Search

Identifying the erosion processes contributing to increased basin fine sediment yield is important for reducing downstream impacts on aquatic ecosystems. However, erosion rates are spatially variable, and much eroded sediment is stored within river basins and not delivered downstream. A spatially distributed sediment budget model is described that assesses the primary sources (hillslope soil erosion, gully and riverbank erosion) and sinks (floodplain and reservoir deposition) of fine sediment for each link in a river network. The model performance is evaluated in a 17,000-km2 basin in south-east Australia using measured suspended sediment yields from eight catchments within the basin, each 100–700km2 in area. Spatial variations within the basin in yield and area-specific yield were reliably predicted. Observed yields and area-specific yields varied by 17-fold and 15-fold respectively between the catchments, while predictions were generally within a factor of 2 of observations. Model efficiency at predicting variations in area-specific yield was good outside forested areas (0.58), and performance was weakly sensitive to parameter values. Yields from forested areas were under-predicted, and reducing the predicted influence of riparian vegetation on bank erosion improved model performance in those areas. The model provided more accurate and higher resolution predictions than catchment area interpolation of measured yields from neighbouring river basins. The model is suitable for guiding the targeting of remediation measures within river basins to reduce downstream sediment yields. [Copyright &y& Elsevier]

The prediction and control of river sediment yield (SY) are critical but challenging tasks. Erosion and sediment transfer in river catchments are controlled by different processes, whose relative importance varies in space and time. We thus put forward that SY can be estimated more efficiently by using explicitly the information contained in the similarity within groups. To test this hypothesis, we developed a novel Bayesian hierarchical model, applied it to a sample of heterogeneous river catchments and compared its fixed-effects and mixed-effects performance incorporating different group levels, namely gauges, rivers, basins and clusters of catchments. With a parsimonious linear model consisting of four variables (specific and extreme discharge, elevation and retention coefficient), we achieved good performance criteria in the calibration (NSE: 0.79–0.85) and in the cross-validation for temporal and spatial prediction (NSE: 0.71 and 0.72, respectively). These results support the promising potential of this technique. • We develop a Bayesian hierarchical framework to model river sediment yield (BaHSYM). • For spatial predictions, we combine BaHSYM with clustering of catchments. • Robust predictions can be achieved with a parsimonious linear model. • The mixed-effects model performs largely better than ordinary linear regression. [ABSTRACT FROM AUTHOR]