Evaluating the skill of AI and physics-based models for February 2020 North Atlantic storm hazards
Hilla Afargan-Gerstman, Rachel W-Y Wu, Alice Ferrini, Daniela I V Domeisen · 2026
Read Evaluating the skill of AI and physics-based models for February 2020 North Atlantic storm hazards by Hilla Afargan-Gerstman, Rachel W-Y Wu, Alice Ferrini, Daniela I V Domeisen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13619441.txt.
Description
Hilla Afargan-Gerstman led a 2026 test of data-driven weather models against physics-based models during a punishing run of North Atlantic storms in February 2020. Three extratropical storms struck the United Kingdom in close succession, pairing damaging winds with heavy rain, a sequence that stresses forecasts more than a single storm does. The team compared both model families on that storm series, looking at pressure patterns, surface winds, and related hazards.
The AI models generally beat the physics-based model on weekly sea-level pressure forecasts, and to a lesser degree on surface winds. Some individual physics-based ensemble members matched or occasionally beat the AI models, however, and weaker links between atmospheric variables in AI output hinted that physical rules were being missed. Forecasters get a balanced verdict: data-driven models are fast and often strong, but reliable multi-hazard warnings still need them paired with physics-based methods.
Catalog record
| CREATOR | Hilla Afargan-Gerstman, Rachel W-Y Wu, Alice Ferrini, Daniela I V Domeisen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |