SHINE: Prediction of Coronal Mass Ejections and Interplanetary Magnetic Fields Using Advanced Artificial Intelligence Techniques

了解和预测猛烈的太阳喷发及其对地球的影响是一个战略性的国家优先事项，因为它影响到人类的日常生活，包括通信、交通、供电系统、国防、太空旅行等。由于太阳仪器的空间和时间分辨率不断提高，研究人员在分析海量空间气象数据方面面临着巨大的挑战，特别是在可操作的近实时利用方面。这个跨学科项目先进的基于人工智能(AI)的工具来预测地球有效的日冕物质抛射(CME)。该项目将资助一名研究生，并对本科生进行指导。该项目由地球科学局和空间气象方案共同资助，以支持在地球科学方面的人工智能/最大限度地推进。该项目支持关于高级人工智能的基础研究，以预测日冕物质抛射及其在1AU附近引起地磁暴的可能性。其目的是(I)利用人工智能来预测太阳活动区是否会产生CME，并估计其通过时间、质量和动能，以及(Ii)基于实时太阳观测，预测1AU附近的磁云(MC)的方向。数据将被图形神经网络(GNN)和集成学习方法使用，以将GNN与其他传统人工智能技术相结合，以预测MC的方向。该项目利用了NASA、NOAA和NSF天文台的数据，包括NSF资助的全球H-阿尔法网络和大熊太阳天文台。这项研究还将利用现有的测量和模型结果，这些结果将通过来自全球日冕场图和非线性无力场建模的额外测量来增强。该项目由地球科学局共同资助，以支持地球科学领域的AI/ML进步。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Understanding and predicting violent solar eruptions and their effect on Earth is a strategic national priority, as it affects the daily life of human beings, including communication, transportation, power supply systems, national defense, space travel, and more. Due to increasing spatial and temporal resolution of solar instrumentation, researchers are facing tremendous challenges in analyzing massive amounts of space weather data, especially for the operational near real-time utilization. This interdisciplinary project advanced artificial intelligence (AI) based tools to forecast geoeffective coronal mass ejections (CMEs). A graduate student will be supported by this project and undergraduate students will be mentored. This project is co-funded by the Directorate for Geosciences to support AI/ML advancement in the geosciences and by the Space Weather program.This project supports fundamental research on advanced AI to forecast CMEs and their potential to cause geomagnetic storms near 1 AU. The objectives are to (i) employ AI to predict whether a solar active region will produce a CME and estimate its transit time, mass, and kinetic energy, and (ii) predict the orientation of magnetic clouds (MCs) near 1 AU, based on real time solar observations. Data will be used by graph neural networks (GNNs) and ensemble learning methods to combine the GNNs with other conventional AI techniques to predict orientations of MCs. The project utilizes data from NASA, NOAA, and NSF observatories, including the NSF-funded Global H-alpha Network and Big Bear Solar Observatory. The study will also utilize existing measurements and model results, which will be augmented with additional measurements derived from global coronal field maps as well as non-linear force-free field modeling.This project is co-funded by the Directorate for Geosciences to support AI/ML advancement in the geosciences.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.