In May 2009, the Planck Satellite was launched with a mission to discover the age of the universe. The goal was to measure cosmic background radiation (CMB), a source of light that traces back to around 380,000 years after the universe’s start at the Big Bang. Those measurements revealed the universe to be around 13.8 billion years old.
But in the past few years, astronomers have improved on classic observations of how fast distant galaxies move away from Earth. These measurements result in an age of the universe hundreds of millions of years younger than the Planck measurements indicated.
In July, the research team behind the National Science Foundation’s Atacama Cosmology Telescope (ACT), which includes Department of Physics and Astronomy Chair Arthur Kosowsky, graduate student Yilun Guan and 40 other participating institutions, jumped into the fray with their own estimates. The team made new maps of the slight variations in the microwave background’s temperature and polarization which are more sensitive and sharper than Planck’s.
The simplest model of the universe that fits the new data from ACT, which is located in Chile, has an age of 13.77 billion years with an uncertainty of plus or minus 40 million years—essentially confirming the Planck results. Papers featuring the findings were posted to the arXiv distribution service and have been submitted to the Journal of Cosmology and Astroparticle Physics.
“Now we’ve come up with an answer where Planck and ACT agree,” says Simone Aiola (A&S ’14G, '16G), a lead author and researcher at the Flatiron Institute’s Center for Computational Astrophysics who holds a PhD in physics from Pitt's Kenneth P. Dietrich School of Arts and Sciences. “It speaks to the fact that these difficult measurements are reliable.”
Kosowsky said the findings add to the confusion surrounding the actual age of the universe but also provide a few clues about where cosmologists should look for answers.
He noted that ACT also essentially confirmed Plank’s estimated Hubble constant, which is the rate at which the universe is expanding. ACT found the universe is expanding at a rate of 67.6 kilometers per second per megaparsec, which means an object located a megaparsec away from Earth (approximately 3.26 million light years) is on average moving away from us at the rate of 67.6 kilometers per second due to the cosmic expansion.
Probably the most interesting possibility for this discrepancy is that our simple model of the universe is wrong.
The Planck team’s Hubble constant was 67.4 kilometers per second per megaparsec and the rates estimated by measuring galaxies range between 70 and 74 kilometers per second per megaparsec, significantly faster than the other results.
“Probably the most interesting possibility for this discrepancy is that our simple model of the universe is wrong. So the inference we’re making about the Hubble parameter from our measurement is based on a model that’s not quite right,” said Kosowsky. “Our basic model of the universe is based on some really simple assumptions about some really simple physics. If you change something you’re probably saying there’s new, undiscovered, fundamental physics that we’re seeing in this discrepancy.”
In the dark
Guan, who is pursuing a PhD in cosmology at Pitt's Dietrich School, is investigating whether the flaw in the model lies in assumptions made about the nature of dark energy, which is what researchers call the unknown force that is accelerating the universe’s expansion.
“In our current model of the universe we treat dark energy as a really simple component that has a simple evolution—there’s this amount of dark energy and it evolves in a certain way. But nobody knows what dark energy is doing, it could be that dark energy is changing from when we measure it at the CMB until we measure it in these local measurements,” he explained. “What we’re doing is thinking about if there’s any way dark energy can be changing that is consistent with measurements from both our local measurements as well as measurements from CMB and other observations that have been made.”
While Guan continues the dark energy research, he is also investigating ways to automate the process of finding the most useful data from telescopes and satellites. On the ACT project, he was responsible for the first levels of data quality assurance, a process that meant analyzing tens of millions of detector time streams collected by ACT each year to determine which ones are to be used to make sky maps.
“Telling good data from bad data requires a lot of expert knowledge. In the past we used to have one expert who had a lot of experience in this area and if he looks at diagnostics and statistics he can tell if something is going wrong. But that’s not a scalable solution,” Guan said.
In 2022, ACT will officially shut down and its research team will shift attention to the Simons Observatory, also located in Chile, which will feature new telescopes, cameras and state of the art detector arrays. Those upgrades will give experts “an order of magnitude more data” to assess. In anticipation of the challenge, last year Guan started work on a machine learning algorithm designed to take on the task.
“It turns out, this is a very easy problem for the machines. We have lots of data that experts told us is bad or good so we can train our algorithms to learn from what experts decide and have these algorithms make these decisions for us,” he said.
For ACT’s upcoming data release next year, which covers information gathered from 2017 through 2019 and will feature four times more data, experts will still lead the way in data quality assurance. However, Guan hopes to have an automated process in place before it’s time to analyze data from Simons.
Kosowsky said the data that comes from the Simons Observatory could hold precisely what’s needed to resolve mysteries surrounding the age of the universe once and for all.
“The fact that we’re doing measurements right now where the pieces are not all completely fitting together makes it really exciting to have the possibility of doing even better measurements of the universe with the next generation experiments,” he said.