The Paris 2024 Olympics Research Demonstration Project

Abstract The Paris 2024 Olympics Research Demonstration Project lasted 5 years and was endorsed by the World Weather Research Programme within the World Meteorological Organization (WMO) to improve urban weather forecasts, using the metropolitan area of Paris as a case study. Meteorological institutes and universities from 10 countries participated. The project had three objectives to increase knowledge of urban summer meteorological hazards, to improve hectometric-scale numerical weather prediction models in cities, and to facilitate the coproduction of weather information for large sporting events. Increased convective activity downwind of Paris was highlighted by a new radar- and lightning-based climatology, and numerical experiments identified its potential driving processes. Collaborative analyses of past heat waves improved air quality models for the Olympics and simulations of thermal comfort variability. Advances in urban-scale modeling enabled real-time intercomparison of seven hectometric- or kilometric-scale atmospheric models that provided daily forecasts throughout the Olympics and Paralympics. Hectometric models showed similarities in convective precipitation characteristics, tending to have an excessive number of small showers and grid-length dependence. Model intercomparisons also showed variability in urban heat island intensity and unexpected differences in urban heat plume extent. A sociological study highlighted the differences in viewpoints between forecasters and sport managers and the importance of transparency in communicating uncertainties. Finally, a decision-making procedure to manage extreme heat contingencies for the “Marathon for All” Olympics public event was developed based on a 100-m grid-length model in collaboration with the weather forecasters’ team. All these insights open new scientific questions in urban climate research and ways forward for future hectometric numerical weather prediction. Significance Statement Forecasting in an urban environment during a global event held in one of the world’s most densely populated cities is a unique meteorological challenge. The real-time intercomparison of high-resolution models conducted during the Olympics, as presented here, demonstrates the potential of hectometric forecasting while also highlighting new scientific questions and operational challenges, such as those associated with event-specific forecasts like the Marathon for All. The findings will appeal to both the numerical weather prediction and general meteorological communities interested in urban meteorology advances and real-time forecasting improvements for major events.