EarthCube Data Capabilities: Machine Learning Enhanced Cyberinfrastructure for Understanding and Predicting the Onset of Solar Eruptions

太空天气是一个术语，用来描述由太阳表面的爆发（如太阳耀斑）引起的太阳系环境条件的变化。了解和预测太阳爆发对国家安全和经济至关重要，因为众所周知，它们会对卫星和配电网络等关键技术基础设施产生不利影响。太阳爆发是由太阳黑子的复杂动力学引起的，太阳黑子通常被称为太阳活动区。本研究的目标是利用地面观测站和卫星任务的先进数据，建立数据基础设施，以表征1970年至今太阳活动区的特性。由物理学家和计算机科学家共同开发的数据库和相关的网络基础设施将利用先进的人工智能和机器学习。通过使用这个先进的数据库，可以更好地了解太阳活动区域以及它们是如何引发太阳爆发的。该项目有重要的教育和培训组成部分，将涉及研究生和初级研究人员。该项目将建立先进的计算机基础设施来描述太阳活动区（ARs）的特征，并应用机器学习工具来预测两种最重要的太阳爆发形式：太阳耀斑和日冕物质抛射（cme）。该项目将解决两个关键的科学问题：(1)哪些参数和物理过程对太阳爆发的发生最重要？(2)利用这些参数预测太阳喷发的准确性如何？这项工作将利用并与先前EarthCube项目下开发的基础设施相连接。它将分析来自大熊太阳天文台（BBSO） 1970年至今的数字化和数字化高分辨率数据，当前卫星任务数据，以及更全面的耀斑和相关ar档案的遗留数据。ar的动态非势性将使用先进的成像和机器学习工具推导。深度学习技术将用于追踪太阳色球层和日冕中的纤维/环结构。将这些与日冕场外推相结合，将为描述ar中的非电位提供新的参数。两个可能与耀斑和日冕物质抛射密切相关的新参数将被推导出来：耀斑产生ARs中的流动运动和磁螺旋注入。基于耀斑/日冕物质抛射的特性和从宿主ar中获得的重要参数，深度学习技术将进一步用于预测耀斑和日冕物质抛射的发生和能量范围。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Space weather is a term used to describe changing environmental conditions in the solar system caused by eruptions on the Sun's surface such as solar flares. Understanding and forecasting of solar eruptions is critically important for national security and for the economy since they are known to have adverse effects on critical technology infrastructure such as satellite and power distribution networks. Solar eruptions are caused by complex dynamics of sunspots which are often called solar active regions. The goal of this research is to build data infrastructure to characterize the properties of solar active regions from 1970 to now using advanced data from ground-based observatories and satellite missions. The database and associated cyberinfrastructure, jointly to be developed by physicists and computer scientists, will utilize advanced artificial intelligence and machine learning. By using this advanced database, a better understanding of the solar active regions and how they trigger solar eruptions will be achieved. The project has significant education and training components that will involve graduate students and junior researchers. The project will build advanced computer infrastructure to characterize solar active regions (ARs) and apply machine learning tools to predict two most significant forms of solar eruptions: the solar flares and coronal mass ejections (CMEs). The project will address two key science questions: (1) Which parameters and physical processes are most important for the onset of solar eruptions? (2) What is the accuracy of using these parameters to predict solar eruptions? The work will utilize and interface with the infrastructure developed under a previous EarthCube project. It will analyze digitized and digital high-resolution data from the Big Bear Solar Observatory (BBSO) from 1970 to now, current satellite mission data, as well as legacy data for a more comprehensive archive of flares and associated ARs. Dynamic non-potentiality properties of ARs will be derived using advanced imaging and machine learning tools. Deep learning techniques will be used to trace fibril/loop structures in the solar chromosphere and corona. Combining these with coronal field extrapolation will provide novel parameters to describe non-potentiality in ARs. Two new parameters will be derived that may be critically linked to flares and CMEs: flow motions and magnetic helicity injection in flare productive ARs. Based on flare/CME properties and important parameters derived from hosting ARs, deep learning techniques will be further adapted to predict the occurrence and energy range of flares and CMEs.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.