FireCastNet: Earth-as-a-Graph for Seasonal Fire Prediction
Dimitrios Michail, Charalampos Davalas, Konstantinos Chafis, Lefki-Ioanna Panagiotou, Ioannis Prapas, Spyros Kondylatos, Nikolaos Ioannis Bountos, Ioannis Papoutsis · 2025
Read FireCastNet: Earth-as-a-Graph for Seasonal Fire Prediction by Dimitrios Michail, Charalampos Davalas, Konstantinos Chafis, Lefki-Ioanna Panagiotou, Ioannis Prapas, Spyros Kondylatos, Nikolaos Ioannis Bountos, Ioannis Papoutsis (2025) 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/arxiv-2502-01550.txt.
Description
Dimitrios Michail and co-authors present FireCastNet, a model that forecasts wildfire burn patterns a full season ahead. As climate change loads the dice for fire weather, early warning guides everything from fuel clearing to insurance. Their design treats the Earth itself as a graph, letting distant climate links speak.
A three-dimensional convolutional encoder reads the SeasFire datacube of climate, vegetation, and human factors. Graph neural networks, kin to weather models, then pass information across the globe's mesh. The system predicts burned area up to six months out, capturing both local sparks and long-range teleconnections. Seasonal foresight could move fire management from reaction to preparation.
Catalog record
| CREATOR | Dimitrios Michail, Charalampos Davalas, Konstantinos Chafis, Lefki-Ioanna Panagiotou, Ioannis Prapas, Spyros Kondylatos, Nikolaos Ioannis Bountos, Ioannis Papoutsis |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-01550