A probabilistic-deterministic approach for assessing climate change effects on infection risks downstream of sewage emissions from CSOs.

The discharge of pathogens into urban recreational water bodies during combined sewer overflows (CSOs) pose a potential threat for public health which may increase in the future due to climate change. Improved methods are needed for predicting the impact of these effects on the microbiological urban river water quality and infection risks during recreational use. The aim of this study was to develop a novel probabilistic-deterministic modelling approach for this purpose building on physically plausible generated future rainfall time series. The approach consists of disaggregation and validation of daily precipitation time series from 21 regional climate models for a reference period (1971-2000, C20), a near-term future period (2021-2050, NTF) and a long-term future period (2071-2100, LTF) into sub-daily scale, and predicting the concentrations of enterococci and Giardia and Cryptosporidium, and infection risks during recreational use in the river downstream of the sewage emissions from CSOs. The approach was tested for an urban river catchment in Austria which is used for recreational activities (i.e. swimming, playing, wading, hand-to-mouth contact). According to a worst-case scenario (i.e. children bathing in the river), the 95th percentile infection risks for Giardia and Cryptosporidium range from 0.08 % in winter to 8 % per person and exposure event in summer for C20. The infection risk increase in the future is up to 0.8 log10 for individual scenarios. The results imply that measures to prevent CSOs may be needed to ensure sustainable water safety. The approach is promising for predicting the effect of climate change on urban water safety requirements and for supporting the selection of sustainable mitigation measures. Future studies should focus on reducing the uncertainty of the predictions at local scale.