Evaluation of suspended load changes using AnnAGNPS and SWAT semi-empirical erosion models

Simulation models are frequently used tools for suspended sediment load quantification in mesoscale catchment areas. To improve the simulation results, it is important to compare and verify different models using measured data. In this study, we tested two continuous simulation, semi-distributed erosion and sediment transport models AnnAGNPS and SWAT for the period of 1995–2004 in the Blšanka river basin. The catchment (374 km2) is an agricultural hop-growing region in the north-western part of the Czech Republic. Both models were calibrated using the first 5 years of data, and then validated using the second 5 years of daily discharge and suspended sediment yield values. For the long-term continuous simulation, the results of SWAT model simulation showed a better agreement to the measured data, while for short-term rainfall-runoff event simulation, especially short duration intensive rainfall events, suspended load was more accurately predicted by AnnAGNPS. Land use changed significantly in parts of the river basin during the observation period. Many hop-gardens were dismantled and arable land was partially converted to grassland and pasture. This enabled us to test the applicability of AnnAGNPS and SWAT models under changing land use conditions. According to both models, suspended sediment load was reduced after the land use changes by 10%–30%, which was in agreement with decreasing sediment discharge observed at the watershed outlet.