Soil loss estimation and empirical relationships for sediment delivery ratios of European river catchments

ABSTRACT

Sediment delivery ratios (SDRs) link soil loss to sediment yields (SYs) of river catchments. At large scales, SDR models rely on simple catchment properties. To assess their sensitivity and uncertainty in European regions, I compiled a sediment database and derived 16 soil loss maps varying the approximation of the factors of the universal soil loss equation (USLE). Additionally to parameterizing the USLE, the sensitivity analysis comprised the choice of the soil loss and SDR models. The USLE maps were compared to a Pan-European Soil Erosion Risk Assessment (PESERA) map and two empirical SDR models were applied for explaining the variability of SDR and SY. Most relevant for the model efficiency was the choice of the soil loss model. Unlike the PESERA map, the USLE with the four-parameter SDR model allowed satisfactory results in all regions. Fewer parameters for estimating SDR were less appropriate. Among the USLE factors, the choice of the K factor was most important. The overall model sensitivity to choosing the USLE map and the SDR model was similar. Based on the sensitivity analysis, a most suitable USLE map for predicting SY of European river catchments is proposed. The uncertainty in USLE estimates resulted in an uncertainty of modelled SY varying from 30% to 60% in different regions. No model realization grasped the variability of SDR and SY, so only regional SDR models were feasible. The model application and evaluation were hampered by unrepresentative sediment data and the inherent limitations of the modelling framework.

Keywords:

• PESERA
• sediment delivery ratio
• sediment yield
• sensitivity analysis
• uncertainty analysis
• universal soil loss equation

Acknowledgements

The Bavarian Environment Agency (namely M. Knott), M. Zessner, and A. Kovács (Vienna University of Technology) are thanked for kindly providing sediment data. The help of C. Bosco (Joint Research Center, Ispra), Tz. Karagiozova, and E. Artinian (National Institute for Meteorology and Hydrology, Bulgaria) is much appreciated.