HR: 16:45h
AN: H24B-04 INVITED     [Abstracts]
TI: A Review of Sampling Issues Relative to the Determination of Suspended Sediment Fluxes in Fluvial Systems Using Rating Curves
AU: * Horowitz, A J
EM: horowitz@usgs.gov
AB: For over 60 years, in the absence of sufficient (e.g., daily values) suspended sediment concentration (SSC) measurements, hydrologists have used site-specific sediment-rating curves to estimate SSC's for subsequent flux calculations. In relatively large- (>1,000,000 km²) and medium-sized (<1,000,000 to >1,000 km²) fluvial systems, this approach is capable of generating accurate suspended sediment flux estimates within ±15 to 20%, provided the requisite temporal resolution is limited to quarterly, or longer. However, some level of sampling is required to develop and calibrate the rating curves. In turn, adequate sampling programs for subsequent flux calculations require a basic understanding of the spatial and temporal variability of SSC. Suspended sediment is not evenly distributed in fluvial cross-sections. Typically, this accrues when sand-sized (>63-μm) particles are present. Their concentrations increase from the riverbanks toward the centroid of flow, as well as from the surface to the riverbed. As a result, depth- and width-integrated sampling is required to obtain representative SSC measurements. In turn, SSC's generated from surrogates (e.g., turbidity), or from automatic point sampling equipment, must be adjusted based on the differences in concentration between concurrent measurements/samples, otherwise they will not be cross-sectionally representative. SSC's also display marked short- and long-term temporal variability. The examination of SSC data from long-term, daily monitoring sites, has provided useful insights into the appropriate sampling strategies for dealing with temporal variability. Short-term temporal variability in SSC is not a major sampling consideration for rating-curve generated flux estimates with temporal resolutions of quarterly or longer. On the other hand, long-term variability can be addressed by hydrologically-based rather than calendar-based sampling. If some 80% of the range of annual flow-conditions can be covered, then reasonably accurate (±15 to 20%) annual fluxes can be produced with as few as 12 samples per year, whereas similarly accurate 5-year estimates can be achieved with as few as 6 samples per year. However, experience also has indicated that the long-term relation between discharge and SSC is not constant; hence, a site-specific rating curve must be readjusted, through the continued collection of SSC data, to maintain predictive accuracy.
DE: 1862 Sediment transport (4558)
DE: 1871 Surface water quality
SC: Hydrology [H]
MN: Fall Meeting 2005