ARENA: Active Region attENtion Aggregation for interpretable solar flare prediction
Abstract
Predicting solar flares is a challenging space-weather forecasting problem with potentially significant impacts on Earth's upper atmosphere, including disruptions to radio communications. We propose Active Region attENtion Aggregation (ARENA), a neural network architecture that predicts solar flares from a variable number of active-region cutouts while learning to identify the region most relevant to the prediction. ARENA is trained using multiple-instance learning under weak supervision, requiring only a flare/no-flare label for the observation and no labels specifying the flare origin. The learned attention mechanism therefore provides a means of localizing the active region driving the prediction without explicit origin annotations. ARENA achieves a heidke skill score of 0.45 which is comparable to other flare prediction approaches. These results demonstrate that active-region-based aggregation can provide an efficient and interpretable approach to solar flare prediction. The code is available at https://github.com/LuisBauer22/ARENA.
Related
- Cross-layer coupling in active-region upflows: Statistical connections between photospheric magnetic field, chromospheric activity, and coronal blueshifts
- Connectivity rewiring and filament formation by active region emergence observed with Solar Orbiter
- Connectivity rewiring and filament formation by active region emergence observed with Solar Orbiter
- The coronal mass ejection productivity of solar active region 13664/8
- Causal Discovery in Photospheric Magnetic Field Time Series for Interpretable Solar Flare Prediction
BibTeX
@misc{bauer2026arena,
title = {{ARENA: Active Region attENtion Aggregation for interpretable solar flare prediction}},
author = {Luis Bauer and Annalena Kofler and Andrés Asensio Ramos and Aparna Venkataramanasastry and Valentín Martínez Pillet and Bernhard Schölkopf},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.32758},
eprint = {2610.32758},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.32758},
}
RIS
TY - UNPB
TI - ARENA: Active Region attENtion Aggregation for interpretable solar flare prediction
AU - Luis Bauer
AU - Annalena Kofler
AU - Andrés Asensio Ramos
AU - Aparna Venkataramanasastry
AU - Valentín Martínez Pillet
AU - Bernhard Schölkopf
PY - 2026
DA - 2026/10/23
PB - arXiv
DO - 10.48550/arxiv.2610.32758
UR - https://arxiv.org/abs/2610.32758
AB - Predicting solar flares is a challenging space-weather forecasting problem with potentially significant impacts on Earth's upper atmosphere, including disruptions to radio communications. We propose Active Region attENtion Aggregation (ARENA), a neural network architecture that predicts solar flares from a variable number of active-region cutouts while learning to identify the region most relevant to the prediction. ARENA is trained using multiple-instance learning under weak supervision, requiring only a flare/no-flare label for the observation and no labels specifying the flare origin. The learned attention mechanism therefore provides a means of localizing the active region driving the prediction without explicit origin annotations. ARENA achieves a heidke skill score of 0.45 which is comparable to other flare prediction approaches. These results demonstrate that active-region-based aggregation can provide an efficient and interpretable approach to solar flare prediction. The code is available at https://github.com/LuisBauer22/ARENA.
ER -