Graph Learning with Multiplicative Aggregation for Solar Filament Fragment Association Modeling in Full-disk Hα Images
Abstract
Solar filaments are prominent features in Hα observations. During their evolution, they often exhibit complex morphological and topological changes, which, due to line-of-sight projection effects, appear as fragmentation and merging. These phenomena make it difficult to reliably track solar filaments using automated methods, as the same structure may appear different over time. To address this issue, we formulate filament fragment association as a graph-based learning problem, where individual fragments are represented as nodes and candidate pairwise relationships are modeled as edges to infer whether fragments belong to the same filament. Furthermore, we propose a novel graph neural network (GNN) that incorporates multiplicative aggregation layers to enhance feature aggregation in the graph, replacing conventional additive aggregation with a multiplicative mechanism to better capture information between fragments. Experimental results demonstrate that the proposed method achieves reliable performance in filament fragment association, validating the effectiveness of the graph-based formulation and the proposed multiplicative aggregation mechanism.
Versions
Related
- Connectivity rewiring and filament formation by active region emergence observed with Solar Orbiter
- Insights into the Kinematic Evolution of Filament-Induced Jets: Coupling of Interchange and Flare Reconnection
- Insights into the Kinematic Evolution of Filament-induced Jets: Coupling of Interchange and Flare Reconnection
- Magnetized Interstellar Molecular Clouds. III. Filament Collisions and Core Formation: Insights into Substructures and Evolution
- Magnetized interstellar molecular clouds - III. Filament Collisions and Core Formation: Insights into Substructures and Evolution
BibTeX
@misc{zhu2026graph,
title = {{Graph Learning with Multiplicative Aggregation for Solar Filament Fragment Association Modeling in Full-disk H$α$ Images}},
author = {GaoFei Zhu and GangHua Lin and Xiao Yang and Suo Liu and Yu Liu and ZhiSheng Gao and Cheng Zeng},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.17291},
eprint = {2610.17291},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.17291},
}
RIS
TY - UNPB
TI - Graph Learning with Multiplicative Aggregation for Solar Filament Fragment Association Modeling in Full-disk Hα Images
AU - GaoFei Zhu
AU - GangHua Lin
AU - Xiao Yang
AU - Suo Liu
AU - Yu Liu
AU - ZhiSheng Gao
AU - Cheng Zeng
PY - 2026
DA - 2026/10/12
PB - arXiv
DO - 10.48550/arxiv.2610.17291
UR - https://arxiv.org/abs/2610.17291
AB - Solar filaments are prominent features in Hα observations. During their evolution, they often exhibit complex morphological and topological changes, which, due to line-of-sight projection effects, appear as fragmentation and merging. These phenomena make it difficult to reliably track solar filaments using automated methods, as the same structure may appear different over time. To address this issue, we formulate filament fragment association as a graph-based learning problem, where individual fragments are represented as nodes and candidate pairwise relationships are modeled as edges to infer whether fragments belong to the same filament. Furthermore, we propose a novel graph neural network (GNN) that incorporates multiplicative aggregation layers to enhance feature aggregation in the graph, replacing conventional additive aggregation with a multiplicative mechanism to better capture information between fragments. Experimental results demonstrate that the proposed method achieves reliable performance in filament fragment association, validating the effectiveness of the graph-based formulation and the proposed multiplicative aggregation mechanism.
ER -