AI Uncovers Stunning Secrets Hidden in the Sun’s Atmosphere
Researchers at the University of Hawaiʻi are utilizing the world’s largest solar telescope, the Daniel K. Inouye Solar Telescope, alongside advanced AI to revolutionize solar research, SciTech Daily revealed yesterday.
The “SPIn4D” project focuses on rapidly analyzing vast datasets to improve predictions of solar phenomena, such as solar storms, which can impact satellites and power grids.
Led by postdoctoral researcher Kai Yang, the team employs deep learning models to process data from the telescope, which generates up to 40 terabytes daily. This AI-driven approach allows for near real-time visualization of the solar atmosphere, significantly speeding up data analysis.
The researchers have created an extensive dataset using over 10 million CPU hours on the NSF’s Cheyenne supercomputer, resulting in 120 terabytes of simulated solar observations. A 13-terabyte subset is already publicly available, with plans to release fully trained AI models as community tools for further solar analysis.
