Is the universe really expanding faster and faster? A team led by Lee Young-uk, a professor in the Department of Astronomy and Space Science and the Center for Galaxy Evolution at Yonsei University, has published findings that directly challenge the "accelerating expansion of the universe" theory, which has been accepted as the standard for 25 years. The study was published on the 16th in MNRAS (Monthly Notices of the Royal Astronomical Society), an international journal issued by the Royal Astronomical Society in the United Kingdom.
In 1998, astronomers analyzed the brightness of Type Ia supernovae exploding in distant galaxies and concluded that the universe is expanding faster and faster. The study was accepted as evidence that the universe is being pushed apart by an unknown force called dark energy, and it became the decisive basis for the 2011 Nobel Prize in physics.
But the Yonsei University team reanalyzed recent observations and presented the conclusion that "the universe is still expanding, but no longer accelerating. It has already entered a decelerating phase."
To find out "how fast the universe is expanding," astronomers have long used Type Ia supernovae as a kind of "standard lamp (standard candle)." Because these supernovae emit nearly the same brightness when they explode, one can calculate "how far away the star is" based on how dim it appears. In other words, brightness serves as a yardstick for measuring distance.
However, the Yonsei University team found that this brightness is less uniform than expected. Brightness varies depending on the "age of the star" that produced the supernova. Supernovae from younger stars appeared slightly dimmer than expected, while those from older stars appeared slightly brighter. This difference had not been properly accounted for in previous studies of cosmic expansion.
Analyzing about 300 supernovae and their host galaxies, the team showed that this phenomenon is a statistically robust result at nearly one in 10 billion (5.5 sigma). In other words, the chance it is a coincidence is effectively "0." This means that the reason supernovae in distant galaxies appear dim is not only because they are far away due to the universe's expansion. Astrophysical factors such as the age of the stars themselves were also influencing the brightness.
The team corrected the supernova data to reflect this effect and recalculated. Then the previously accepted "cosmological constant model," in which dark energy exists at a constant level, no longer fit. Instead, the model of "dark energy weakening over time," proposed recently by the DESI (Dark Energy Spectroscopic Instrument) project, fit much better.
In other words, instead of the prevailing view that "dark energy is constantly pushing the universe outward," a new interpretation becomes possible: "dark energy is gradually weakening, so the universe's expansion rate is slowing."
The Yonsei University team compared the "supernova brightness correction results" with other cosmic observations. To do this, they used baryon acoustic oscillation (BAO) and cosmic microwave background (CMB) data. BAO is the imprint of oscillations in matter from when the universe was very young. By looking at the size of this imprint, one can know how fast the universe expanded in the past. The CMB is the light left over just after the Big Bang. Analyzing its pattern reveals the early state of the universe. Put simply, supernovae show "the universe today," while BAO and CMB show "the universe in the past."
The team analyzed the three datasets together. As a result, the "standard cosmological model" used like a textbook (the model that the universe is accelerating at a constant rate) was shown to be inconsistent by a statistical difference of 9 sigma. "9 sigma" means the probability that this result is a coincidence is nearly zero. The criterion scientists accept as "almost certain" is 5 sigma. In other words, the team's conclusion that "cosmic expansion is no longer accelerating; it has already entered a decelerating phase" is statistically very robust.
To check whether other errors might still be hidden in the supernova data, the team is also conducting an additional experiment called an "evolution-free" cosmology test. Simply put, it reanalyzes only young galaxies whose stellar ages are similar. This completely removes brightness changes due to stellar age (evolution effects). The initial results of the first phase reportedly support the Yonsei University team's claim that "cosmic expansion is already decelerating."
Yonsei University doctoral candidate Son Jun-hyeok and research professor Jeong Cheol, who played key roles in this study, said, "If we can measure the ages of about 20,000 new supernova host galaxies that the LSST survey telescope will discover within the next five years, supernova-based cosmology will become far more precise than it is now."
The Yonsei University team said that if these findings are confirmed through further verification, it could trigger a fundamental paradigm shift in cosmology for the first time in 27 years since the discovery of dark energy in 1998. They said it could open a new chapter in studies to reveal the nature of dark energy, the Hubble tension, and the history and fate of the universe's expansion.
This research was supported by the University-Centered Research Institute Program and the Mid-Career Researcher Program of the National Research Foundation of Korea (NRF).
References
MNRAS (2025), DOI: https://doi.org/10.1093/mnras/staf1685