基于海量观测数据的太阳爆发事件预报建模研究

Solar activities are the source of the Sun-Earth space environment changes, and the social development and national security could be affected by the space weather hazards. On the one hand, the physical characteristics of solar active region are analyzed, and the non- potentiality, the displacement parameter between the center of photospheric free magnetic energy and the center of the magnetic field and the magnetic bounded degree above active regions are calculated as predictors of solar eruptive events. On the other hand, multi-band coronal patterns are automatically extracted by the deep learning method from the massive extreme ultraviolet (EUV) images. Combined with the magnetic field parameters of active regions and characteristics extracted from EUV coronal images, a solar eruption events forecasting model, which considers the solar flares and coronal mass ejections together, is proposed. This model greatly improves the short-term forecasting ability of solar eruptions, and provides better space weather forecasting services.

太阳活动是日地空间环境变化的源头，由太阳爆发引起的空间天气灾害会对高科技依赖的人类社会和国家安全产生重要的影响。本项目一方面分析太阳活动区的物理特性，提取活动区磁场非势性，活动区自由磁能偏离磁场中心的程度以及活动区上方的磁场束缚程度等物理参量，形成太阳爆发活动预报的物理参量；另一方面从积累的大量日冕极紫外观测图像中，利用深度学习方法自动提取活动区日冕的多波段观测特性形成太阳爆发预报的参量。结合太阳活动区物理参量及自动提取的日冕多波段特征参量，建立综合考虑太阳耀斑和日冕物质抛射的太阳爆发事件预报模型。该模型的建立极大地提高了短期太阳爆发活动的预报能力，能够更好地为空间天气预报服务。

太阳活动是日地空间环境变化的源头，由太阳爆发引起的空间天气灾害会对高科技依赖的 人类社会和国家安全产生重要的影响。本项目一方面分析太阳活动区的物理特性，提取活动区磁场非势性等物理参量；另一方面从积累的大量日冕极紫外观测图像中，利用深度学习方法自动提取活动区日冕的多波段观测特性形成太阳爆发预报的参量。最终建立综合考虑太阳耀斑和日冕物质抛射的太阳爆发事件预报模型。本项目建立了太阳活动区三维磁场数据集和多波段图像数据集，提取了活动区多波段图像的爆发先找特征，并建立了太阳爆发活动预报模型。该模型的建立极大地提高了短期太阳爆发活动的预报能力，能够更好地为空间天气预报服务。