© Reidar Hahn, Fermilab
The best cosmic map yet of the universe's make-up finds 24 per cent less dark matter than we thought and could call for a rewrite of physics
The fate of the universe has never been certain, but it just became even less so. That's due to a disagreement between a new map of today's universe and an existing map of the early universe. The mismatch either means one of the measurements is wrong or, disturbingly, that we need to rewrite physics.

The results, which are part of the Dark Energy Survey (DES), charted the distribution of matter across 26 million galaxies in a large swathe of the southern sky.

"This is one of the most powerful pictures of the universe today that we've ever had," says Daniel Scolnic at the University of Chicago, who is a part of the 400-person DES collaboration but wasn't involved in this work.

It is so powerful because knowing this distribution helps us understand the cosmic game of tug of war between dark energy, the mysterious force accelerating the universe's expansion, and dark matter, the hidden extra mass in the universe. Dark energy tends to pull each galaxy apart, while dark matter's gravity brings each galaxy together. From the relative strengths of these effects, we can predict how the cosmos will change in the future.

The new results are based on data from the 4-metre Victor M. Blanco Telescope in Chile. It was the scope's first observing season for the DES, lasting a mere six months. The survey is slated to run for five years, meaning that DES's measurements can only improve. But its first results have already revealed a tension with earlier findings.

In 2013, astronomers used ESA's Planck satellite to chart the universe's dark contents across the early cosmos, around 380,000 years after the big bang.

"We've had really good baby pictures of the universe and now, with this, it's like the first time we've had really good selfies," Scolnic says.

Comparing these two snapshots allows us to piece together how the universe evolved from its early state to the present - and make predictions about what the future may hold. Although the spread of the universe's dark contents may have changed, current models of cosmology suggest that the proportion of it remains constant. DES's first findings, however, might suggest otherwise.

Take dark matter, for example. Planck pegged it at 34 per cent of the energy of the early universe, but DES finds that today it only amounts to 26 per cent (arxiv.org/abs/1708.01530). That could mean dark matter is losing the cosmic game of tug of war to dark energy - a result that would force a radical rewrite of physics.

"If [the two different answers] don't go away, we're seeing the first signs of what could be a very serious problem in the cosmological model," says David Spergel at Princeton University.

"If dark energy increases over time, galaxies, their stars and the atoms within us will be torn to shreds"

Astronomers are hesitant to make too grandiose a claim based on the first data set. Statistically speaking, there's only a slight tension between the results from today's universe and the early universe. But Scolnic points out that this tension is one in a line of many. Results from the South Pole Telescope, for example, similarly disagree with Planck.

"It's hard to believe that this is a coincidence," Scolnic says. And if it isn't a coincidence, the results could mean that dark energy actually grows stronger with time - an idea that has wild implications for our future.

Previously, we assumed that galaxies would remain forever bound despite the universe continuing to expand. But, should dark energy increase with time, then one day galaxies will also expand until they are torn to shreds - followed by stars, planets and even the atoms within us.

"That would be a disturbing conclusion," Spergel says. But he advises against panicking just yet.

The latest mismatch could mean that one or both of the measurements is wrong. If so, it might disappear with more data. Then again, it might not. And the fate of the universe hangs in the balance.