![[2602.22298] AviaSafe: A Physics-Informed Data-Driven Model for Aviation Safety-Critical Cloud Forecasts](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22298] AviaSafe: A Physics-Informed Data-Driven Model for Aviation Safety-Critical Cloud Forecasts
Summary
AviaSafe introduces a physics-informed, data-driven model for aviation cloud forecasts, enhancing prediction accuracy for critical hydrometeor species essential for aviation safety.
Why It Matters
This research addresses significant challenges in aviation safety by improving cloud forecasting accuracy, which is crucial for flight operations and safety. The model's ability to distinguish between cloud species can lead to better decision-making in aviation route optimization, ultimately enhancing safety and efficiency in air travel.
Key Takeaways
- AviaSafe predicts cloud microphysical species critical for aviation safety.
- The model integrates a physics-based Icing Condition index for enhanced accuracy.
- It outperforms traditional numerical models in key forecasting variables.
- The hierarchical architecture improves cloud spatial distribution predictions.
- Forecasting individual cloud species aids in aviation route optimization.
Computer Science > Machine Learning arXiv:2602.22298 (cs) [Submitted on 25 Feb 2026] Title:AviaSafe: A Physics-Informed Data-Driven Model for Aviation Safety-Critical Cloud Forecasts Authors:Zijian Zhu, Qiusheng Huang, Anboyu Guo, Xiaohui Zhong, Hao Li View a PDF of the paper titled AviaSafe: A Physics-Informed Data-Driven Model for Aviation Safety-Critical Cloud Forecasts, by Zijian Zhu and 4 other authors View PDF HTML (experimental) Abstract:Current AI weather forecasting models predict conventional atmospheric variables but cannot distinguish between cloud microphysical species critical for aviation safety. We introduce AviaSafe, a hierarchical, physics-informed neural forecaster that produces global, six-hourly predictions of these four hydrometeor species for lead times up to 7 days. Our approach addresses the unique challenges of cloud prediction: extreme sparsity, discontinuous distributions, and complex microphysical interactions between species. We integrate the Icing Condition (IC) index from aviation meteorology as a physics-based constraint that identifies regions where supercooled water fuels explosive ice crystal growth. The model employs a hierarchical architecture that first predicts cloud spatial distribution through masked attention, then quantifies species concentrations within identified regions. Training on ERA5 reanalysis data, our model achieves lower RMSE for cloud species compared to baseline and outperforms operational numerical models on certain...
