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

Paper No. 3
Presentation Time: 2:00 PM


AALTO, Rolf, Earth and Space Sciences, Univ of Washington, Seattle, WA 98195-1310 and NITTROUER, Charles A., School of Oceanography, Univ of Washington, Seattle, WA 98195-7940, aalto@u.washington.edu

Lowland floodplains store suspended sediment carried over bank during floods. Because floodplain sedimentation may account for the majority of the net sediment deposition within large sand-bedded river-floodplain systems, it is important to develop a means to quantify this process. Recent studies have demonstrated that the natural radionuclide Pb-210 can be used to determine floodplain accumulation rates over decadal-to-century timescales with a Constant Input Concentration and Constant Sedimentation rate model (CICCS). To investigate the application of the CICCS method to a large, dynamic, pristine river system with extensive forested floodplains, 265 vertical cores were collected throughout the floodplains along ~2,000 km of the Beni and Mamore Rivers in northern Bolivia. Rather than measuring bulk-averaged down-core Pb-210 activity to estimate average accumulation rate according to the standard CICCS procedure, the cores were instead sectioned vertically to obtain a high-resolution record of Pb-210 activity and sediment accumulation. Over 95% of these profiles depict a history of episodic deposition with decadal recurrence intervals, and few profiles exhibit the steady-state accumulation assumed by the CICCS model. Because these discrete deposition events are large, few of the cores ever reach supported background Pb-210 activity at depth, another requirement of CICCS. Furthermore, the input activity of Pb-210 is not constant, both at a site and down-channel. The Pb-210 content of sediment varies directly with fine sediment content, which is highly variable within many of the cores. Without a discretely sampled, clay-normalized Pb-210 activity profile, it is difficult to evaluate the suitability of CICCS for each core. To accommodate these complexities, a more labor intensive approach is proposed for dating sedimentation using discrete down-core, clay-normalized measurements of Pb-210 activity, coupled with a geomorphic model of Pb-210 input during large floods. Combined with a field determination of supported Pb-210 activity, collection of multiple terrace cores to determine the meteoric input of Pb-210, and a technique for independently dating the “meteoric cap” above discrete depositional events, this approach to Pb-210 geochronology provides a convenient and robust approach to dating floodplain sediment.