Geo-referenced simulations of E. coli in a sub-catchment of the Vecht River using a probabilistic approach

The proportion of wild swimmers at non-official bathing sites has increased during the Covid-19 pandemic. Bathing water quality at designated sites is monitored through analysis of the concentration of fecal indicator bacteria such as E. coli. However, non-official sites are generally not monitored. In a previous work, steady state modelling of E.coli was achieved at catchment scale, enabling a comparison of expected concentrations along an entire catchment for longtime average. However, E. coli concentrations can vary over several orders of magnitude at the same monitoring site throughout the year. To capture the temporal variability of E. coli concentrations on the catchment scale, we extended the existing deterministic E. coli sub-module of the GREAT-ER (Geo-referenced Exposure Assessment tool for European Rivers) model for probabilistic Monte-Carlo simulations. Here, selected model parameters are represented by probability distributions instead of fixed values. Wastewater treatment plant (WWTP) emissions and diffuse emissions were parameterized using selected data from a previous monitoring campaign (calibration data set) and instream processes were modeled using literature data. Comparison of simulation results with monitoring data (evaluation data set) indicates that predicted E. coli concentrations well-represent median measured concentrations, although the range of predicted concentrations is slightly larger than the observed concentration variability. The parameters with the largest influence on the range of predicted concentrations are flow rate and E. coli removal efficiency in WWTPs. A comparison of predicted 90th percentiles with the threshold for sufficient bathing water quality (according to the EU Bathing Water Directive) indicates that year-round swimming at sites influenced by WWTP effluents is advisable almost nowhere in the study area. A refinement of the model can be achieved if quantitative relationships between the WWTP removal efficiency and both, the treatment technologies as well as the operating parameters are further established.