2026-10-14T23:38:40Z https://www.scidb.cn/oai

10.57760/sciencedb.space.00328 2026-09-29T11:19:05Z

2026-09-29 The Deep Learning–based Dual-branch Multimodal Fusion Model for Solar Flare Prediction doi:10.57760/sciencedb.space.00328 en The dataset is associated with solar flare forecasting research. It includes SDO/HMI SHARP active-region magnetograms, the corresponding magnetic-field parameters, a supplemented solar flare catalog, and solar flare forecasting labels.The Magnetogram data include active-region magnetograms in the cylindrical equal-area (CEA) projection, recorded from May 2010 to May 2023 at a cadence of 96 minutes.The Flare-catalog was constructed from the NOAA/NCEI GOES X-ray Flare Catalog. For flare events without associated NOAA active-region numbers, the corresponding NOAA active-region numbers were supplemented using the SolarSoft event archive and corresponding SDO observations.The Flare label file was generated from the Flare-catalog using a predefined 24-hour forecasting window. Samples are labeled as flare or non-flare according to whether a C-class or stronger flare occurs within the subsequent 24 hours. For multiclass forecasting, flare samples are further categorized as C-, M-, or X-class according to the highest flare class within the forecasting window. Solar flare; Solar activity; Solar forecasting; Deep learning Limin Zhao Xingyao Chen Xiaoshuai Zhu Dong Zhao Yihua Yan PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank