It’s one of the biggest puzzles in modern astronomy: Based on multiple observations of stars and galaxies, the universe seems to be flying apart faster than our best models of the cosmos predict it should. Evidence of this conundrum has been accumulating for years, causing some researchers to call it a looming crisis in cosmology.
Now a group of researchers using the Hubble Space Telescope has compiled a massive new dataset, and they’ve found a-million-to-one odds that the discrepancy is a statistical fluke. In other words, it’s looking even more likely that there’s some fundamental ingredient of the cosmos—or some unexpected effect of the known ingredients—that astronomers have yet to pin down.
“The universe seems to throw a lot of surprises at us, and that’s a good thing, because it helps us learn,” says Adam Riess, an astronomer at Johns Hopkins University who led the latest effort to test the anomaly.
The conundrum is known as the Hubble tension, after astronomer Edwin Hubble. In 1929 he observed that the farther a galaxy is from us, the faster it recedes—an observation that helped pave the way toward our current notion of the universe starting with the big bang and expanding ever since.
Researchers have tried to measure the universe’s current rate of expansion in two primary ways: by measuring distances to nearby stars, and by mapping a faint glow dating back to the infant universe. These dual approaches provide a way to test our understanding of the universe across more than 13 billion years of cosmic history. The research has also uncovered some key cosmic ingredients, such as “dark energy,” the mysterious force thought to be driving the universe’s accelerating expansion.
But these two methods disagree on the universe’s current expansion rate by about 8 percent. That difference might not sound like much, but if this discrepancy is real, it means the universe is now expanding faster than even dark energy can explain—implying some breakdown in our accounting of the cosmos.
The researchers’ findings, described in several studies submitted last week to The Astrophysical Journal, use specific types of stars and stellar explosions to measure the distance between us and nearby galaxies. The dataset includes observations of 42 different stellar explosions, more than double the next-biggest analysis of its kind. According to the team’s work, the tension between their new analysis and results from measurements of the early cosmos has reached five sigma, the statistical threshold used in particle physics to confirm the existence of new particles.
Other astronomers still see room for possible errors in the data, which means it’s still possible the Hubble tension is just an artifact.
However, “I do not know how this large of an error is hiding at this point, and if it is, it’s just something no one has suggested,” says team member Dan Scolnic, an astronomer at Duke University. “We’ve checked every idea that’s been presented to us, and nothing’s doing the trick.”
Cosmic microwaves and the distance ladder
The Hubble tension comes from attempts to measure or predict the universe’s current rate of expansion, which is called the Hubble constant. Using it, astronomers can estimate the age of the universe since the big bang.
One way of getting the Hubble constant relies on the cosmic microwave background (CMB), a faint glow that formed when the universe was just 380,000 years old. Telescopes such as the European Space Agency’s Planck observatory have measured the CMB, providing a detailed snapshot of how matter and energy were distributed in the early universe, as well as the physics that governed them.
Using a model that predicts many of the universe’s properties with spectacular success—known as the Lambda Cold Dark Matter model—cosmologists can mathematically fast-forward the infant universe as seen in the CMB and predict what today’s Hubble constant should be. This method predicts that the universe should be expanding at a rate of about 67.36 kilometers per second per megaparsec (a megaparsec equals 3.26 million light-years).
By contrast, other teams measure the Hubble constant by looking at the “local” universe: the more modern stars and galaxies that are relatively close to us. This version of the calculation requires two kinds of data: how quickly a galaxy is receding from us, and how far away that galaxy is. That in turn requires astronomers to develop what’s known as a cosmic distance ladder.
The new studies’ cosmic distance ladder, assembled by Riess’s research group SH0ES, starts with measurements of the distances between us and certain kinds of stars called Cepheid variables. Cepheids are valuable because in essence they act as strobe lights of known wattage: They brighten and dim regularly, and the brighter the Cepheids, the more slowly they pulsate. Using this principle, astronomers can estimate the intrinsic brightnesses of even more distant Cepheids based on their pulsation rates and ultimately calculate the stars’ distances from us.
To extend the ladder even farther, astronomers have added rungs based on stellar explosions called type 1a supernovae. By studying galaxies that host both Cepheids and type 1a supernovae, astronomers can work out the relationship between the supernovae’s brightnesses and their distances. And because type 1a supernovae are much brighter than Cepheids, they can be seen at much greater distances, letting astronomers extend their measurements to galaxies deeper in the cosmos.
Accounting for variation
The trouble is, accurately measuring all of these stars and supernovae is devilishly complicated. Technically speaking, not all Cepheids and type 1a supernovae look exactly the same: Some may have different compositions, different colors, or different types of host galaxies. Astronomers have spent many years figuring out how to account for all this variability—but it’s extremely difficult to know with certainty that some hidden source of error isn’t pushing its thumb on the scales.
To address these concerns, a research team called the Pantheon+ collaboration exhaustively analyzed 1,701 observations of type 1a supernovae collected since 1981. The analysis included efforts to quantify all known uncertainties and sources of bias.
“We care about, like, what the weather and seeing of a telescope was like in November 1991—that is tough,” says Duke University’s Scolnic, who co-leads Pantheon+ with Harvard-Smithsonian Center for Astrophysics researcher Dillon Brout.
The team’s findings fed into the new analysis by Riess and his SH0ES colleagues. After performing a similarly exhaustive cross-check of factors that could affect observations of Cepheids, the team generated its sharpest estimate yet for the Hubble constant: 73.04 kilometers per second per megaparsec, plus or minus 1.04. That’s about 8 percent higher than the value inferred from the Planck observatory’s measurements of the CMB.
The team also went to great lengths to test outside scientists’ ideas for why its Hubble constant estimate is higher than Planck’s. In all, the researchers ran 67 variants of their analysis—many of which made the tension worse.
“We’ve listened, I think, carefully to a lot of concerns and issues,” Riess says. “This isn’t just a ‘shazam’ … We’ve done a lot of deep dives down rabbit holes.”
The unknown universe
In recent years, though, the University of Chicago’s Wendy Freedman has been working on an estimate that doesn’t rely on pulsing stars. Instead, she uses a specific group of red giant stars, which also act like light bulbs of known wattage. Building off of these alternate “standard candles,” or objects with known intrinsic brightnesses, Freedman’s independent estimate of the Hubble constant is 69.8 kilometers per second per megaparsec—in the middle of the other two measurements.
Despite the team’s careful work, Freedman says that undiscovered errors could still be affecting the analysis, perhaps creating an illusory tension. She adds that some sources of uncertainty are also unavoidable. For one, there are only three galaxies close enough to the Milky Way whose distances we can measure directly, and the base of the cosmic distance ladder rests on this trio.
“Three’s a small number, but that’s what nature has given us,” Freedman says.
The Pantheon+ and SH0ES teams have taken a long look at Freedman and others’ results, and some of their various analyses examine what happens if Freedman’s preferred stars are added to the cosmic distance ladder, along with Cepheids and type 1a supernovae. According to their work, including these extra stars slightly lowers the estimate for the Hubble constant—but it does not eliminate the tension.
And if the Hubble tension really does reflect our physical reality, then explaining it will probably require adding another item to our list of the universe’s fundamental ingredients.
One of the leading theoretical contenders, called early dark energy, proposes that about 50,000 years after the big bang, there was a brief flare-up of dark energy. In principle, a short blip of extra dark energy could alter the expansion of the early universe enough to resolve the Hubble tension without messing too much with the standard model of cosmology.
But in the process, cosmologists’ estimates for the age of the universe would fall from the current 13.8 billion years to about 13 billion years.
“There’s a lot of questions about why you have to introduce this one new thing that just shows up and disappears–that feels a little funny,” says Mike Boylan-Kolchin, an astrophysicist at the University of Texas at Austin. “But we’re in a place where, if these things are really that discrepant, maybe we have to start looking in the funny corners of the universe.”
For now, there’s no slam-dunk evidence for early dark energy, though some hints have reared their heads. In September the Atacama Cosmology Telescope, a facility in Chile that measures the cosmic microwave background, claimed that a model including early dark energy fits its data better than the standard cosmological model does. The Planck telescope’s data disagree, so future observations will be required to get to the bottom of the mystery.
Other observatories should also help clarify the Hubble tension. ESA’s Gaia satellite, for example, has been mapping the Milky Way since 2014, generating increasingly precise distance estimates between us and many of our galaxy’s stars, including Cepheids. And the upcoming James Webb Space Telescope—due to launch later this month—should help astronomers double-check Hubble’s measurements of certain stars.
“We’re working at the edge of what is possible,” Freedman says. “We will get to the bottom of this.”