The latest, most precise, estimate of the universe’s current rate of expansion - a value known as the Hubble constant - comes from observations by the European Space Agency’s Gaia mission, which is conducting the most detailed ever three-dimensional survey of the Milky Way.

The data has allowed the rate of expansion to be pinned down to a supposed accuracy of a couple of percent. However, this newest estimate stands in stark contradiction with an independent measure of the Hubble constant based on observations of ancient light that was released shortly after the Big Bang. In short, the universe is getting bigger quicker than it should be.

The mismatch is significant and problematic because the Hubble constant is widely regarded as the most fundamental number in cosmology.

“The fact the universe is expanding is really one of the most powerful ways we have to determine the composition of the universe, the age of the universe and the fate of the universe,” said Professor Adam Riess, at the Space Telescope Science Institute in Baltimore, Maryland, who led the latest analysis. “The Hubble constant quantifies all that into one number.”

In an expanding universe, the further away a star or galaxy is, the quicker it is receding. Hubble’s constant – proposed by Edwin Hubble in the 1920s – reveals by how much.

So one approach to measuring it is by observing the redshifts of bright supernovae, whose light is stretched as the very space it is travelling through expands. A challenge, though, is pinpointing the exact distance of these stars.

Riess, who shared the 2011 Nobel Prize for Physics for providing evidence that the expansion of the universe is accelerating, is part of a team focussed on developing ultra-precise methods for measuring distances.

The latest Gaia observations have advanced this effort by identifying dozens of new Cepheid stars, which have the special feature that their light flickers at a rate that is directly linked to their brightness at source. So through observing the pulsations of these so-called standard candles, scientists can work out their original luminosity and, therefore, how far away they and their native galaxies are.

The new data puts the Hubble constant at 73, which translates to galaxies moving away from us 73km per second faster for each additional megaparsec of distance between us and them (a megaparsec is about 3.3m light-years).

However, a separate estimate of Hubble comes from observations of the Cosmic Microwave Background, relic radiation that allows scientists to calculate how quickly the universe was expanding 300,000 years after the big bang.

“The cosmic microwave background is the light that is the furthest away from us that we can see,” said Riess. “It’s been travelling for 13.7bn years... and it’s telling us how fast the universe was expanding when the universe was a baby.”

Scientists then use the cosmic equivalent of a child growth chart (a computational model that roughly describes the age and contents of the universe and the laws of physics) to predict how fast the universe should be expanding today. This gives a Hubble value of 67.

Until recently, scientists had hoped that as measurements became more precise, this mismatch would narrow, but instead the difference has widened and the latest calculation gives a chance of only 1 in 7,000 of the discrepancy being down to chance. “If this continues to hold up we may be dealing with what we call new physics of the universe,” said Riess.

What form might this take? One proposal is that dark energy, believed to be accelerating the expansion of the universe, is becoming more potent. Scientists are not sure why this would happen - since space is being stretched out, one might expect its strength to be diluted instead. Another possibility is that a fourth, not yet observed, variety of neutrino could have skewed calculations. Dark matter could also be to blame if it turns out to interact more with normal matter than current models predict.

“I’m not in the business of ensuring that everything fits,” said Riess. “I think: ‘Ah this is very interesting’.”

More prosaic explanations have also been put forward. John Peacock, professor of cosmology at the University of Edinburgh, said: “Beyond a certain level of complexity you have to be open to the possibility that there may be little assumptions that might not be quite right.”

“I’m sticking with [a Hubble value of] 70 for the foreseeable future,” he added.

The crisis in cosmology, as it was described a meeting of the American Physical Society last month, could soon be resolved through new measurements of the Hubble constant based on gravitational wave observations by the Ligo collaboration. “Within the next five years, we’ll probably nail this,” said Peacock.

By Hannah Devlin, Science correspondent