{"id":2651,"date":"2025-01-28T07:26:36","date_gmt":"2025-01-28T06:26:36","guid":{"rendered":"https:\/\/wp.unil.ch\/iaunil\/climate-and-cyclones-harnessing-ai-for-physics-to-better-predict-extreme-weather-conditions\/"},"modified":"2025-12-10T13:41:16","modified_gmt":"2025-12-10T12:41:16","slug":"climate-and-cyclones-harnessing-ai-for-physics-to-better-predict-extreme-weather-conditions","status":"publish","type":"post","link":"https:\/\/wp.unil.ch\/iaunil\/en\/climate-and-cyclones-harnessing-ai-for-physics-to-better-predict-extreme-weather-conditions\/","title":{"rendered":"Climate and cyclones: harnessing AI for physics to better predict extreme weather conditions"},"content":{"rendered":"\n

Original text plublished on: https:\/\/wp.unil.ch\/geoblog\/2024\/03\/modeliser-la-physique-atmospherique-prevoir-la-formation-de-cyclones-tropicaux-et-predire-le-climat-futur\/<\/a><\/em><\/p>\n\n

\n
\n

Climate models must handle vast amounts of data and complex atmospheric processes. Climate physicist at UNIL, Tom Beucler incorporates neural networks into his simulations to reduce computational cost while increasing their precision. Guided by the laws of physics, his hybrid models \u2014 at the intersection of AI and meteorology \u2014 enable better forecasting of extreme events such as tropical cyclones.By combining scientific rigor, computational power, and ethical reflection, this approach opens a new era for climate research. <\/p>\n<\/div>\n\n\n\n

To improve models thanks to neural networks while respecting laws of physic<\/h3>\n\n\n\n

The atmosphere obeys physical laws that can be formulated as equations. Atmospheric complexity stems from the vast array of interacting variables and scales. In seeking to model this complexity, Tom Beucler soon confronts the limitations of computational resources. By incorporating neural networks, he reduces the computational burden while enabling the integration of heterogeneous data sources \u2014 such as high-resolution simulations, meteorological radar, and satellite imagery. This approach enhances and streamlines the representation of intricate processes, making atmospheric models more efficient and accessible. <\/p>\n\n\n\n

Tom Beucler notes that calculations based solely on data and algorithms can sometimes produce results that violate physical laws \u2014 for instance, he has obtained outputs that failed to satisfy the conservation of mass and energy. To prevent such inconsistencies, he incorporates physical knowledge into his AI models through physics-guided machine learning, where AI systems operate on data within a framework constrained by physical principles, thereby avoiding incoherent outcomes. <\/p>\n\n\n\n

\"Beucler<\/figure>\n\n\n\n

Using statistical modeling tools, Tom Beucler and his team have already achieved improvements in simulating the formation and evolution of tropical cyclones. Whereas conventional models tend to generate a high number of false positives \u2014 forecasting cyclones that do not materialize \u2014 data-consistent models retain correct predictions while substantially lowering the rate of false detections.<\/p>\n\n\n\n

\r\n