KAIROS Presents AI-Powered Volcanic Plume Forecasting at IAVCEI Scientific Assembly

The KAIROS project was highlighted last week at the IAVCEI 2025 Scientific Assembly in Geneva through an oral presentation by Hugues Brenot of the Royal Belgian Institute for Space Aeronomy (BIRA-IASB).

His talk, titled "KAIROS AI-digital solution: a novel approach to forecasting the transport of SO₂-rich volcanic plumes," was delivered as part of the session on Operational hazard forecasts: source parameters and model evaluation.

AI-Enhanced Forecasting for Aviation and Hazard Response

Volcanic eruptions release large quantities of sulphur dioxide (SO₂), posing a serious threat to aviation safety and public health. The KAIROS system addresses this challenge by combining:

• Advanced AI algorithms

• Historical data over several months

• Near-real-time satellite inputs (within 10 minutes to 3 hours)

• Observations from ultraviolet and infrared sensors on GEO and LEO satellites

  
  

These elements work together to produce high-resolution, time-sensitive SO₂ transport forecasts, enabling earlier and more informed decision-making for air traffic management, civil protection authorities and government agencies monitoring volcanic hazards.

This research is the result of a multi-institutional collaboration, including the Royal Belgian Institute for Space Aeronomy (BIRA-IASB), Belgium; AI Methods, Spain; University Carlos III of Madrid (UC3M), Spain; and Meteomatics, Germany. Together with the rest of the KAIROS project partners, the team is building a digital forecasting system that provides more accurate, timely, and operationally useful information for managing the impacts of weather hazards such as volcanic SO₂ plumes.