Jakarta, INTI – Efforts to mitigate the devastating impact of tropical storms have entered a new era. Google, the tech giant from the United States, is developing a new artificial intelligence (AI) model capable of predicting the formation of tropical cyclones with a higher level of accuracy. Developed in collaboration with the U.S. National Hurricane Center (NHC), the technology is expected to provide more precise early warnings for communities in disaster-prone areas.
Earlier Predictions, Faster Evacuations
The AI model, developed by Google DeepMind and Google Research, is designed to forecast the path, size, and intensity of storms up to 15 days before they occur. In its initial trials, the model was able to generate up to 50 storm trajectory scenarios, offering a more comprehensive overview of potential impacts.
The collaboration with the NHC aims to integrate this AI technology into official forecasting systems so that people in affected regions have more time to prepare and evacuate if necessary.
Weather Lab: A Testing Ground for AI Forecasts
Google has introduced an interactive website called Weather Lab to showcase the capabilities of its AI model. On this platform, users can compare forecasts generated by Google’s AI with widely used physics-based models such as the ECMWF. Although still in the research and development phase, the website highlights AI’s vast potential in supporting future weather prediction.
“We hope this data can help improve NHC forecasts and deliver earlier and more accurate warnings for tropical cyclone hazards,” Google stated in its report.
Trained on Global Data, Competing with Leading Physics Models
The AI model is trained using data from Europe’s ERA5 archive, which contains hundreds of millions of weather observations collected from around the world. This vast dataset combined with predictions from traditional weather models—helps the AI learn and recognize storm patterns.
Google’s previous model, GenCast, even outperformed one of the leading physics-based models, ECMWF, by 97.2% according to a study published in Nature in December 2024. This achievement shows AI's enormous potential as a highly accurate weather forecasting tool in the future.
Not a Replacement, But a Complementary Tool
Despite its promise, Google emphasizes that this AI model is not meant to replace existing weather observation methods. Instead, it is intended to complement them. AI-powered detection can be used alongside traditional monitoring techniques to provide a more complete picture of extreme weather events.
Collaborations with institutions such as the Cooperative Institute for Research in the Atmosphere at Colorado State University, as well as researchers from the UK and Japan, are part of Google’s ongoing efforts to refine this technology amidst increasingly unpredictable climate conditions.
Conclusion
Google’s initiative to develop AI for tropical storm forecasting highlights the critical role of technological innovation in addressing global climate change. With its ability to detect and predict severe weather earlier, this model could become a vital tool in saving lives and reducing damage caused by natural disasters.
Read More:Deputy Minister Urges Diaspora Youth in London to Build Indonesia’s AI Future