JW-FD: A 15-Year Multimodal Dataset for Solar Flare Forecasting
Abstract
Solar flares drive severe space weather hazards, and forecasting their occurrence remains a central challenge for both heliophysics and operational space weather services. Data driven methods require long horizon datasets in which images, magnetic features, and flare labels are coregistered in space and time. We present JW-FD (JW-Flare Dataset), a 15 year multimodal release spanning 1~January~2011 through 31~December~2025, constructed from SDO/HMI line of sight magnetograms, NOAA Solar Region Summary reports, and NOAA X-ray flare event lists. The dataset comprises 3{,}064 independent active regions and 1{,}991{,}247 coregistered magnetogram crops, together with FITS, PNG, CSV, and MP4 modalities. Each sample provides 29 magnetic features linked to configurable flare labels under a strict pre-eruption window spanning seven forecast horizons and four GOES intensity thresholds. An 8:1:1 split at the active region level is adopted to prevent temporal leakage between partitions. PNG branches are released at six magnetic saturation thresholds, and internal Transformer experiments on ≥C1.0 forecasting suggest Bth=1000~G as a preliminary default, although the optimal saturation is model and task dependent. The open source construction pipeline is available at https://github.com/Xiaoxuan-1/JW-FD.
Versions
Related
- Short-term solar flare forecasting with a hybrid ConvNeXtV2 and transformer–temporal convolutional network
- SHAPE: Solar Heliospherical Analysis and Prediction of Eruptions A machine learning approach to solar flare forecasting
- A multi-source dataset for solar energetic particle forecasting
- ARENA: Active Region attENtion Aggregation for interpretable solar flare prediction
- Direct Injection of Chemically-Enhanced Plasma into a Solar Flare Arcade
BibTeX
@misc{mingfu2026jwfd,
title = {{JW-FD: A 15-Year Multimodal Dataset for Solar Flare Forecasting}},
author = {Shao Mingfu and Lin Jiaben and Wang Hui and Tong Liyue and Yang Chen and Zhang Yin and Li Yuyang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2608.19195},
eprint = {2608.19195},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.19195},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - JW-FD: A 15-Year Multimodal Dataset for Solar Flare Forecasting
AU - Shao Mingfu
AU - Lin Jiaben
AU - Wang Hui
AU - Tong Liyue
AU - Yang Chen
AU - Zhang Yin
AU - Li Yuyang
PY - 2026
DA - 2026/09/30
PB - arXiv
DO - 10.48550/arxiv.2608.19195
UR - https://arxiv.org/abs/2608.19195
AB - Solar flares drive severe space weather hazards, and forecasting their occurrence remains a central challenge for both heliophysics and operational space weather services. Data driven methods require long horizon datasets in which images, magnetic features, and flare labels are coregistered in space and time. We present JW-FD (JW-Flare Dataset), a 15 year multimodal release spanning 1~January~2011 through 31~December~2025, constructed from SDO/HMI line of sight magnetograms, NOAA Solar Region Summary reports, and NOAA X-ray flare event lists. The dataset comprises 3{,}064 independent active regions and 1{,}991{,}247 coregistered magnetogram crops, together with FITS, PNG, CSV, and MP4 modalities. Each sample provides 29 magnetic features linked to configurable flare labels under a strict pre-eruption window spanning seven forecast horizons and four GOES intensity thresholds. An 8:1:1 split at the active region level is adopted to prevent temporal leakage between partitions. PNG branches are released at six magnetic saturation thresholds, and internal Transformer experiments on ≥C1.0 forecasting suggest B_th=1000~G as a preliminary default, although the optimal saturation is model and task dependent. The open source construction pipeline is available at https://github.com/Xiaoxuan-1/JW-FD.
ER -