The AI-Powered Solar Eruption Center of Excellence in Research and Education (SEC) will collaborate with NASA, New York University, and IBM to develop AI and machine learning tools aimed at improving the prediction of solar flares and coronal mass ejections (CMEs). These efforts will enhance understanding of these solar events and advance forecasting capabilities.

The funding, provided through NASA's Minority University Research and Education Project (MUREP) Institutional Research Opportunity (MIRO), is part of a $45 million initiative to support research at 21 universities across the U.S. NJIT is one of seven minority-serving institutions to receive a portion of the funds, which will also support educational programs related to space science.

"This grant establishes a first-of-its-kind hub where cutting-edge advances in AI, and space weather research and education converge," said Haimin Wang, ISWS director and NJIT physics professor, who will lead the project. "By harnessing AI-enabled tools to investigate the fundamental nature of space weather, we aim to significantly enhance our ability to interpret observational data from the Sun to forecast major solar eruptions accurately and in near real-time, a capability beyond our reach up to this point."

Bo Shen, SEC associate director and NJIT engineering professor, added, "We aim to push the boundaries of interpretable AI and physics-informed learning by integrating physics knowledge with advanced AI tools, ensuring that models not only make accurate predictions but also provide insights aligned with fundamental physical principles."

Solar flares and CMEs, fueled by magnetic energy, are key drivers of space weather phenomena, including solar geomagnetic storms that can impact satellite operations and power grids on Earth. However, predicting these events has been challenging due to limited understanding of their triggering mechanisms in the Sun's atmosphere.

To address this, the SEC will combine NASA's solar observations with AI and machine learning techniques, offering new insights into the buildup of magnetic energy in active solar regions that lead to eruptions.

The center also plans to compile an extensive dataset of solar activity spanning several 11-year solar cycles, providing researchers with a more comprehensive understanding of flare and CME precursors.

"A major hurdle in understanding solar eruption mechanisms is the limited data on large events like X-class flares," Wang explained. "Building a large, homogeneous dataset of solar activity using advanced machine learning methods allows us to study these major events with unprecedented resolution and cadence, ultimately revealing eruption mechanisms and unlocking better space weather predictions."

In addition to advancing solar weather forecasting, the SEC will establish education and outreach programs to provide research opportunities for students at various levels, from K-12 to graduate studies. The center will partner with local MSIs, such as Kean University and Essex County College, to offer workshops, summer boot camps, and other STEM-focused initiatives.

Related Links
NJIT Center for Solar-Terrestrial Research
Solar Science News at SpaceDaily