Author(s): Thomas Hoffmann; Gudrun Hillebrand; Jurgen Schmegg; Stefan Vollmer
Linked Author(s): Gudrun Hillebrand
Keywords: Suspended sediment; Turbidity; Calibration; Floods; Sediment monitoring in Germany; River Moselle; River Rhine
Abstract: Traditionally, suspended sediment in German inland waterways has been monitored using daily water samples taken manually at~70 sampling locations starting in 1965. A major limitation of the manual sediment monitoring is related to the frequency of sediment sampling, which is limited to one sample per day and restricted to the working hours of the Federal Waterways and Shipping Administration. Thus short-term extreme events, which may transport significant fractions of the annual load, may be strongly underrepresented. Here we present results from two stations measuring turbidity at a 15 min interval at the River Moselle and the River Rhine. The turbidity measurements are calibrated using filtered water/sediment samples that were taken at irregular intervals and were conducted simultaneously with the traditional, manual sampling design for 14 to 36 months. A major aim of this study is to test various calibration methods and to compare annual loads derived using the daily water sampling and the 15 min automated turbidity measurements. The stratified linear regression of the turbidity (TR) and the suspended sediment concentration (SSC), based on the filtered water samples, resulted in the lowest root mean square error (~20%) of the tested calibration methods. The stratification of the calibration samples takes into account thatSSC/TR-ratios change in response to changing dominant grain sizes during various discharge stages. The comparison shows that automated high-frequency measurements result in~25% higher annual sediment loads compared to daily sediment sampling.
Year: 2017