Science innovation

One focus for innovation in the coming years is artificial intelligence, which is driving change in many fields such as remote sensing and operational weather forecasting.

“What is happening today in terms of artificial intelligence and machine learning is not only changing today’s technology and the way we work with new methods and new tools, but it will also change the way meteorological services will operate and how we plan and design future satellites,” says EUMETSAT’s Chief Scientist Paolo Ruti.

Machine learning, which is a subset of AI, is based on learning new insights from existing data. For weather forecasting, the data that are needed are variables such as temperature, humidity, wind and pressure from different sources including weather satellite observations.

“In general, AI and cloud infrastructures, such as the European Weather Cloud, are opening up access to weather data and changing the way organisations and users interact with data and the way their work is structured and developed. 

“At the same time we are also seeing innovation in the commercial sector, as companies such as Google, Huawei and NVIDIA have been developing new machine learning-based weather forecasting models with weather data and having some success,” Ruti explains. 

Simultaneously, traditional weather forecasting is also advancing due to continuous improvements in existing technologies and satellite observations.

“The Meteosat Third Generation weather satellites will herald a new era in weather forecasting and Earth observation. They are the most complex and innovative meteorological geostationary satellite systems ever developed and they will provide 50 times more data than the current Meteosat fleet.”