Time required to obtain representative point-integrated suspended sediment samples

Suspended sediment sampling by conventional bottle sampling methods remains one of the primary ways in which sediment input is measured for sediment budgets. Depth- and point-integrated methods are commonly assumed to accurately represent suspended sediment transport and to contain minimal error, however a recent study of depth-integrated sampling has identified considerable uncertainty between grain-size classes of multiple samples. The uncertainty surrounding point-integrated sampling methods is less well understood. This field study examines at-a-point variability in suspended sediment concentration, grain-size distribution, and grain-size moments to determine if traditional point-integrated methods accurately provide a representative sample of suspended sediment. Here, we present observations of suspended sediment from the sand bedded portion of the Fraser River at Mission, British Columbia, Canada, using a LISST laser-diffraction instrument. Measurements were obtained during the 2010 and 2013 freshet to address scales of sediment transport variability related to turbulence and hydraulic phenomena, respectively. Our observations indicate that the minimum sampling time to obtain representative indicators of suspended sediment transport ranges between 100 to 200 seconds. We are exploring the scales of variability embedded in the flow and the uncertainty in point-integrated sampling. Our goal is to understand the magnitude and the sources of error in suspended sediment measurements, to better understand the sources of uncertainty in sediment budgets.