Saved in:
Bibliographic Details
Main Authors: Ge, Zongyuan, Zhang, Chenwaner, Li, Haoyang, Zhang, Hantai, Zhou, Wei, Gu, Wenxin, Wang, Zhaoming
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.24038
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917532870377472
author Ge, Zongyuan
Zhang, Chenwaner
Li, Haoyang
Zhang, Hantai
Zhou, Wei
Gu, Wenxin
Wang, Zhaoming
author_facet Ge, Zongyuan
Zhang, Chenwaner
Li, Haoyang
Zhang, Hantai
Zhou, Wei
Gu, Wenxin
Wang, Zhaoming
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2605_24038
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Aurora Hunter: A Two-Stage Framework for Probabilistic Visibility Forecasting
Ge, Zongyuan
Zhang, Chenwaner
Li, Haoyang
Zhang, Hantai
Zhou, Wei
Gu, Wenxin
Wang, Zhaoming
Space Physics
Earth and Planetary Astrophysics
Instrumentation and Methods for Astrophysics
Machine Learning
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.
title Aurora Hunter: A Two-Stage Framework for Probabilistic Visibility Forecasting
topic Space Physics
Earth and Planetary Astrophysics
Instrumentation and Methods for Astrophysics
Machine Learning
url https://arxiv.org/abs/2605.24038