Florence Rabier (the Director of the European Centre for Medium Range Weather Forecasting)
Title: How AI is revolutionising weather prediction
Abstract: This seminar will tell the story of the European Centre for Medium Range Weather Forecasts (ECMWF) based in Reading, from its beginning in 1975 to the current days, from working on the first CRAY-1 supercomputer in Europe in the late 1970s, to the implementation of the first fully operational AI weather model.
Weather prediction is complex, with ever expanding data from satellites and surface sensors and sophisticated computer models based on the laws of physics. The last few decades have seen significant improvements in the ability to make accurate predictions, with the introduction of optimal control to provide the initial conditions of the model and ensemble forecasts to describe a range of future scenarios. However AI has the potential to drastically improve weather forecasting, making it even more accurate and faster.
While traditional models can provide accurate forecasts, they rely on immense computational power which limits their use. AI’s ability to predict the future based on vast amounts of data at unprecedented speed is transforming the sector.
This seminar will describe the way traditional physics-based weather prediction works and how it can be combined with AI to move to next-generation forecasting.
ECMWF, an intergovernmental organisation supported by 35 countries in the European area, works collaboratively with National Meteorological Agencies in its Member States, providing cutting edge monitoring of the Earth system and weather forecasts from days to months ahead to its users. ECMWF also works closely with the European Union and operates the Copernicus Atmosphere Monitoring Service and the Copernicus Climate Change Service, providing key indicators of climate change to policy makers and the media.