Astronomers have mapped the ages of 70,000 stars spanning our galaxy, ushering in a new era of galactic archaeology.

In an archaeological dig, scientists sift through the shards and bones that lay scattered in the dirt. They carefully brush off layers of dust, note each item’s position, and take samples for carbon dating. Now, with a new technique in hand, astronomers are doing much the same, but their dig site is a bit bigger: the Milky Way Galaxy.

At the American Astronomical Society meeting in Kissimmee, Florida, Melissa Ness (Max Planck Institute for Astronomy, Germany) announced her team's release of a catalog that maps the ages of 70,000 red giant stars scattered throughout our galaxy. Like carbon-dating, these luminous beacons and the new method of measuring their ages will revolutionize the field of galactic archaeology.

Milky Way age plot
This image shows 70,000 red giant stars as colored dots embedded in a simulation of the Milky Way Galaxy. The color scale shows stellar age: red denotes the oldest stars, which formed some 12 billion years ago when the Milky Way was young and small, while blue shows younger stars that formed more recently, when the Milky Way was big and mature.
G. Stinson (MPIA)

Digging Up the Milky Way

“Big Data” has become a buzzword everywhere from the corner office to the genetics lab, but astronomy’s been in the business of big data since the advent of the Sloan Digital Sky Survey (SDSS) in 2000. In recent years, the Sloan telescope turned its watchful eye from the farthest reaches of the cosmos to our own galaxy. Since 2011, a survey dubbed APOGEE has sighted 150,000 red giant stars across the Milky Way, splitting starlight across the visible-light range to collect detailed spectra for every star.

But sifting through all that data isn’t easy. And that’s particularly true when it comes to measuring something more indirect like age — stars are secretive about the number of years under their belt.

One way to date a star is to measure the seismic waves that ripple across its surface. These starquakes create pulsations in brightness that change as the star ages. But doing these observations requires time — the kind of time that Kepler had when it stared at hundreds of thousands of stars for four years.

Another way is more indirect. The outer layers of stars like the Sun boil with convective fervor, and when these stars evolve to become red giants, that roiling plasma extends its reach down to scrape the star’s core. Just-fused elements get dredged up from the core and carried to the surface, theoretically revealing how long the star has been around.

Marie Martig (also at the Max Planck Institute for Astronomy) and colleagues took the best of both worlds when they examined 1,475 red giants with both seismic masses from Kepler and dredged-element data from Sloan spectra. This set of overlapping data became the training set for an algorithm dubbed “The Cannon” (named in honor of Annie Jump Cannon). After teaching The Cannon to classify stars’ ages by their spectra, Ness, Martig, and colleagues set it loose on the full dataset. It returned the age for some 70,000 stars strewn throughout the Milky Way.

Galactic Archaeology

Star dating, like the carbon dating that revolutionized archaeology, provides astronomers with an essential tool to piece together our galaxy’s history. In a way, it’s still a crude tool — the mass estimates for any given star can be off by as much as 40%. But combined, tens of thousands of stars beat those errors down and provide useful insight. Already, the team has been able to confirm that our galaxy’s spiral disk formed from the inside out, bolstering accepted theory.

The youngest red giant stars assemble along the galactic plane, a skeleton of sorts that’s encased by older stars. Farther from the galaxy’s center, this “backbone” of younger stars flares outward, away from the plane. These patterns are the hallmark of a disk that started small and grew slowly outward, as though whoever was pouring the pancake batter first ladled into the center and then added more and more batter on the outer edges.

These results prove The Cannon works and lay the foundation for future studies, says Daniel Majaess (St. Mary’s University and Mount St. Vincent University, Canada). And Ness, who has already had lots of requests for the catalog, expects those future studies will be coming out soon.


Image of Anthony Barreiro

Anthony Barreiro

January 11, 2016 at 2:04 pm

Is the metallicity of these stars a factor in determining their ages? Younger stars generally have more heavy elements in them than older stars, and the differing proportions of these heavy elements should show up in the stars' spectra. Or does the fact that they're all red giants mean that they've all created their own heavy elements? This is very interesting work, and a very impressive graphic, but I'm still trying to understand it.

You must be logged in to post a comment.

Image of Monica Young

Monica Young

January 13, 2016 at 1:24 pm

Hi Anthony - an excellent question. My understanding from Melissa Ness's presentation and the paper is that the team takes metallicity into account in an empirical way. That is, when they feed the training set of stellar spectra to the machine, they tell it what the age and mass of each star is. The machine then "learns" how features in a star's spectrum relate to its age. In that way, it's able to not only capture the dredge-up process but also the role that metallicity plays, event though the machine (and the astronomers, for that matter) don't fully understand how stars age.

You must be logged in to post a comment.

Image of Anthony Barreiro

Anthony Barreiro

January 13, 2016 at 4:00 pm

Fascinating, thanks. I guess we're well and truly into the age of big data and machine learning. A hundred years ago Annie Jump Cannon could look at spectra and figure out how to classify stars. Now we need a supercomputer and an algorithm that rewrites itself as we feed it more data.

You must be logged in to post a comment.

Image of Kevin


January 15, 2016 at 5:49 pm

Wouldn't younger stars have far FEWER heavy elements in them? As a star ages it continually fuses its fuel from hydrogen to helium to carbon and so on until only iron remains at which point its mass determines its fate.

I feel sorry for the stars that are too light to become a supernova..."Your brother applied himself and became massive enough to go supernova. Are you telling your father and I that you are content to become just another red giant?"

You must be logged in to post a comment.

Image of

January 16, 2016 at 1:45 pm

Where are WE in that simulated image of the Milky Way? Are we in the red side or the blue side of the Galaxy? Can we not see young, hot stars to further distances than old red giants? Doesn't interstellar dust scatter blue light more than red?

You must be logged in to post a comment.

You must be logged in to post a comment.