Predicting microbiological contamination in urban rivers: an event-based indicator approach applied to the Seine river during the 2024 Paris olympic and paralympic games

Urban rivers are subject to microbiological contamination following the discharge of waste water after heavy rains. These peaks of faecal contamination lead to health risks, particularly in urban bathing areas. To prevent these risks, European legislation requires early warning systems to be operated. This study proposes a new methodology for implementing an early warning system based on high-frequency measurements of bacteriological contamination proxies and hydro-meteorological data. By analysing these high-frequency data collected over three summer seasons (2021-2023) in the Seine River, Paris, France, the proposed framework defines rainfall events and their impact on microbial water quality. Dimensionality reduction techniques, including Principal Component Analysis (PCA) and Manifold ISOMAP, are applied and compared to identify key explanatory characteristics and build a predictive model. A global rain parameter (GRP) is derived, which effectively classifies rainfall events based on their potential impact on water quality. The method was successfully tested during the 2024 Paris Olympic and Paralympic Games, demonstrating its potential for real-time water quality management.