Has the Universe’s Expansion Started Slowing Down?

For nearly three decades, astronomers have been convinced that our universe is accelerating – expanding at an ever-faster rate due to a mysterious force known as dark energy.

This startling cosmic acceleration, discovered in 1998 by observing distant stellar explosions called Type Ia supernovae, upended expectations and earned the 2011 Nobel Prize in Physics. But new findings published in Monthly Notices of the Royal Astronomical Society (MNRAS, 2025) suggest a surprising twist: the universe’s expansion may already be slowing down. If confirmed, this would mark the biggest shift in cosmology since dark energy’s discovery.

In this post, we examine the history of cosmic expansion and acceleration, the role of Type Ia supernovae as standard candles of the cosmos, and the implications of this new research. We’ll break down the difference between the classic Lambda-CDM model (with constant dark energy) and an evolving dark energy model (w0–wa CDM) that might better explain the data. Finally, we’ll look at the possible fates of the universe under these models – from eternal acceleration to a decelerating, perhaps even recollapsing, cosmos.

Discovery of an Accelerating Universe: A Brief History

In the early 20th century, Edwin Hubble discovered that galaxies are receding from one another, implying that the universe is expanding. For decades, scientists assumed this expansion should be slowing down under gravity’s pull. After the Big Bang, the expansion indeed decelerated as gravity reined it in.

However, in 1998, two independent teams studying distant Type Ia supernovae found something astounding: those far-off explosions were dimmer than expected, suggesting they were farther away than our models predicted. The only explanation was that the expansion of the universe had sped up over the last few billion years.

This was a shock. Gravity, produced by all matter and energy, always pulls inward – it should slow cosmic expansion, not reverse it. To account for an accelerating universe, cosmologists resurrected an idea Einstein once proposed and then discarded: the cosmological constant (Λ).

In modern terms, this constant is part of the “Lambda-CDM” model – our standard model of cosmology. It corresponds to dark energy, an enigmatic form of energy that makes up roughly 70% of the universe and acts like a repulsive force or “anti-gravity” on large scales. In ΛCDM, dark energy (Λ) is a fixed, constant property of space that began dominating the universe a few billion years ago, driving a renewed acceleration.

Under the Lambda-CDM model, the future seemed clear: the universe would continue to expand at an ever-faster rate. Galaxies not gravitationally bound to us would rush away until they vanish from view. Over trillions of years, stars would burn out, matter would decay, and the cosmos would drift into a cold “big freeze” or heat death – dark, empty, and eternal. This grim fate, though distant, was accepted as the most likely outcome given an endlessly accelerating universe.

Type Ia Supernovae – Cosmic Mileposts and “Standard Candles”

Type Ia supernovae (SNe Ia) are spectacular exploding stars that astronomers use as standard candles to measure cosmic distances. They occur when a white dwarf star in a binary system detonates in a runaway fusion reaction. Crucially, these explosions reach nearly the same peak brightness every time. By comparing how bright a Type Ia supernova appears from Earth to how bright it truly is, astronomers can figure out its distance – much like judging a car’s distance by the dimness of its headlights. This makes SNe Ia excellent cosmic mileposts for mapping the expansion of the universe.

However, standard candles must truly be standard. Astronomers calibrate Type Ia supernova luminosities using methods that account for slight variations (for example, brighter explosions fade differently than dimmer ones). The 1998 discovery of acceleration relied on the assumption that these calibrated supernovae behave the same everywhere in the universe. But what if there’s an overlooked difference in these stellar explosions with a hidden bias affecting their brightness? A new study from Yonsei University suggests exactly that.

The Progenitor Age Bias: Young vs. Old Stars

Researchers led by Junhyuk Son and Young-Wook Lee examined over 300 galaxies that hosted Type Ia supernovae. They measured the ages of the star populations in those host galaxies. The finding was striking: even after standard calibration, supernovae born from younger stellar populations tended to appear slightly fainter, while those from older stars appeared brighter. In other words, the true brightness of a Type Ia supernova isn’t one-size-fits-all – it can depend on the age of the stars that produced it.

This effect is called a progenitor age bias. Imagine two standard candles with the same true luminosity: one comes from an old galaxy (older stars), and one from a young galaxy. The study finds that the one from the young galaxy looks dimmer to us, after accounting for all the usual factors. If left uncorrected, this bias would lead us to overestimate the supernova’s distance. And if many distant supernovae happen to come from younger stellar systems (which makes sense – the farther away we look, the earlier the cosmic time we are seeing, so stars were generally younger in the past), they would all seem a bit too dim. That dimming was previously interpreted as evidence that expansion has accelerated to carry them so far away. However, part of that dimming may be intrinsic to the stellar age, rather than solely due to cosmic speed-up.

The Yonsei team found this age-related brightness bias with extremely high confidence (99.999%), indicating it’s highly unlikely to be a fluke. It also appears that a commonly used fix – a correction for the host galaxy’s mass – did not account for the age effect. (Galaxies with higher mass tend to host slightly different supernova properties, so scientists often correct for that, but here age plays a distinct role.) As the researchers put it, “this systematic bias is largely uncorrected by the commonly used mass-step correction, as progenitor age and host galaxy mass evolve very differently with redshift”.

A New Picture: Correcting the Bias Reveals a Slower Expansion

So, what happens if we correct the supernova data for this age bias?

The research team applied an age-based correction to the supernova distances as a function of redshift (which serves as a proxy for looking back in time – high redshift means an earlier universe). Once they did that, the results were astounding: the corrected supernova data no longer fit the standard ΛCDM model of a universe that is constantly accelerating.

In the classic Lambda-CDM model, dark energy is a constant that causes the universe to continually accelerate. But the bias-corrected supernova distances started to deviate from ΛCDM’s predictions. Instead, they lined up much better with an alternative model – one in which dark energy’s strength changes over time. Specifically, the supernova observations fell in line with a model supported by the Dark Energy Spectroscopic Instrument (DESI) project, which uses entirely different methods (baryon acoustic oscillations and the cosmic microwave background) to probe cosmic expansion.

According to the Yonsei analysis, once the age bias is removed, the supernova evidence no longer shows an accelerating expansion at present. In fact, the data suggest the universe’s expansion has already transitioned into a decelerating phase today. The authors combined the corrected supernova data with results from baryonic acoustic oscillations (BAO), which are ripples in the matter distribution of the early universe, and measurements of the cosmic microwave background (CMB) radiation.

Together, these three probes paint a consistent picture that rules out the standard ΛCDM cosmology with overwhelming significance. In statistical terms, the tension with ΛCDM exceeds 9 sigma (>9σ), far beyond the usual threshold to claim a major discovery.

“Most surprising of all,” the researchers noted, “the combined analysis indicates that the universe is not accelerating today as previously thought, but has already transitioned into a state of decelerated expansion”.

In other words, the cosmic acceleration might have already stopped; the universe could be slowing its expansion rate in the present epoch, rather than continuing to speed up.

It’s important to stress that the universe is still expanding – it hasn’t stopped growing larger. But if these results hold, the expansion rate is no longer increasing; instead, it’s in the process of easing off. This nuance is crucial. Earlier analyses (like those from the DESI project) hinted that while acceleration would eventually slow in the far future, the universe was still accelerating at present. The new age-corrected supernova data suggest the slowdown may have already begun, in line with what BAO and CMB measurements were quietly indicating as well.

ΛCDM vs. w0–wa CDM: Constant vs. Changing Dark Energy

What kind of cosmological model fits a slowing expansion? The data support a scenario in which dark energy is not constant but instead evolves over time. In cosmological jargon, the alternative model is described by parameters (w0, wa) and is often called the w0–wa CDM model. These parameters allow the “equation of state” of dark energy (the relationship between its pressure and density, often denoted by w) to change as the universe expands.

In the Lambda-CDM model, dark energy has w = –1 at all times – that’s the hallmark of a true cosmological constant. In a w0–wa model, w0 represents the current value of the dark energy equation-of-state (which might be close to –1), and wa represents how much w changes with time (with wa = 0 recovering the constant case). If observations prefer a w0–wa model, it means dark energy today might not be exactly the same “strength” or effect as it was in the past. It could be, for instance, that dark energy was stronger in earlier epochs and is weakening over time, causing the acceleration to wane. This is exactly what the DESI project’s recent findings were pointing toward.

The Yonsei team’s corrected supernova data align with the DESI BAO+CMB results, suggesting that the impact of dark energy evolves rather than remaining constant. Notably, if dark energy is weakening now, it would mean the expansion is decelerating now (no current acceleration), a striking departure from the consensus of the last 25 years.

On the other hand, if dark energy had been stronger in the past, it might have caused a period of even more intense acceleration in earlier times (some models even allow for a transient “phantom” dark energy phase, where expansion could be extremely fast). The history of the universe might then include a rise and fall of the influence of dark energy.

It’s worth noting that an evolving dark energy (sometimes called “quintessence” in theories) has been considered by cosmologists for years. The w0–wa parameterization is a common method for testing it. Until recently, ΛCDM with w = –1 fit observations well enough that no variation was confirmed. These new data hint that the true nature of dark energy could be more complex – and that our standard model might need revision.

Possible Fates of the Universe Under Different Models

What does all this mean for the future of the universe? The fate of the cosmos is tied to whether the expansion continues to accelerate, slows down, or reverses its course. Here are several scenarios, each tied to a different cosmic model:

Big Freeze (Eternal Acceleration): If dark energy is a true constant (w = –1 forever, as in ΛCDM), the expansion will continue to accelerate without end. Galaxies outside our local cluster will move away from us at an increasingly faster rate. Trillions of years from now, nearly all other galaxies will slip beyond our visible horizon. Stars will burn out, black holes will evaporate, and the universe heads toward a heat death – essentially an eternal, ever-colder big freeze where nothing new forms. This was long the most accepted fate given an accelerating universe.

Coasting or Slowing Expansion: If dark energy weakens over time (w > –1 in the future), cosmic acceleration could gradually slow and possibly come to a halt. The universe would continue expanding, but at a steady or decelerating rate – sometimes called a “coasting” universe. In this scenario, there is no big crunch, but also no runaway acceleration. Galaxies would continue to drift apart, but not at an exponential rate. The expansion might approach a constant rate or slow to essentially zero expansion rate far in the future. The age-biased supernova results, combined with BAO/CMB, suggest that we might already be entering a phase of decelerated expansion. If this holds, the long-term outlook could be more benign than the big freeze – perhaps galaxies remain visible for longer – but eventually, star formation will cease, and the universe will still grow dark over unimaginably long timescales.

Big Crunch (Reversal and Collapse): An even more dramatic outcome would occur if dark energy not only weakens but becomes negligible or even turns attractive (for example, if w rises above –1 to 0 or positive in the far future, meaning gravity’s pull isn’t being counteracted at all). In that case, gravity from matter could slowly overcome expansion. The universe might stop expanding and begin to contract. Over time, galaxies would start moving closer, eventually crashing together as the universe collapses in a fiery “big crunch” – essentially the Big Bang in reverse.

Is this likely? It’s considered unlikely with the matter we have (since current data suggests the universe’s overall density is not enough to recollapse on its own). But if dark energy’s effect diminished drastically or flipped sign, a recollapse isn’t ruled out by theory. Some optimists speculated that a dynamic dark energy could allow for a big crunch instead of heat death, but most scientists urge caution – we have no evidence yet that dark energy will disappear entirely or become gravitationally attractive.

Big Rip (Runaway Acceleration): On the opposite extreme, if dark energy were to grow stronger (for instance, w < –1, called “phantom energy”), the expansion could not only continue accelerating but accelerate to catastrophic levels. In a far-future big rip scenario, galaxies, stars, planets, and eventually atoms themselves could be torn apart by the accelerating expansion of space.

For a time, early analyses of some data suggested phantom energy as a possibility, but it’s no longer a favored scenario. Interestingly, the DESI data that hinted at evolving dark energy also suggested that long ago (in an earlier epoch) dark energy might have behaved phantom-like. If true (and that’s very speculative), it’s a reminder that dark energy’s behavior could swing in unexpected ways. However, most models that allow a big rip are considered speculative, and current observations (including the new supernova findings) lean toward weaker dark energy, not stronger.

It’s important to remember that no matter which scenario, the universe is not likely to provide a cozy, eternal haven. Even if cosmic expansion slows or stops, stars will still age and die, galaxies will exhaust their gas for making new stars, and entropy will increase. As cosmologist Katie Mack wryly noted, even a halted expansion doesn’t avoid an eventual apocalypse of some form – the arrow of time guarantees a slide toward disorder. The real question is how the journey unfolds: a slow fade-out in an ever-expanding space, or a more dramatic end, such as re-collapse or ripping apart.

Caution and Curiosity: Awaiting Further Evidence

While these new findings are exciting and thought-provoking, they are still preliminary. Extraordinary claims – like “the universe’s expansion is no longer accelerating” – require extraordinary evidence. The study by Son, Lee, Chung, et al. has opened a critical conversation, but the cosmology community will now scrutinize these results with a fine-tooth comb. Other research teams will need to verify the progenitor age effect and see if they too find the same trend. Independent data will be crucial to confirm (or refute) that the acceleration era is indeed ending.

The Yonsei group is already working on what they call an “evolution-free test” to bolster their case. This test involves using a subset of supernovae that originate from young, near-identical galaxies (whose star populations are all approximately the same age) across a wide range of distances. By focusing only on supernovae in these uniformly young environments, they aim to eliminate the age bias from the outset and determine whether the cosmic acceleration signal is present or not. Early results from this approach support their main conclusion so far.

The next few years offer a fantastic opportunity to gather more evidence. The brand-new Vera C. Rubin Observatory in Chile has just begun its sky survey operations. Rubin’s 8.4-meter telescope and gigantic camera will discover tens of thousands of new supernovae and their host galaxies every year. The Yonsei researchers estimate that within five years, data on more than 20,000 supernova host galaxies will allow an extremely precise check of their findings. With that many supernovae, astronomers can directly measure how supernova brightness correlates with host galaxy age across the universe, and decisively test for any bias.

The Vera C. Rubin Observatory in Chile’s Andes mountains is poised to revolutionize supernova discovery. Beginning operations in 2025, it will discover thousands of new Type Ia supernovae, enabling astronomers to measure the ages of their host galaxies. This flood of data will help test whether the universe’s expansion is truly slowing down, as the latest study suggests.

Other projects are also contributing: the Dark Energy Spectroscopic Instrument (DESI) continues to map millions of galaxies to refine BAO measurements, and new space telescopes, such as Euclid (launched in 2023) and the Nancy Grace Roman Space Telescope (planned launch in the mid-2020s), will provide independent data on cosmic expansion. These will all help cross-check the supernova results with those from lensing, galaxy clustering, and other sources.

Cosmologist and science communicator Dr. Katie Mack cautions against jumping to conclusions about the universe’s ultimate fate based on these early results. Dynamic dark energy models are tricky – even if we confirm that dark energy is changing, we won’t immediately know how it will behave in the future. As she points out, with a true cosmological constant, we have a clear (if grim) prediction for the universe’s ultimate fate, but with evolving dark energy, the uncertainty is significantly greater. The recent findings could signal the first big paradigm shift in cosmology since the late 1990s, but science demands verification.

A New Chapter in Cosmology?

If the universe has indeed begun to slow its expansion, it would significantly reshape our understanding of cosmic history and fate. It may help resolve some growing tensions in cosmological data (for example, certain measurements of the universe’s expansion rate and geometry have been puzzlingly inconsistent with ΛCDM, a phenomenon known as the Hubble tension).

A slowing expansion could bring certain datasets into better alignment. It would also raise new questions: What is dark energy, truly? Why would it weaken over time – what physics causes that? Did the acceleration era last only a few billion years and end in our cosmic “recent” epoch? Is it possible that we live at the transition between acceleration and deceleration?

For now, dark energy remains as enigmatic as ever – perhaps even more so. The one-two punch of the DESI project’s hints and the new supernova analysis suggests dark energy might not be a simple, static thing. The coming years of data will be crucial. As scientists, we must be prepared for surprises: the universe often defies our expectations, and that’s what makes exploring it so exciting.

Bottom line: A new study finds that correcting for the age of stars that gave rise to distant supernovae makes the universe’s expansion appear to be slowing down, rather than speeding up. This aligns supernova data with other cosmological clues from galaxies and the early universe. If confirmed, it would overturn the longstanding idea of ongoing acceleration and suggest dark energy is evolving over time. Such a claim needs further testing, but it’s sparking a wave of curiosity and cautious excitement in the cosmology community.

Are we witnessing the beginnings of a paradigm shift about the fate of the cosmos? Future observations (and plenty of hard work) will tell. Until then, the expansion of the universe – once thought to be a settled story of runaway acceleration – has become the center of a fascinating, evolving mystery once again.

Explore more

Check out these great videos that really get at the scale of the expansion of the universe