Model-based prediction of bathing water quality in a lake polluted by fecal coliform bacteria from combined sewer overflows

In the European Union, recreational water quality is regulated by the bathing water directive, which requires authorities to regularly take water samples to identify fecal bacterial pollution, which can compromise the bathing water quality at designated recreational bathing water areas. Using a case study from Lake Knudsøin Denmark, this paper shows that the bathing water quality occasionally is compromised by overflows from combined sewer systems. Due to the randomness in the frequency of overflow occurrences, the entailing decrease in bathing water quality is not normally detected by the regulatory sampling campaigns. By dedicated sampling campaigns conducted in this project in the proximity of outlets and recreational bathing areas and by hydrodynamical transport modeling, it is shown that the transport patterns of pollutants are crucially and rapidly dependent on dynamic conditions. The occasional short-term pollution by fecal bacteria and consequent decrease in bathing water quality are, therefore, challenging to capture (both spatially and temporally) by regular water sampling. Rather than increasing the frequency of water sampling, an online model-based warning system for fecal bacteria contamination in bathing areas is proposed. This warning system framework includes: (1) dynamical modeling of combined sewer overflow based on rainfall over a catchment area, (2) a hydrodynamical model that simulates current fields in multiple vertical layers based on wind forcing and water fluxes, and (3) a particle dispersion model which provides an estimate of pollutant concentrations. The output from the model has shown potential to issue bathing prohibition if there is a risk of fecal bacteria concentrations below the criteria for good bathing water quality.