Adversarial Observations in Weather Forecasting
By: Erik Imgrund, Thorsten Eisenhofer, Konrad Rieck
Potential Business Impact:
Tricks weather AI to show fake storms.
AI-based systems, such as Google's GenCast, have recently redefined the state of the art in weather forecasting, offering more accurate and timely predictions of both everyday weather and extreme events. While these systems are on the verge of replacing traditional meteorological methods, they also introduce new vulnerabilities into the forecasting process. In this paper, we investigate this threat and present a novel attack on autoregressive diffusion models, such as those used in GenCast, capable of manipulating weather forecasts and fabricating extreme events, including hurricanes, heat waves, and intense rainfall. The attack introduces subtle perturbations into weather observations that are statistically indistinguishable from natural noise and change less than 0.1% of the measurements - comparable to tampering with data from a single meteorological satellite. As modern forecasting integrates data from nearly a hundred satellites and many other sources operated by different countries, our findings highlight a critical security risk with the potential to cause large-scale disruptions and undermine public trust in weather prediction.
Similar Papers
On the Predictive Skill of Artificial Intelligence-based Weather Models for Extreme Events using Uncertainty Quantification
Atmospheric and Oceanic Physics
AI weather forecasts warn of floods and heatwaves.
Forecasting Fails: Unveiling Evasion Attacks in Weather Prediction Models
Machine Learning (CS)
Makes weather AI forecasts easier to trick.
Data-Efficient Ensemble Weather Forecasting with Diffusion Models
Machine Learning (CS)
Makes weather forecasts better using less data.