European Weather Organization Collaborates with Huawei for AI-Based Forecasting
European Centre for Medium-Range Weather Forecasts (ECMWF) has partnered with Huawei Technologies, a Chinese tech giant, to bring artificial intelligence (AI) into weather forecasting. The collaboration aims to revolutionize the industry by integrating AI models into the daily forecasts provided by the European weather agency.
Huawei’s AI model, called Pangu-Weather, has been launched on ECMWF’s website. The model offers global weather predictions with high accuracy and faster processing speeds compared to traditional numerical prediction methods. Pangu-Weather has shown exceptional performance in predicting typhoon paths, heatwaves, cold waves, and other extreme weather conditions.
Tian Qi, Huawei’s chief project scientist, highlighted the model’s advantages, including lower computational costs and improved accuracy indicators. According to a study published in the scientific journal Nature, the Pangu model demonstrated higher precision in its forecasts and a prediction speed that was 10,000 times faster than traditional methods.
Pangu-Weather has already proven its capabilities by accurately predicting the trajectory of Typhoon Doksuri and the sharp turn of Typhoon Khanun. Its forecasts have become a valuable reference for professional forecasters and weather enthusiasts worldwide.
Huawei’s ultimate goal is to develop a next-generation AI weather forecasting framework using Pangu-Weather. The company aims to enhance weather prediction accuracy, speed, and response to extreme weather conditions. Discussions are also underway with the China Meteorological Administration for collaboration in this field.
Although the European Union has been considering a ban on using companies deemed to present a security risk, ECMWF recognizes the potential of AI models as a quiet revolution in weather forecasting. Dr. Florian Pappenberger, forecast department director at ECMWF, believes that data-driven modeling based on machine learning can significantly improve traditional numerical weather prediction.
AI models can uncover new atmospheric evolution patterns from vast datasets, providing insights that were previously impossible for humans to perceive. They have the potential to break the monopolies of traditional forecasting methods and overcome the limitations in numerical weather prediction accuracy.
Despite its strengths, Pangu-Weather may not accurately predict specific landing points of typhoons due to different initial data and the need for further consideration of physical laws. Thus, numerical models still play a crucial role in predicting typhoon strength.
The collaboration between ECMWF and Huawei marks an exciting advancement in weather forecasting. By embracing AI, the European weather agency aims to enhance the accuracy and speed of its predictions, ultimately helping humanity respond better to extreme weather conditions.
Weather enthusiasts, meteorologists, and the general public can now access Pangu-Weather’s 10-day global weather forecasts on ECMWF’s website. The integration of AI models into weather forecasting represents a significant step forward, ensuring more reliable and timely information for decision-making in various sectors affected by weather conditions.
