Astronomers cannot agree on how fast the universe is expanding

This suggests cosmology might be wrong about something fundamental

IT IS ONE of the biggest mysteries in cosmology—and getting bigger all the time. Ever since Edwin Hubble, an American astronomer, published observations of distant galaxies in 1929, scientists have known that the universe is expanding. For almost 30 years they have known that the expansion is accelerating (that discovery, made in 1998, was honoured with a Nobel prize in 2011). What they cannot agree on, though, is how fast it is currently growing.

Photograph: NASA, ESA, CSA, and J. Lee (NOIRlabs)

That present-day rate of expansion is known as the Hubble constant. Measure it one way, and it comes to around 73 kilometres per second per megaparsec (km/s/mpc; a megaparsec is the distance travelled by light in about 3.3m years, and a value of 73 means that objects 1mpc away recede from an observer at 73 kilometres per second). But measure it another way and the answer is closer to 67km/s/mpc.

That cosmologists cannot agree on one of the most elementary facts about the universe is striking enough. But that uncertainty produces others, too: it makes it impossible to calculate an exact age for the universe, for one thing, or to be certain of its exact size. And the discrepancy refuses to go away, no matter how many times astronomers re-check their measurements, upgrade their instruments, or think of new ways to attack the problem.

The Hubble tension, as the discrepancy between the two sets of measurements is known, “has got stronger every year for the past decade”, says Dan Scolnic, an astronomer at Duke University, in North Carolina. Some astronomers think one set of measurements or the other will turn out to be wrong. Others believe that the tension is a hint of deeper problems with the scientific description of the universe, known as the standard model of cosmology.

There are, broadly speaking, two ways to work out the Hubble constant. One involves measuring the modern universe directly, working out how far away distant galaxies are and how quickly they are receding. It is this technique that gives the higher value of 73. The second is to look at the cosmic microwave background radiation (CMB), an aftershock of the Big Bang. The CMB reflects the large-scale structure of the early universe. Given those starting conditions, astronomers can crank the handle on their cosmological models to predict how fast the universe should be expanding today. This kind of work is where the figure of 67 comes from.

So which is correct? One possibility is that astronomers in the first camp are getting their measurements of the modern universe wrong. The speed with which distant galaxies are receding is relatively straightforward to measure. Just as the pitch of an ambulance’s siren appears to change as it approaches and then speeds away, light emitted by galaxies will have longer wavelengths—and so appear redder—the faster they are receding. For that reason, it is measurements of distances that come in for the most scrutiny.

Distance measurements on galactic scales are notoriously tricky. The most common method is to combine several different techniques into something called the cosmic distance ladder, in which the farthest object measurable by one technique is used to calibrate the next. The lowest rungs are nearby stars, Earth’s distance from which can be measured by trigonometry. Higher rungs are formed by what astronomers call standard candles—stars known as Cepheid variables, for example, or certain supernovae—whose absolute brightness is known, and whose distance can therefore be inferred by how dim or bright they appear from Earth.

There are plenty of subtleties that can skew such measurements, says Wendy Freedman, an astronomer at the University of Chicago, who specialises in measuring the Hubble constant. Interstellar dust absorbs light in some wavelengths more than others, which has to be corrected for. The “metallicity” of individual Cepheids—astronomer-speak for the degree to which they contain elements other than hydrogen and helium—can influence their brightness. The specific kind of supernovae needed for distance measurements are relatively uncommon, so the sample used for distance measurements is rather small. Extraordinary claims, says Dr Freedman—such as the idea that two sets of bulletproof measurements disagree with each other—require extraordinary evidence. But the evidence so far, she says, is not quite extraordinary enough.

Others take the opposite view. “I think the idea that these measurements are wrong was more viable a few years ago,” says Adam Riess, an astronomer at the Space Telescope Science Institute in Baltimore (and one of the winners of that 2011 Nobel). As more astronomers have become interested in the Hubble tension, they have cross-checked the distance ladder measurements in other ways. Every rung has been double-checked using different standard candles, says Dr Riess, and yet the tension persists.

A paper published in June further complicated matters. It did not rely on a distance ladder of any sort. Instead it examined beams of light from bright astronomical objects called quasars. If a massive object lies between the source of that light and Earth, its gravitational effects will cause different beams of light to take different amounts of time to travel to Earth. Examining those differences lets astronomers work out how far the beams have travelled. The method came up with a value of the Hubble constant very similar to studies that rely on the old-fashioned distance ladder. That means, says Dr Riess, that if some unknown confounder is throwing off the distance measurements, it would have to be throwing off several fundamentally different sorts of measurements at once.

Some astronomers, therefore, think it is the early-universe technique that is at fault. The worry here is less about erroneous readings—the CMB has been measured and re-measured with increasing accuracy by a string of satellites since the 1990s, as well as ground-based telescopes in Chile and at the South Pole, all of which agree. The suspicion is rather that something may be wrong with the cosmological theory into which those measurements are fed.

That theory, called Lambda-CDM (LCDM) holds that the visible portion of the universe—galaxies, planets, starlight and the rest—makes up just 5% of the total. The remainder is supposedly split between “dark energy”, a force that opposes gravity at long distances and which drives the expansion of the universe (the “lambda” in LCDM), and “dark matter”, which cannot be seen but whose presence can be inferred from its gravitational effects on galaxies (CDM stands for “cold dark matter”).

LCDM might be counterintuitive. But it is very successful at predicting everything from the abundance of simple chemical elements to the distribution of galaxies and patterns within the CMB, all with high precision. Replacing it with something that is equally good but which can also predict a Hubble constant in line with present-day measurements is a tall mathematical order.

Still, there is no shortage of candidates. Some speculate that dark energy’s potency might change over time. That would mean that attempts to model today’s universe from the CMB—which assume that the nature of dark energy has not changed since the Big Bang—have been misguided. A paper presented at a meeting of the Royal Astronomical Society last month suggested that the Milky Way might sit within a giant, comparatively empty region of space, which would make the Hubble constant appear larger than it really is.

For now none of these theories has knocked LCDM off its perch. Astronomy, then, is at an impasse. It is possible that some inspired theoretician will emerge tomorrow with an idea that can solve the problem. Failing that, astronomers must fall back on the hope that yet more data will provide some vital clue. A string of new telescopes, such as the Vera Rubin Observatory in Chile or the Nancy Grace Roman Space Telescope, due to fly no later than May 2027, may offer a vital insight. But if the past few decades are any guide, they are as likely to simply re-confirm the Hubble tension as they are to resolve it.