分两阶段预测极光可见性，准确率达90%以上。
Aurora Hunter: A Two-Stage Framework for Probabilistic Visibility Forecasting
- 先判断极光是否发生，再评估观测条件是否允许看见。
- 在特罗姆瑟和基鲁纳数据上准确率分别达93.7%和90.5%。
- 适合极光旅游规划与空间天气研究者使用。
极光可见性预测对空间天气研究和极光旅游至关重要。某地某夜的可见性取决于两个因素：一是极光是否真实发生（由太阳风-磁层耦合驱动），二是观测条件是否允许肉眼识别（主要受云量和月光影响）。我们提出 Aurora Hunter，一种两级级联框架，将这两个因素解耦。第一阶段用 XGBoost 预测极光发生的概率，基于51个物理驱动特征，使用2015-2023年特罗姆瑟+基鲁纳联合数据（约16,600小时样本）训练，标签来自特罗姆瑟AI全景图像分类器。第二阶段用逻辑回归预测在极光发生前提下可观测的概率，使用21个云量与月光特征，仅在极光发生时段数据上训练。级联公式为：P(可见) = P(发生) × P(清晰|发生)，在特罗姆瑟测试集（2019–2020）达到ROC-AUC 0.937，独立基鲁纳测试集（2024）达0.905，比单阶段基线提升0.087。留出的斯卡比翁数据（2022–2025）验证了跨站点泛化能力。SHAP分析显示，Kp与夜侧交互、时区位置、极光卵距离是主导因子（合计贡献39%）。原型系统：https://aurora-hunter.onrender.com。
原文摘要 · Abstract (English)
Forecasting aurora borealis visibility matters for space weather research and aurora tourism. Visibility at a site and night depends on two distinct factors: (1) whether aurora is physically occurring, driven by solar wind-magnetosphere coupling, and (2) whether observing conditions allow naked-eye detection, mainly cloud cover and lunar illumination. We present Aurora Hunter, a two-stage cascade that decouples these factors. Stage 1 predicts P(occurring) with XGBoost using 51 physics-driven features trained on joint Tromso+Kiruna data (about 16,600 hourly samples, 2015-2023) with labels from the Tromso AI all-sky image classifier. Stage 2 predicts P(clear observation given occurring) with logistic regression using 21 cloud-cover and lunar-illumination features trained only on aurora-occurring hours. The cascade P(visible)=P(occurring)*P(clear|occurring) reaches ROC-AUC 0.937 (Tromso test, 2019-2020) and 0.905 (independent Kiruna, 2024), improving a single-stage baseline by +0.087. Held-out Skibotn data (2022-2025) confirm cross-site generalization. SHAP identifies the Kp x nightside interaction, MLT position, and auroral oval distance as dominant predictors (39% combined). Prototype: https://aurora-hunter.onrender.com.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。