Artificial Intelligence (AI) and Machine Learning (ML) have the potential to revolutionize weather forecasting by accurately predicting extreme weather events, as noted in a report coordinated by the World Meteorological Organization (WMO). The report, titled “United in Science: Reboot Climate Action,” highlights that AI/ML models have outperformed traditional physics-based methods in forecasting variables like tropical cyclones and the El Niño Southern Oscillation (ENSO) up to three years in advance.
As global warming trends continue to intensify, disrupting weather patterns worldwide, the report emphasizes the urgency of adopting these advanced technologies. Current projections estimate a global temperature rise of up to 3°C above pre-industrial levels, unless substantial climate actions are taken.
While AI/ML models provide faster and cost-effective forecasting by utilizing reanalysis and observational data, they face challenges such as limited data availability, low model resolution, and ethical concerns around transparency and access. The report suggests that despite these challenges, AI and ML can significantly improve the accuracy of forecasts, especially for single-day rain events, which often vary from predictions.
India has recently launched “Mission Mausam” to enhance forecast precision using AI/ML alongside improved weather models. The initiative includes deploying additional radars and technology to improve nowcasting, enabling more localized and timely weather warnings.
Experts stress that human expertise is still crucial in interpreting AI-driven forecasts, as models can have systematic biases. AI and ML methods can help refine model outputs, reducing errors and enhancing decision-making processes in weather prediction.
