basin of Knowles Creek, has a surveyed valley floor sediment volume of 69,466 m3. An average residence time for this valley floor sediment of 320 yrs assumes a long-term average denudation rate of 0.1 mm/yr, as found in other studies in the OCR. Basins of similar size in Hoffman Creek and Cedar Creek have average sediment residence times of 460 yrs and 360 yrs respectively. We hypothesize that by storing sediment, headwater tributary valley floors in the OCR act as a buffer between the episodic debris flow signal and larger basin reaches. In this study, we attempt to determine the residence time distribution of sediment in a headwater basin by radiocarbon dating. We assume that storage age distribution approximates a residence time distribution. Mapped channel bank exposures and valley storage serve as a grid for generating random coordinates for sampling of carbon dates. These random coordinates are weighted to sample wider valley floor reaches more frequently than narrower reaches. Points are collapsed to randomly chosen right or left stream bank exposures. The stream is assumed to be a random cut through the valley fill. The utility of this method is to limit random sampling to known exposures, minimize the number of carbon samples to date (thus reducing the cost), and provide a relatively unbiased distribution of ages that defines the residence time of most of the sediment in the basin weighted to volume. Two preliminary radiocarbon dates (not chosen by this method) suggest a wide distribution in Bear Creek. Stream banks at distances of 2,391 m and 1,545 m from the basin divide yielded 2-sigma calendar calibrated age ranges of 5,220-4,820 BP and 3,840-3,640 BP respectively.