How Cosmologists Determined That the Universe Is Expanding Faster Than Anyone Thought

A team of astronomers from the Space Telescope Institute and Johns Hopkins University led by Nobel laureate Adam Reiss has confirmed data that the universe is expanding significantly faster than previously thought.

As detailed in a forthcoming paper for The Astrophysical Journal, Reiss and his colleagues used four years’ worth of data from the Hubble Space Telescope to determine that the universe is expanding about 9 percent faster than other leading measurements predicted—a wild mismatch in a field as precise as cosmology.

The data used to reach this conclusion is the most accurate measurement of the expansion of the universe since it was discovered to be expanding nearly a century ago.

The results raise profound questions about what could be causing the mismatch between predictions about the acceleration of the expanding universe, and may lead to fundamental insights about the nature of dark energy.

“The community is really grappling with understanding the meaning of this discrepancy,” Reiss, who shared a Nobel Prize for discovering that the universe expands at an accelerating rate in 1998, said in a statement.

The universe looked a whole lot different a century ago compared to today. Back then, astronomers had no way of measuring objects outside our own galaxy, and mistakenly characterized the galaxies they could see as other stars or clouds of gas in the Milky Way.

In 1913, Harvard astronomer Henrietta Leavitt discovered Cepheid variables, stars whose brightness increased and dimmed in a consistent way. These types of stars could be used as a cosmic ruler to measure the distance of celestial objects by observing how light waves are ‘stretched’ by the expansion of space. This insight was used by the astronomer Edwin Hubble in 1923 to measure a Cepheid variable star in the Andromeda nebula, thereby proving the existence of a galaxy outside of our own for the first time.

Hubble went on to discover many more galaxies over his career and noticed something remarkable about the relationship between them: As the distance between two galaxies increases, so does the relative speed at which they are moving away from one another. This observation was the first evidence that the universe was expanding, even if Hubble wasn’t necessarily the first to realize it. The rate at which the universe is expanding was called the Hubble constant, which is somewhat misleading since Reiss and his colleagues demonstrated that the universe is expanding at an accelerating—not constant—rate in the late 90s.

Until Hubble’s death in 1953, and for a few decades afterwards, the actual value of the Hubble constant was a subject of intense debate among cosmologists and thought to be anywhere from 50 to 100 kilometers per second per megaparsec.

A megaparsec is a measurement of distance, equal to about 3.3 million light years. So if the Hubble constant has a value of 50 kilometers per second per megaparsec that means that for every 3.3 million light years an object is away from an observer, its speed increases by 50 kilometers per second. So an object that is one megaparsec away is receding at a rate of 50 km/s, while an object 100 megaparsecs away is receding at 100 times that speed, or 5000 km/s.

From 2009 to 2013, the European Space Agency operated the Planck space observatory, whose main mission was to map the cosmic microwave background. The CMB is radiation from just 380,000 years after the Big Bang, and is the earliest snapshot we have of the universe. Using data from Planck’s measurements of the CMB, physicists were able to predict the rate at which the universe has expanded since the Big Bang. According to these ultraprecise measurements of the early universe, they predicted the Hubble constant to be approximately 67 kilometers per second per megaparsec today.

There was only one problem—the Planck data was derived by measuring incredibly distant objects in the early universe, but when Reiss and his colleagues measured galaxies that were closer to us in space back in 2011, they found the universe to be expanding at a rate of around 74 kilometers per second per megaparsec.

The discrepancy between the Hubble constant measured for near and far galaxies posed a major problem for astronomers: They couldn’t explain it.

“Both results have been tested multiple ways, so barring a series of unrelated mistakes, it is increasingly like that this is not a bug, but a feature of the universe,” Reiss said.

Explaining the difference between these two values of the Hubble constant has been the focus of Reiss and his team—called Supernova H0 for the Equation of State—for over a decade. Yet in order to understand why these values differ, Reiss and his colleagues needed to refine their measurements.

To do this, the astronomers measured eight Cepheid variable stars in the Milky Way using the Hubble Space Telescope. These stars are between 6,000 and 12,000 light years away, roughly 10 times further than other Cepheid stars studied with the Hubble Space Telescope, which were at most 1,600 light years away.

When Henrietta Leavitt first discovered these types of stars over a century ago, she used periodic fluctuations in their brightness to determine their distance. This method would also be used by Reiss and co, but not until after they independently calculated the distance to these eight new Cepheid stars independently of their brightness using a tool called parallax, which measures distance based on the change in an object’s position from an observer’s point of view.

Parallax was first discovered by the ancient Greeks and used to measure the distance to the moon, but the measurement challenge faced by Reiss was far more daunting. He and his colleagues were measuring tiny wobbles in the eight Cepheid stars, and each wobble is only equal to 1/100 of a single pixel in Hubble’s camera. To put this in perspective, that’s like measuring the movement of a grain of sand from 100 miles away.

To accomplish this, Reiss created a scanning technique that measured the positions of the eight stars one thousand times per minute every six months for four years. This essentially means Hubble is dragging its camera across a star, capturing the image as a streak of light.

“This method allows for repeated opportunities to measure the extremely tiny displacements due to parallax,” Reiss said. “You’re measuring the separation between two stars, not just in one place on the camera, but over and over thousands of times, reducing the errors in measurement.”

The team then calculated the distance to these eight nearby Cepheid stars based on parallax and cross-referenced this to the calculated distance based on their brightness. This cross-referencing resulted in unprecedented precision when measuring Cepheid variable stars in distant galaxies. Based on this data, Reiss and his colleagues derived the most accurate measurement of the Hubble constant ever: 74 kilometers per second per megaparsec.

This is the same value they had derived in 2011, but this time around their degree of uncertainty is far smaller: In 2009, Reiss’ measurements carried a 4.7 percent uncertainty, in 2011 it was a 3.3 percent and the most recent measurement uncertainty is only 2.4 percent.

Okay, so now we have incredibly accurate data for the rate of the universe expanding, but that still doesn’t explain the discrepancy between the values from Planck and Hubble—but Reiss has some ideas. An immediately obvious candidate is dark energy, the stuff that makes up 95 percent of the universe and has already been shown by Reiss and his fellow Nobel laureates to contribute to the accelerating expansion of the universe. In this case, it may be forcing galaxies apart with growing strength, which would mean that the rate of acceleration itself is not constant.

Another theory attributes the acceleration discrepancy to a class of subatomic particles that move around the speed of light and are affected only by gravity, known as dark radiation. Finally, there’s a chance that dark matter interacts with ‘normal’ matter more strongly than previously assumed. All of these theories could account for the discrepancies in the speed of the universe’s expansion because they would change over time, meaning that values observed in the early universe by Planck would be different than values observed closer to the present time by Hubble.

Reiss and his colleagues hope to contribute to this problem’s resolution with still more data from Hubble, as well as data from ESA’s Gaia space telescope, which is dedicated to measuring the position and distances of stars with unprecedented precision.

“Ordinarily, if every six months you try to measure the change in position of one star relative to another at these distances, you are limited by your ability to figure out exactly where the star is,” Reiss said. “This precision [in Gaia] is what it will take to diagnose the cause of this discrepancy.”