AI Unearths New Supernova Type from Star-Black Hole Collision
A groundbreaking astronomical discovery, made possible by an artificial intelligence system akin to a music recommendation engine, has unveiled what may be an entirely new class of supernova. This celestial explosion, observed in July 2023 and designated SN 2023zkd, appears to be the dramatic demise of a colossal star potentially attempting to engulf a nearby black hole.
The supernova was initially detected by the Zwicky Transient Facility, a comprehensive sky survey operating from the Palomar Observatory in California. Crucially, this detection wasn’t accidental. The Zwicky facility was specifically directed to the event by an advanced algorithm named Lightcurve Anomaly Identification and Similarity Search (LAISS). This AI, notably modeled on the same principles as Spotify’s music recommendation system, excels at sifting through vast amounts of astronomical data to pinpoint unusual activity in the night sky, much like Spotify suggests new songs based on a user’s listening habits.
Early detection of supernovae is paramount for scientists aiming to understand their complete lifecycle—from their explosive genesis to their eventual fading. In this instance, LAISS flagged unusual brightenings in the star’s light signature months before the actual explosion, according to study co-lead authors Alex Gagliano, a postdoctoral researcher at the Institute For AI and Fundamental Interactions, and Ashley Villar, a supernova researcher and assistant professor at the Harvard-Smithsonian Center for Astrophysics. This timely alert allowed a network of major observatories to pivot their instruments towards the nascent event, capturing a rich tapestry of data across a wide spectrum of light wavelengths.
Astronomers have proposed several theories for what unfolded, but the most compelling evidence suggests SN 2023zkd resulted from a massive star orbiting a black hole some 730 million light-years from Earth. As these two celestial bodies spiraled closer, the star was subjected to immense gravitational forces. The prevailing hypothesis posits that the star, under this extreme stress, effectively tried to absorb the black hole, leading to its catastrophic explosion. While another possibility, known as “spaghettification”—where the black hole’s immense gravity tears the star apart—was considered, the collected data does not support this scenario as strongly.
Further analysis of the massive star’s chemical composition revealed that it had not shed all of its outer layers before its demise. This finding hints at a more complex and “messier” interplay in binary star systems than previously assumed, as Gagliano noted. Understanding how companion interactions influence the explosions of massive stars remains a significant challenge for current astronomical models. While acknowledging the rarity of such observed events, Gagliano emphasized that the data strongly indicates the involvement of a binary system, making a star-black hole interaction highly probable.
The power of LAISS lies in its ability to identify statistical outliers by comparing the light properties of celestial objects against an extensive reference database of known phenomena. When LAISS identifies a promising candidate, an automated bot alerts astronomers via a dedicated messaging channel, facilitating rapid investigation of the most compelling and unusual discoveries.
The light pattern of SN 2023zkd itself proved highly unusual. Initially, it brightened like a typical supernova before dimming. However, it then brightened again, drawing significant attention. Archival data further revealed peculiar pre-explosion behavior: the star, which had maintained a consistent brightness for an extended period, had gradually grown brighter over the four years preceding its explosion. Scientists theorize this pre-explosion brightening originated from excess material the star was shedding. The subsequent peaks in brightness after the explosion likely occurred as the supernova’s shockwave successively collided with lower-density gas and then a dense cloud of dust in the surrounding environment.
The presence of a black hole was inferred not only from the intricate structure of the surrounding gas and dust but also from the peculiar stellar brightening observed in the years leading up to the explosion. Without the assistance of LAISS, these crucial early signatures of the surrounding disk and the black hole companion would likely have been missed. AI systems like LAISS are increasingly indispensable, enabling astronomers to consistently uncover rare explosions and trace their origins without relying on chance. The findings of this groundbreaking research were formally published on August 13 in The Astrophysical Journal.