Quantifying uncertainty in morphologically-derived bedload transport rates for large braided rivers: insights from high-resolution, high-frequency digital elevation model differencing
Abstract
Repeat surveys of channel morphology provide a means to quantify fluvial sediment storage and enable inferences about changes in long-term sediment supply, watershed delivery and bed level adjustment; information vital to support effective river and land management. Over shorter time-scales, direct differencing of fluvial terrain models may also offer a route to predict reach-averaged sediment transport rates and quantify the patterns of channel morphodynamics and the processes that force them. Recent and rapid advances in geomatics have facilitated these goals by enabling the acquisition of topographic data at spatial resolutions and precisions suitable for characterising river morphology at the scale of individual grains over multi-kilometre reaches. Despite improvements in topographic surveying, inverting the terms of the sediment budget to derive estimates of sediment transport and link these to morphodynamic processes is, nonetheless, often confounded by limited knowledge of either the sediment supply or efflux across a boundary of the control volume, or unobserved cut-and-fill taking place between surveys. This latter problem is particularly poorly constrained, as field logistics frequently preclude surveys at a temporal frequency sufficient to capture changes in sediment storage associated with each competent event, let alone changes during individual floods. In this paper, we attempt to quantify the principal sources of uncertainty in morphologically-derived bedload transport rates for the large, labile, gravel-bed braided Rees River which drains the Southern Alps of NZ. During the austral summer of 2009-10, a unique timeseries of 10 high quality DEMs was derived for a 3 x 0.7 km reach of the Rees, using a combination of mobile terrestrial laser scanning, aDcp soundings and aerial image analysis. Complementary measurements of the forcing flood discharges and estimates of event-based particle step lengths were also acquired during the field campaign. Together, the resulting dataset quantifies the evolution of the study reach over an annual flood season and provides an unprecedented insight into the patterns and processes of braiding. Uncertainties in the inferred rates of bedload transport are associated with the temporal and spatial frequency of measurements used to estimate the storage term of the sediment budget, and methods used to derive the boundary sediment flux. Results obtained reveal that over the annual flood season, over 80% of the braidplain was mobilised and that more than 50% of the bed experienced multiple cycles of cut and fill. Integration of cut and fill volumes event-by-event were found to be approximately 300% of the net change between October and May. While significant uncertainties reside in estimates of the boundary flux, rates of bedload transport derived for individual events are shown to correlate well with total energy expenditure and suggest that a relatively simple relationship may exist between the driving hydraulic forces at the reach scale and the geomorphic work performed.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFMEP33E..03B
- Keywords:
-
- 1825 HYDROLOGY Geomorphology: fluvial;
- 1862 HYDROLOGY Sediment transport;
- 1855 HYDROLOGY Remote sensing;
- 1873 HYDROLOGY Uncertainty assessment