Variation in turbidity with precipitation and flow in a regulated river system – River G ta lv, SW Sweden

The turbidity variation in time and space is investigated in the downstream stretch of the river G ta lv in Sweden. The river is heavily regulated and carries the discharge from the largest fresh water lake in Sweden, lake V nern, to the outflow point in G teborg Harbour on the Swedish west coast. The river is an important waterway and serves as fresh-water supply for 700 000 users. Turbidity is utilised as an indicator to ensure sufficient quality of the intake water to the treatment plant. The overall objective of the study was to investigate the influence of rainfall, surface runoff, and river water flow on the temporal and spatial variability of the turbidity in the regulated river system by employing statistical analysis of an extensive data set. Six-year long time series of daily mean values on precipitation, discharge, and turbidity from six stations along the river were examined primarily through linear correlation and regression analysis, combined with nonparametric tests and analysis of variance. The analyses were performed on annual, monthly, and daily basis, establishing temporal patterns and dependences, including seasonal changes, impacts from extreme events, influences from tributaries, and the spatial variation along the river. The results showed that there is no simple relationship between discharge, precipitation, and turbidity, mainly due to the complexity of the runoff process, the regulation of the river, and the effects of lake V nern and its large catchment area. For the river G ta lv, significant, positive correlations between turbidity, discharge, and precipitation could only be found during periods with high flow combined with heavy rainfall. Local precipitation does not seem to have any significant impact on the discharge in the main river, which is primarily governed by the precipitation at catchment scale. The discharge from the lake V nern determines the base level for the turbidity in the river, whereas local surface runoff and tributary discharge induced by rainfall govern the temporal variability in turbidity. Autocorrelation analysis indicates a temporal persistence in turbidity of about 10 days. The results also show that erosion in the main river, from the river bed and banks, is not a dominant contributor to the suspended sediment transport in the river. Further studies on the correlation between turbidity and suspended sediment transport and in relation to erosion processes are suggested.