Reanalysis data fusion-based tropical cyclone formation prediction network
Ying Hai, Yuan Peng, Yuchen Wang, Zhiyong Cheng · 2026
Read Reanalysis data fusion-based tropical cyclone formation prediction network by Ying Hai, Yuan Peng, Yuchen Wang, Zhiyong Cheng (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/PMC13575148.txt.
Description
Ying Hai and three colleagues build a neural network that forecasts whether a tropical cloud cluster will grow into a tropical storm, in a 2026 paper. Early warning of cyclone formation helps coastal regions prepare. Their system learns from weather reanalysis data, which blends past observations into gridded maps.
Data from two reanalysis sources were turned into images at several scales and labelled with known formation events. A pyramid attention module picks out formation-related features, a fusion step weights and merges the scales, and a light convolutional network predicts whether a storm forms and where. At 24 hours ahead, detection reached 0.865 with a false alarm ratio of 0.326 across five test splits. Location error averaged about 437 kilometres.
Catalog record
| CREATOR | Ying Hai, Yuan Peng, Yuchen Wang, Zhiyong Cheng |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 71374 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |