A group of astronomers in the US has made a new and more precise measurement of the universe’s rate of expansion by using NASA’s Hubble Space Telescope (HST) to observe miniscule shifts in the apparent position of stars known as Cepheid variables. The group’s results reinforce a disagreement over the value of the Hubble constant as measured directly and as calculated via observations of primordial radiation – a disparity, say the researchers, which likely points to new physics.
In his pioneering work of the 1920s Edwin Hubble observed that galaxies further away from Earth recede more quickly, as measured by their red-shifted radiation. This implied that the universe was expanding, and that expansion has since been described by the Hubble constant, which states how many kilometres per second faster galaxies move apart from one another for every megaparsec, or 3.25::million light-years, of distance between them.
Measurements of the famous constant were imprecise until the launch of the HST, which allowed scientists to pin down a value of 72±8 in 2001. That result has since been improved upon by Adam Riess at the Space Telescope Science Institute in Baltimore, US, and colleagues, who from 2009 have reported a series of improved values thanks to data from the HST’s Wide Field Camera 3 – arriving at 73.2±1.8 in 2016.
Shortly after the Big Bang
The Hubble constant can also be deduced by calculating the universe’s rate of expansion shortly after the Big Bang using data from the cosmic microwave background (CMB) and then extrapolating to the present assuming certain properties of dark matter and dark energy. This CMB-derived value is in clear disagreement with the HST value. In 2016, the European Space Agency’s CMB-measuring Planck satellite reported a value of 66.9±0.6, implying that the cosmos ought to be expanding more slowly today than is observed.
The mismatch has now been reinforced by new results from Riess and colleagues, who have looked at Cepheid variables. These stars pulsate at a rate fixed by their intrinsic brightness, which means their apparent brightness can be used to work out how far away they are. They can also be used to calibrate the (known) brightness of type 1a supernovae, given that both are visible in some nearby galaxies, with such supernovae in turn being used to establish the distance to further-flung galaxies. This process creates a billion-parsec long “distance ladder” used to calculate the Hubble constant.
Since astronomers must initially calibrate the Cepheids themselves, the first (and hardest) rung on the ladder involves independently measuring the distance to these objects. This is done using parallax, the apparent change in position of an object compared to the background stars as seen by a moving observer. The distance between object and observer is obtained via triangulation – combining the (apparent) change in the object’s position with that of the observer.
More distant objects
Previously, Riess and colleagues had measured the parallax of Cepheids lying just a few hundred light-years from Earth. They have now turned their attention to more distant objects – eight Cepheids situated between 6000-12,000 light-years away (although still within the Milky Way). These are particularly well suited to the distance ladder since they pulsate at the lower rates characteristic of Cepheids found together with type 1a supernovae in other galaxies.
Riess’s group measures parallax by observing each Cepheid twice a year, with the Earth (and with it the HST) on opposite sides of its orbit around the Sun. But because this change in position is tiny compared to the distance separating the stars and Earth, the parallax is correspondingly minute – amounting to just one hundredth the size of a single pixel on Wide Field Camera 3.
To get around this problem, rather than taking a snapshot of each Cepheid the researchers instead scanned the camera across it as the HST moved in its orbit, so spreading the light over 4000 pixels. As Riess explains, doing so overcomes the fact that each pixel is like a well and fills up after receiving a certain number of photons. “You get more photons altogether by scanning,” he says.
“Conspiracy of errors”
Using this approach, the group calculate a Hubble constant of 73.5 ±1.7, which is a 3.7σ disagreement with the Planck results. This means that there is a 1 in 5000 chance that the disparity is a statistical fluke. What is more, Riess points out, over the last couple of years independent probes have confirmed both the distance-ladder and CMB results – gravitational lensing and baryon acoustic oscillations, respectively. “There would have to be a series of systematic errors in techniques that have nothing to do with each other,” he says. “And once you start to think about a conspiracy of errors that doesn’t look very likely.”
Dating the universe
As to what new physics might be responsible for the disparity, Riess says that it could be caused by hypothetical “sterile neutrinos”, interactions with dark matter, or a strengthening over time of dark energy (which accelerates the universe’s expansion). He adds that the team will use the HST to measure more Cepheids and that data from ESA’s Gaia satellite, due to be released in April, should contain parallax information from around 200 such stars – thus further reducing the Hubble constant’s uncertainty and potentially narrowing down the source of the disparity, he says.
Chuck Bennett of Johns Hopkins University in the US, who led the team on Planck’s predecessor WMAP, is cautious. He says that the new result “places even further stress on some potential cracks in the standard model of cosmology” but argues that more work needs to be done. “Unfortunately, none of the commonly discussed potential modifications to the standard model seem to solve the tensions while also being compelling.”