“It would be incredibly exciting if there was new physics,” Freedman said. “I have a secret in my heart that I hope there is, that there’s a discovery to be made there. But we want to make sure we’re right. There’s work to do before we can say so unequivocally.”
That work involves reducing possible sources of error in measurements of the cosmic expansion rate. One of the biggest sources of that uncertainty has been the distances to nearby stars — distances that the new parallax data appears to all but nail down.
In a paper posted online last night and submitted to The Astrophysical Journal, Riess’s team has used the new data to peg the expansion rate at 73.2 kilometers per second per megaparsec, in line with their previous value, but now with a margin of error of just 1.8%. That seemingly cements the discrepancy with the far lower predicted rate of 67.
Freedman and Madore expect to publish their group’s new and improved measurement of the cosmic expansion rate in January. They too expect the new data to firm up, rather than shift, their measurement, which has tended to land lower than Riess’s and those of other groups but still higher than the prediction.
Since Gaia launched in December 2013, it has released two other massive data sets that have revolutionized our understanding of our cosmic neighborhood. Yet Gaia’s earlier parallax measurements were a disappointment. “When we looked at the first data release” in 2016, Freedman said, “we wanted to cry.”
An Unforeseen Problem
If parallaxes were easier to measure, the Copernican revolution might have happened sooner.
Copernicus proposed in the 16th century that the Earth revolves around the sun. But even at the time, astronomers knew about parallax. If Earth moved, as Copernicus held, then they expected to see nearby stars shifting in the sky as it did so, just as a lamppost appears to shift relative to the background hills as you cross the street. The astronomer Tycho Brahe didn’t detect any such stellar parallax and thereby concluded that Earth does not move.
And yet it does, and the stars do shift — albeit barely, because they’re so far away.
It took until 1838 for a German astronomer named Friedrich Bessel to detect stellar parallax. By measuring the angular shift of the star system 61 Cygni relative to the surrounding stars, Bessel concluded that it was 10.3 light-years away. His measurement differed from the true value by only 10% — Gaia’s new measurements place the two stars in the system at 11.4030 and 11.4026 light-years away, give or take one or two thousandths of a light-year.
The 61 Cygni system is exceptionally close. More typical Milky Way stars shift by mere ten-thousandths of an arcsecond — just hundredths of a pixel in a modern telescope camera. Detecting the motion requires specialized, ultra-stable instruments. Gaia was designed for the purpose, but when it switched on, the telescope had an unforeseen problem.
The telescope works by looking in two directions at once and tracking the angular differences between stars in its two fields of view, explained Lennart Lindegren, who co-proposed the Gaia mission in 1993 and led the analysis of its new parallax data. Accurate parallax estimates require the angle between the two fields of view to stay fixed. But early in the Gaia mission, scientists discovered that it does not. The telescope flexes slightly as it rotates with respect to the sun, introducing a wobble into its measurements that mimics parallax. Worse, this parallax “offset” depends in complicated ways on objects’ positions, colors and brightness.
However, as data has accrued, the Gaia scientists have found it easier to separate the fake parallax from the real. Lindegren and colleagues managed to remove much of the telescope’s wobble from the newly released parallax data, while also devising a formula that researchers can use to correct the final parallax measurements depending on a star’s position, color and brightness.
Climbing the Ladder
With the new data in hand, Riess, Freedman and Madore and their teams have been able to recalculate the universe’s expansion rate. In broad strokes, the way to gauge cosmic expansion is to figure out how far away distant galaxies are and how fast they’re receding from us. The speed measurements are straightforward; distances are hard.
The most precise measurements rely on intricate “cosmic distance ladders.” The first rung consists of “standard candle” stars in and around our own galaxy that have well-defined luminosities, and which are close enough to exhibit parallax — the only sure way to tell how far away things are without traveling there. Astronomers then compare the brightness of these standard candles with that of fainter ones in nearby galaxies to deduce their distances. That’s the second rung of the ladder. Knowing the distances of these galaxies, which are chosen because they contain rare, bright stellar explosions called Type 1a supernovas, allows cosmologists to gauge the relative distances of farther-away galaxies that contain fainter Type 1a supernovas. The ratio of these faraway galaxies’ speeds to their distances gives the cosmic expansion rate.
Parallaxes are thus crucial to the whole construction. “You change the first step — the parallaxes — then everything that follows changes as well,” said Riess, who is one of the leaders of the distance ladder approach. “If you change the precision of the first step, then the precision of everything else changes.”
Riess’s team has used Gaia’s new parallaxes of 75 Cepheids — pulsating stars that are their preferred standard candles — to recalibrate their measurement of the cosmic expansion rate.
Freedman and Madore, Riess’s chief rivals at the top of the distance ladder game, have argued in recent years that Cepheids foster possible missteps on higher rungs of the ladder. So rather than lean too heavily on them, their team is combining measurements based on multiple kinds of standard-candle stars from the Gaia data set, including Cepheids, RR Lyrae stars, tip-of-the-red-giant-branch stars and so-called carbon stars.
“Gaia’s [new data release] is providing us with a secure foundation,” said Madore. Although a series of papers by Madore and Freedman’s team aren’t expected for a few weeks, they noted that the new parallax data and correction formula appear to work well. When used with various methods of plotting and dissecting the measurements, data points representing Cepheids and other special stars fall neatly along straight lines, with very little of the “scatter” that would indicate random error.
“It’s telling us we’re really looking at the real stuff,” Madore said.