On Thursday, May 21, the release of the annual forecasts from the US National Oceanic and Atmospheric Administration (NOAA) for the Atlantic hurricane season is eagerly anticipated. However, these forecasts could lose precision and reliability due to a lack of data. Many specialists are concerned about the budget cuts implemented by the Trump administration, which affect various areas of research, particularly climate-related research.
Despite this, in December, NOAA introduced brand new weather forecasting models based on artificial intelligence (AI), with attractive features: they are said to be faster, more efficient, and more precise. However, these promising models need data to train.
“Reducing funds allocated to climate research has a negative effect on weather forecasts and hinders their development,” explained Craig McLean, former NOAA research director, in an interview with The Guardian.
More Data Than Ever Needed
The British journal expresses concern that the disruption of Earth observation systems – such as reducing satellite launches, meteorological balloons, oceanic buoys, and cutting researchers who analyze data – is happening at a time when forecasts “will be most necessary” as extreme weather events are expected to become more frequent and intense in the future, due to climate change.
Meteorologist Chris Gloninger warns: “If the government continues to increase its reliance on AI-based models while reducing the amount of data that fuels them, government forecasts could be compromised.”
Indeed, the new AI-based models are still insufficient to predict extreme weather events, as they rely on historical data from a time when such events were rare. Consequently, “they tend to predict weather phenomena similar to those observed in the past,” explains The Guardian, emphasizing the importance of acquiring new data.
