
For the first time, the collision of two exoplanets was documented in real-time by analyzing significant fluctuations in the light emitted by their host star—an event astronomers have hoped to witness for years.
Anastasios Tzanidakis from the University of Washington was reviewing archived telescope data from 2020 when he stumbled upon a seemingly unremarkable star exhibiting highly unusual behavior. This object, cataloged as Gaia20ehk, is situated approximately 11,000 light-years from Earth, near the constellation of Leo.
It was a stable main-sequence star, quite similar to our own Sun, which would typically suggest its light output should be constant and predictable. However, the data revealed that the star had begun to flicker unpredictably.
Tzanidakis detailed that the star’s light curve remained flat and steady until three substantial dips in its brightness were observed starting in 2016. The situation became chaotic around 2021, when, in his words, the star “went completely bonkers.” The researcher stressed the exceptional nature of this occurrence, noting that Sun-like stars simply do not behave this way, prompting the team to question the true source of what was happening.
Upon thorough examination, the reason for this flickering had nothing to do with the star itself. The cause was a massive influx of rock and dust that seemed to emerge from nowhere, orbiting within the system, passing in front of the distant star, and partially and unevenly obscuring the light reaching our instruments. The most plausible origin for such a huge amount of debris turned out to be an even more astonishing event: the catastrophic impact of two planets.
Tzanidakis himself expressed amazement at the ability of various telescopes to record this collision at the precise moment it unfolded, mentioning that records of planet-on-planet impacts of any kind are scarce, and none resemble the collision believed to have formed the Earth-Moon system. The opportunity to observe such moments elsewhere in the galaxy, he added, would provide invaluable insights into the formation of our own world. A detailed analysis of the star was published in The Astrophysical Journal Letters.
The process of planet formation involves the gravitational accretion of material—be it dust, gas, ice, or rocky debris—into an orbit around a newborn star. Solar systems in their nascent stages are turbulent environments where planets frequently collide and either shatter or are ejected into space. It is through this violent process, lasting roughly one hundred million years, that systems like ours eventually shed excess planets and settle into a state of equilibrium.
Despite the presumed frequency of these collisions, witnessing one firsthand in a distant solar system requires an uncommon blend of patience and luck. Planetary orbits must align such that the bodies pass precisely between the star and us, causing the resulting debris to partially block the stellar light. The signature flickering created by this event also persists for several years.
The study’s lead author, James Davenport of the University of Washington, highlighted Tzanidakis’s working style, which leverages archival data collected over decades to uncover slow-developing phenomena—astronomical stories unfolding over a decade-long timescale. Davenport pointed out that very few researchers search for events in this manner, suggesting that many potential discoveries are currently within reach.
As the principal investigator, Tzanidakis had focused his research on extreme stellar variability over time. In a prior study, he had already identified a binary star system obscured by a large dust cloud that caused a seven-year-long eclipse. However, the behavior of Gaia20ehk presented an unprecedented puzzle. The specific nature of the star’s fluctuations, marked by brief dips in brightness followed by a chaotic period, had never been observed before. The team was stumped until Davenport suggested utilizing data from a different telescope to search for infrared light, rather than visible light.
The comparison proved highly illuminating. Tzanidakis observed that the infrared light curve was the precise inverse of the visible light curve. While the visible light began to flicker and dim, the infrared emission experienced a sudden surge. This evidence indicated that the matter obscuring the star was intensely hot, glowing brightly in the infrared spectrum. A catastrophic planet-planet collision could certainly generate sufficient heat to account for this energy output. Furthermore, the specific type of collision could also explain the initial brightness dips that were detected.
Tzanidakis hypothesized that this initial behavior resulted from the two planets slowly spiraling closer together. The early phase involves a series of glancing blows that do not generate much infrared energy. Later, the final, devastating impact occurs, after which the associated infrared radiation spikes dramatically.
There are also suggestions that this impact bears a striking resemblance to the event estimated to have created the Earth and Moon about four and a half billion years ago. The resulting debris cloud orbits Gaia20ehk at a distance of approximately one astronomical unit, matching the separation between the Earth and the Sun. At this range, material could eventually cool enough to condense and give rise to something akin to our Earth-Moon system. Both Tzanidakis and Davenport acknowledge they won’t be able to confirm this until the dust settles in that system—a process that may take anywhere from several years to millions of years.
In the meantime, this discovery serves as a call to action to detect more collisions of this nature. The powerful Simonyi Survey Telescope, housed at the Vera C. Rubin Observatory and funded by the National Science Foundation and the U.S. Department of Energy, is poised to be the ideal instrument for this task when its Legacy Survey of Space and Time begins later this year. Preliminary projections by Davenport estimate that Rubin will be capable of discovering up to a hundred new collisions over the next decade. This vast influx of data could ultimately help narrow the search for habitable worlds beyond our Solar System.
Davenport reflected on the rarity of the event that led to the existence of Earth and the Moon—a question he considers fundamental to astrobiology. The Moon appears to be one of those magical components that renders Earth habitable, by helping shield it from some asteroids, driving the ocean tides, creating climatic variations that allow for global mixing of chemical and biological processes, and potentially even influencing plate tectonic activity. The frequency of these dynamic processes is currently unknown, but documenting more impacts like the one currently observed will begin to provide answers to that question.