Skip to content
Starts With A Bang

How ‘Cosmic DNA’ Revealed Exoplanet Siblings Raised In The Same Nursery

How two seemingly distinct exoplanet systems turned out to be related.

Practically every star in the Milky Way has a similar origin story. At some point in the past, a molecular cloud of gas — mostly hydrogen and helium but enriched with the heavy elements from prior generations of stars — began to contract under its own gravity. As that cloud contracts, it radiates heat away, becoming dense enough in places that mass begins to accumulate in a runaway process. After millions of years, proto-stars and then full-fledged stars begin to form, and the race is on: between gravity, trying to grow and form as many stars as possible, and from the radiation from the newly formed stars, which works to boil the neutral matter off and prevent stars from forming further.

This process, as we understand it, is what occurs within star-forming regions, giving rise to new star birth and a familiar sight to astronomers: star clusters. These star clusters typically live only a short while, then dissociate, and the stars get randomly distributed throughout the galaxy. Tracing them back to their original nursery is often too complex a task, but recent advances may have just made that possible. For the first time, two stars housing exoplanets — Kepler 52 and Kepler 968 — have been traced back to their “parental” star cluster, and we’ve indeed confirmed it: these two mature systems are only just now leaving their childhood homes. Here’s how we know.

During its primary mission, NASA’s Kepler observed the same patch of sky for years. As a result, while watching more than 100,000 stars in its field of view at once, it discovered thousands of star systems with planets all their own. (JON LOMBERG (ARTWORK), NASA (KEPLER DIAGRAM))

When the Kepler mission first began observing the sky, the plan was simple, straightforward, and brilliant. It would point its telescopic eye at the same region of space, over and over again, for years at a time. As it observed this area on the sky — located along one of the arms of our galactic plane — it collected data on more than 100,000 stars simultaneously. For most of these stars, their planets orbited out-of-the-plane that intersected with our line-of-sight. As long as the star wasn’t inherently variable and none of the planets passed in front of the star’s disk as they orbited, the brightness of each star would remain constant.

But with over 100,000 stars to view, even relatively rare configurations could be abundantly found. Even though only a small percentage of stars were aligned fortuitously, so that (at least) one or more of their inner planets did pass in front of the star’s disk during its orbit from our perspective, we could identify a periodic dimming of the star. If this transit event occurred repeatedly and could be followed up with a complementary measurement, this interesting event could get promoted first to an exoplanetary candidate and then to a confirmed exoplanet.

This figure shows the number of systems with one, two, three, planets, etc. Each dot represents one known planetary system. As of 2017, we knew of more than 2,000 one-planet systems, and progressively fewer systems with many planets. In the subsequent years, these numbers have continued to increase, with over 4,000 total exoplanets now known. (NASA/AMES RESEARCH CENTER/WENDY STENZEL AND THE UNIVERSITY OF TEXAS AT AUSTIN/ANDREW VANDERBURG)

Since its launch just over a decade ago, NASA’s Kepler discovered thousands of stars that housed one or more planets around it, with our current exoplanet total now exceeding 4,000 total planets. Two of these stars, in nearly the same region of the sky, seemed to be both typical and unremarkable in many similar ways: Kepler 52 and Kepler 968.

Kepler 52 has three known exoplanets around it, with the farthest one about half as far away as Mercury is from our Sun. Kepler 52, the star, is less massive and luminous than our Sun (about 54% as massive), and is the most massive kind of M-type star: right on the border between what makes a red dwarf, which will never fuse helium into carbon, and a K-type star, which will someday get there.

Kepler 968, on the other hand, has two known exoplanets that are in extremely tight orbits: only separated from their parent star by about ~10% of the Sun-Mercury distance. Kepler 968 is a bit more massive of a star, at 76% the mass of our Sun, and is a full-fledged K-class star: between the Sun-like G-type and the low-mass M-type.

The (modern) Morgan–Keenan spectral classification system, with the temperature range of each star class shown above it, in kelvin. Our Sun is a G-class star, producing light with an effective temperature of around 5800 K and a brightness of 1 solar luminosity. Stars can be as low in mass as 8% the mass of our Sun, where they’ll burn with ~0.01% our Sun’s brightness and live for more than 1000 times as long, but they can also rise to hundreds of times our Sun’s mass, with millions of times our Sun’s luminosity and lifetimes of just a few million years. The first generation of stars should consist of O-type and B-type stars almost exclusively, and may contain stars up to 1,000+ times the mass of our Sun. (WIKIMEDIA COMMONS USER LUCASVB, ADDITIONS BY E. SIEGEL)

These two stars, on the surface, seem unrelated to one another. They’re in nearby but distinct parts of the sky, they’re both a little more than 1,000 light-years away, and their ages, based on data from the European Space Agency’s Gaia mission, are extremely poorly constrained. These are both evolved stars, with indications that:

  • they’re fusing hydrogen into helium in their cores,
  • they only have fully-formed planets around them, not protoplanetary disks or other rich sources of dusty debris,
  • and that their solar systems are mature, like our own.

If this were all we knew about these stars, we’d probably call it a day. Like many stars, they appear to have a system of planets around them, aren’t associated with any sort of star cluster, and have lots of uncertainties about their properties. Sure, we know each of the star’s masses and the orbital properties of the planets, but beyond that, it’s very difficult to infer things like their ages, their rotation periods, their metallicities, or how they’re moving relative to us and one another; the Kepler data, and even the follow-up data to confirm the existence of these exoplanets, doesn’t tell us all that much on their own.

The ESA’s Gaia mission has measured the positions and properties of hundreds of millions of stars near the galactic center, and is finding evidence of some of the oldest stars known to humanity present in this environment. It has also identified enormous, extended, diffuse star clusters, perhaps thousands of them throughout the Milky Way, that had never been identified before. (ESA/GAIA/DPAC)

However, these stars and their planetary systems haven’t only been observed by Kepler, but also from NASA’s TESS — the Transiting Exoplanet Survey Satellite — and from the Zwicky Transient Facility. With the combined data from three separate observatories, scientists were able to measure two very important properties about these stars:

  1. their rotation rates, determining how quickly each star takes to spin a complete revolution about its own axis,
  2. and the mass of the parent star, inferred by the properties of the orbiting planets.

These two pieces of information, combined, are tremendously interesting. The reason is straightforward: when stars are born, they rotate rapidly; it only takes a few hours to a few days for them to complete a full spin of 360°. However, over time, their magnetic fields cause their rotation rate to slow. If you’re born spinning rapidly, your magnetic field will slow you down more quickly. Also, if you’re a less massive star, your rotation rate lengthens more easily than if you’re more massive, which leads to an interesting phenomenon. Once your set of newborn stars is more than about 100 million years old, all the stars more massive than a certain threshold will display a nice, clean correlation between their masses and their rotation rates, with the specifics of that correlation highly dependent on the ages of the stars. As star clusters age, the more massive stars evolve, leaving only the less massive, less luminous members behind.

The stars present and absent within a recently born cluster reveal its age. Initially, their distribution follows the long, curved line from lower-right to upper-left. As the stars age, the ones at the top left evolve up and towards the right, with greater ages further lowering the “turn-off” point on the curve. The bluest, brightest stars are also the shortest lived. (CHRISTOPHER TOUT, NATURE 478, 331–332 (2011))

At the same time, the last few years have yielded a relative surprise for astronomers as far as stars go. The ESA’s Gaia mission, designed to exquisitely measure the properties of more than a billion stars in our galaxy — how far away they are, their positions, their motions over time, their colors, their parallaxes, etc. — started to find star clusters with properties we’ve never seen before. Whereas the star clusters we’re most familiar with are either tight, compact, roughly ball-like collections of stars, Gaia discovered more than 1,000 new star clusters that are instead spread out over wide areas: as though they collapsed along filaments, rather than from ellipsoidal clouds of gas.

One of those new star clusters is known as Theia 520, whose stars are approximately 350 million years old. All told, the cluster itself is around 1,200 light-years away, but it’s elongated and widely spread out over space. Instead of compact and rich, it’s diffuse and sprawling. For that reason, it’s a terrible object to view with your eyes through a telescope. However, it is a sparkling example of this new type of cluster. Like many of the new ones, it has tidal tails, diffuse distributions, and features that appear to be driven by evolution. Some of these clusters, in fact, are so elongated that they span over a thousand light-years from end-to-end. Theia 520 is one of them, and these two stars, Kepler 52 and Kepler 968, are actually found on the extreme outskirts of the cluster itself.

The Hyades, the closest star cluster to Earth, may not be what we’ve long thought. Traditionally, we’ve pictured the Hyades as an originally spheroidal star cluster that’s in the process of dissociating, or breaking up, and that’s why its stars are so extended. But with the recent identification of long, filamentary star clusters as perhaps the dominant time, maybe the Hyades is one of these extended, filamentary star clusters instead. (ESA/GAIA/DPAC, CC BY-SA 3.0 IGO; ACKNOWLEDGEMENT: S. JORDAN/T. SAGRISTA)

It’s only because of the fact that we have so many new, cutting-edge observatories that all complement one another that we were able to synthesize this picture together so holistically.

  • From ESA’s Gaia and its view of individual stars, we can get astrometry and photometry data, teaching us the star’s position, color, and a little bit about its motion.
  • From Kepler, TESS, and the Zwicky Transient Facility, we can determine the orbits of the planets around the star, the rotation data of the star, and the mass of the star in question.
  • And from the Keck telescope, the Sloan Digital Sky Survey’s APOGEE instrument, and China’s LAMOST telescope, we can get spectroscopic data, which helps inform us about the star’s metallicity (how many and what kinds of heavy elements are inside) and other detailed stellar properties.

In the modern era of large data sets, one helpful features is that all of these various observatories have already had their data digitized, and it’s all freely and publicly available to researchers anywhere in the world. From this position, a team of researchers led by Dr. Jason Curtis at Columbia University were able to draw some extraordinary conclusions.

Kepler-52, in purple, and Kepler 968, in dark blue, seem relatively unrelated. Both have numerous exoplanets and are found in approximately the same region of the sky, but we never knew that they were part of a large, diffuse star cluster until extremely recently. (JASON CURTIS, MARCEL AGÜEROS, ET AL.)

First off, Kepler 52 and Kepler 968 are, in fact, part of a much larger, enormous, but diffuse star cluster: Theia 520. If they formed from the same cloud of gas, you’d expect them to all:

  • have the same ages to within only a few million years,
  • to follow the same mass vs. rotation period correlation,
  • and to all have approximately the same heavy element content, or metallicities, as one another.

This is precisely what we see. Theia 520 consists of about ~400 stars, strewn about across a large area of the sky. The metallicities of the stars are thus far difficult to obtain, but for the seven different stars where metallicity measurements exist, they’re all consistent both with one another and with having a comparable heavy element fraction to our own Sun. And, as we’ve already seen, they all follow the mass-rotation period correlation we showed earlier, with Kepler 52 and Kepler 968 matching Theia 520 extremely well. This leaves one conclusion as the overwhelmingly favored one: these two star systems, the Kepler 52 and the Kepler 968 systems, are actually siblings of one another.

Four different star clusters and their stars, plotted with rotation period vs. mass. Note how tight the correlations are at hight masses, and how they only start to depart from the main curve at very low masses that haven’t had time to spin down yet. Our Sun, for comparison, spins with a rotation period of 25 days at the equator and 33 days at the pole; it has spun down spectacularly. (JASON CURTIS, MARCEL AGÜEROS, ET AL.)

This is pretty extraordinary! With the rotation periods and masses measured for 130 separate stars in Theia 520 — about a third of the identifiable stars inside — we were able to pin down the ages of the stars inside to extreme precision: they’re 350 million years old, with an uncertainty of only ~50 million years on that figure. This makes Kepler 52 and Kepler 968 incredibly valuable systems, as young planetary systems appear to be rare.

Travel the Universe with astrophysicist Ethan Siegel. Subscribers will get the newsletter every Saturday. All aboard!

In fact, by observing a number of the stars inside Theia 520, we find a remarkable happenstance: the stars within Theia 520 that are home to detected planets are preferentially located on the outskirts of this diffuse cluster, while the stars located closer to the cluster center don’t appear to have planets. While this is just one such cluster with only a few hundred stars, making it difficult to draw broad conclusions, it certainly is suggestive that there might be a greater pattern at play here.

A selection of the globular cluster Terzan 5, a unique link to the Milky Way’s past. Incredibly old stars can be found within globular clusters, relics of some of the first ‘bursts’ of star formation to occur in our vicinity of the Universe. It would not be unsurprising if there were a higher percentage of stars containing exoplanets in the cluster’s outskirts than towards the center. (NASA/ESA/HUBBLE/F. FERRARO)

As Dr. Curtis put it, “This is only the beginning. Gaia has shown that the solar neighborhood is teeming with [these diffuse stellar] populations, some stretching hundreds of light-years in space in elongated patterns, others arranged in more amorphous distributions, and some that are dense clusters with halos and tails. Like Theia 520, some of these groupings are home to already known planets, with many more waiting to be discovered with the ongoing TESS survey.”

By measuring the rotation rates and masses of stars, we can determine their ages to excellent precisions. This new research takes us a step further: into territory where we can identify large, elongated, diffuse star clusters, even ones spread out across more than a thousand light-years, that we can confidently trace back to a single origin in time. It’s proof that we can identify which stars, even stars separated by great distances, were born together, from the same star-forming region. And it offers hope, for perhaps the first time, that if we can gather enough high-quality data, even 4.5 billion years after the fact, we just might be able to someday find our long-lost stellar siblings as well. With the power of massive data sets, open science, a lot of technique and a little luck, we might soon discover that we’re a lot less lonely in the Universe than we’ve ever imagined.

Starts With A Bang is written by Ethan Siegel, Ph.D., author of Beyond The Galaxy, and Treknology: The Science of Star Trek from Tricorders to Warp Drive.


Up Next