A team led by the University of California, Santa Cruz has identified a rare stellar explosion, SN 2023zkd, using artificial intelligence to detect unusual supernovae in real time. The discovery was made in July 2023 with a new AI algorithm that flagged the event quickly enough for astronomers to begin immediate follow-up observations.
Ryan Foley, associate professor of astronomy and astrophysics at UC Santa Cruz, said, “Something exactly like this supernova has not been seen before, so it might be very rare. Humans are reasonably good at finding things that ‘aren’t like the others,’ but the algorithm can flag things earlier than a human may notice. This is critical for these time-sensitive observations.”
The Young Supernova Experiment (YSE), operated by Foley’s team at UC Santa Cruz, surveys about 4% of the night sky every three days and has discovered thousands of cosmic explosions shortly after they occur. The YSE uses telescopes at Haleakalāa Observatory in Hawaiʻi as part of its operations.
Researchers believe SN 2023zkd resulted from a collision between a massive star and a black hole companion. As energy dissipated from their orbit, their separation decreased until gravitational stress triggered the supernova as the star was partially consumed by the black hole.
Lead author Alexander Gagliano, a fellow at the NSF Institute for Artificial Intelligence and Fundamental Interactions, stated, “Our analysis shows that the blast was sparked by a catastrophic encounter with a black hole companion, and is the strongest evidence to date that such close interactions can actually detonate a star.”
An alternative explanation considered by scientists is that the black hole completely tore apart the star before it could explode on its own. In both scenarios, only one heavier black hole remains after the event.
SN 2023zkd occurred approximately 730 million light-years away. While initially appearing as an ordinary supernova with one burst of light, further monitoring revealed an unexpected second brightening months later. Archival data indicated gradual brightening over four years before the explosion—an uncommon pattern among supernovae.
Analysis conducted partly at UC Santa Cruz showed that material shed by the star prior to its death shaped how light from the explosion evolved over time. Early brightening resulted from interaction with low-density gas; later brightening came from sustained collision with denser material likely arranged in a disk around the system. These findings suggest strong gravitational influence from a compact companion such as a black hole.
Foley described how he worked closely with Gagliano on interpreting spectra from SN 2023zkd: “Our team also built the software platform that we use to consolidate data and manage observations. The AI tools used for this study are integrated into this software ecosystem,” Foley said. “Similarly, our research collaboration brings together the variety of expertise necessary to make these discoveries.”
Enrico Ramirez-Ruiz of UC Santa Cruz led theoretical work on this project; V. Ashley Villar from Harvard contributed AI expertise; other partners included researchers from Center for Astrophysics | Harvard & Smithsonian and MIT as part of YSE.
Funding came from several sources including National Science Foundation (NSF), NASA, Moore Foundation, and Packard Foundation. Several students involved were or are NSF graduate research fellows.
Despite these achievements, Foley expressed concern about uncertain funding: “The uncertainty means we are shrinking,” he said, “reducing the number of students who are admitted to our graduate program—many of them being forced out of the field or to take jobs outside the U.S.”
Foley added that while predicting future applications is difficult, similar AI techniques could be useful beyond astronomy: “You can easily imagine similar techniques being used to screen for diseases, focus attention for terrorist attacks, treat mental health issues early, and detect financial fraud,” he explained. “Anywhere real-time detection of anomalies could be useful, these techniques will likely eventually play a role.”
