2500 light years from Earth in the constellation of Cancer lies Messier 67, an open star cluster that is now known to be home to at least three planets. The new worlds, found using the HARPS spectrograph on the European Southern Observatory’s 3.6-meter instrument at La Silla, come as the result of an observation program covering 88 selected stars in the cluster over a period of six years. The finding is noteworthy because we have so few known planets in star clusters of any kind. Moreover, one of these planets orbits a truly Sun-like star.
Image: This wide-field image of the sky around the old open star cluster Messier 67 was created from images forming part of the Digitized Sky Survey 2. The cluster appears as a rich grouping of stars at the centre of the picture. Credit: ESO/Digitized Sky Survey 2 / Acknowledgement: Davide De Martin.
I’m cautious about calling anything ‘Sun-like’ given how loosely that term has been used over the years, but ESO astronomers say the cluster star YBP1194 fits the bill: It has a similar mass, and shows both chemical abundances and temperatures very close to Sol’s. Of the three discovered worlds, two orbit G-class stars similar to the Sun (the other is YBP1514), while the third orbits the red giant S364. The first two have roughly one-third the mass of Jupiter, orbiting their host stars in seven and five days respectively, while the third, more massive than Jupiter, orbits the red giant in 122 days.
Because most stars are thought to emerge from clusters, the small number of planets found in them has been a puzzle, spurring the recent work, which was led by Anna Brucalassi (Max Planck Institute for Extraterrestrial Physics). Says Brucalassi:
“In the Messier 67 star cluster the stars are all about the same age and composition as the Sun. This makes it a perfect laboratory to study how many planets form in such a crowded environment, and whether they form mostly around more massive or less massive stars.”
Messier 67 contains about 500 stars and is an open cluster, a stellar grouping that has emerged from a single gas and dust cloud in the relatively recent past. Such clusters are normally found in the spiral arms of galaxies like ours. Globular clusters, on the other hand, are the much larger, spherical collections of stars that orbit around the center of the galaxy. Although a handful of planets have been found in open clusters (Messier 44 and NGC 6811 are other examples), no planets have yet turned up in the far more ancient globular clusters.
The lack of detected planets in open and globular clusters has been under discussion for some time now. From the paper:
To explain the dichotomy between field and cluster stars, it has been suggested that the cluster environment might have a significant impact on the disk-mass distribution. Eisner et al. (2008), studying disks around stars in the Orion Nebula Cluster (ONC), proposed that most of these stars do not possess sufficient mass in the disk to form Jupiter-mass planets or to support an eventual inward migration.
Brucalassi’s work, however, leads in a different direction. The paper continues:
van Saders & Gaudi (2011), in contrast, found no evidence in support of a fundamental difference in the short-period planet population between clusters and field stars, and attributed the non-detection of planets in transit surveys to the inadequate number of stars surveyed. This seems to be confirmed by the recent results.
Planets in open star clusters, in other words, are likely to be as common as those around isolated stars, a finding that draws not just from Brucalassi and team’s work but also from a number of recent observations discussed in the paper. The researchers continue to study M67 to examine the mass and chemical makeup of stars with and without planets.
The paper is Brucalassi et al., “Three planetary companions around M67 stars,” accepted for publication in Astronomy & Astrophysics. See also Pasquini et al., “Search for giant planets in M67 I. Overview” (preprint). And take note of Henry Cordova’s “The SETI Potential of Open Star Clusters,” which ran all the way back in 1995 in Vol. 1, No. 4 of SETIQuest, an early and prescient contribution.
Yes they have: the planet orbiting the pulsar/white dwarf binary PSR B1620-26/WD J1623-266 is in the globular cluster Messier 4. Haven’t heard that this claim has been disproven, but to be fair the discovery paper also overlooks this planet in their list of previous cluster planet detections.
And thus we get one step closer to Isaac Asimov’s “Nightfall”….
http://www.uni.edu/morgans/astro/course/nightfall.pdf
However, as Pasquini (of ESO and co-author of the paper) notes, “These new results show that planets in open star clusters are about as common as they are around isolated stars — but they are not easy to detect. The new results are in contrast to earlier work that failed to find cluster planets, but agrees with some other more recent observations. A footnote in the ESO release further emphasizes, “This detection rate of 3 planets in a sample of 88 stars in Messier 67 is close to the average frequency of planets around stars that are not members of clusters.”
If you would be so kind…is this to imply that these open clusters may be islands where civilizations of similar age are positioned at distances that even chemical rockets could traverse? Please say yes!
I’ve always been depressed by the immensity of the universe…that we can’t reach AC in our or anyones lifetime…but at least if there were these “little” high density pockets where life could intermingle, I would somehow take some comfort in that. Call me weird…its what I think about.
A very interesting study indeed.
When i see what HARPS can do i can’ t wait for its successor ESPRESSO to see first light in 2016.
HARPS can detect Radial velocity of about 1m/s, and ESPRESSO on the VLT will achieve around 10 cm/s which is what Earth causes on the sun ( 9cm/s).
And that’s not the end of the story: CODEX on the EELT will achieve 1cm/s.
If intellegent life has taken hold in some of these globular clusters then they could be a good candidates for observation as there will most likely be plenty of electromagnetic leakage.
@coacervate: I share your wishful thinking.
To a lesser extent this could, of course, also be the case for (solar type) binary stars, such as 16 Cygni A and B, Zeta 1 and 2 Reticuli, and …. guess! (if not too close for long-term stable planetary orbits).
Robert Bradbury wrote a paper on how ETI at Kardashev Type II and III level civilizations might use globular star clusters for resources due to the relative closeness of so many star systems. I guess the same would go for open clusters too.
“Globular Clusters and Astroengineering” (July, 2001)
I cannot find the paper I mention as it seems Bradbury’s archives disappeared shortly after his death, a major loss if they are truly gone. I hope someone will correct me on this.
This preserved paper he wrote on Matroishka Brains does contain some relevant discussions here:
http://www.enthea.org/planetp/mbrain1.html
The fact that we can find any exoplanets at all in clusters through our rather crude detection methods shows there are likely far more worlds in those celestial structures than previously thought. Whether they have native life forms or visitors mining resources is another question to answer.
There is the possibility of black holes been quite common in Globular Clusters.
http://msutoday.msu.edu/news/2013/msu-led-research-finds-another-black-hole-in-a-star-cluster/
If they didn’t cause any trouble they would be great for lensing events, they could possibly see us long before we see them.
And their night sky wow.
If I’m doing the math right, Messier 67 is about 4200 cubic ly (10 ly radius) with about 500 stars, that averages to about 8 cubic ly/star, which would imply an average separation of about 2 ly (less toward the center, more toward the periphery). Would that be enough separation to preclude frequent gravitational disruptions and life-destroying blasts of radiation from other stars (including all those red giants)? If so, then the hypothetical inhabitants of some of those Sol-like systems might have to travel only 3 light-years or less to reach another star very similar to their own.
There was a paper last year about survivability or otherwise of planetary sytems and density of stellar environment. Unfortunately the result is that planetary orbits in globular clusters and the galaxy central kpc are too crowded. The orbit of earth like planets around sun like stars would be disrupted long before complex life evolved.
One hole in this though, is that I don’t think M-dwarfs were considered. Maybe planets in tight orbits in the HZ of such stars can survive in dense environments.
Gerry (and kzb):
yes, 2 ly distance is *plenty* of separation for stable planetary orbits. I actually came to almost 2.5 ly average distance.
In fact, I find this distance disappointingly great: even in our own part of the MW galaxy the average distance between two stars is roughly 3 to 4 ly.
So, with respect to distance, there is not much advantage in being in an open cluster. On the other hand, this particular open cluster, M67, contains a large proportion of solar type stars of medium age, that makes it more interesting.
I’m not sure these stellar density calculations are particularly useful.
Firstly the question is what you are using for the definition of the cluster radius – half-mass radius, tidal radius, core radius?
Secondly you neglect the central condensation of the cluster, which is probably the far more important effect.
I really don’t think you can use those calculations to conclude that the open cluster environment does not have particularly different stellar separations to the field.
ljk cited Asimov’s “Nightfall”. Two others are worth citing:
Isaac Asimov and Robert Silverberg collaborated on a novel-length retelling of the story.
http://www.amazon.com/Nightfall-Isaac-Asimov/dp/0553290991
And Dean McLaughlin wrote a novel with his own version of the story.
http://www.amazon.com/Dawn-Dean-McLaughlin/dp/B0051BSLAO/ref=sr_1_3?s=books&ie=UTF8&qid=1389993923&sr=1-3
Maybe a star system not originally part of the cluster could be gravitationally captured by the cluster. If it’s on the outer edges, it might not be disrupted quite as much as a system inside the cluster.
@stephen January 17, 2014 at 17:30
‘Maybe a star system not originally part of the cluster could be gravitationally captured by the cluster. If it’s on the outer edges, it might not be disrupted quite as much as a system inside the cluster.’
It is quite possible for a globular cluster to capture a star system as they have been around a long, long time. I can’t see there been great disruption to a planetary system once it has formed and then been captured even with the high star density it should still hold on to them.
It is an OPEN cluster, so the stars move together around the Milky Way, they don’t circle each other. Which makes me ask:
How come an open cluster can have the same age as the Sun? The Sun has made about 15-20 orbits around the Milky Way by now and the open cluster it once belonged to has dispersed all over the galaxy.
@Glaas January 18, 2014 at 12:47
‘It is an OPEN cluster, so the stars move together around the Milky Way, they don’t circle each other. Which makes me ask:
How come an open cluster can have the same age as the Sun?
They may have been more tightly bound in the past and plus we don’t see the ones that have broken apart.
‘The Sun has made about 15-20 orbits around the Milky Way by now and the open cluster it once belonged to has dispersed all over the galaxy.’
It is possible that the Alpha Centauri system was part of the cloud that the sun was born from, a twin if you like.
http://www.esa.int/Our_Activities/Space_Science/A_cool_discovery_about_the_Sun_s_next-door_twin
Off topic if there were intelligent alien life in these globular clusters would they look at the black holes that frequent them as potential gods.
@Gerry January 16, 2014 at 21:27
‘If I’m doing the math right, Messier 67 is about 4200 cubic ly (10 ly radius) with about 500 stars, that averages to about 8 cubic ly/star, which would imply an average separation of about 2 ly (less toward the center, more toward the periphery).
The density of the stars at the centre is on the order of the solar system! what a sight it would be to stand on a planet in orbit around one of these core stars.
http://www.eso.org/public/news/eso0107/
Messier 67 is a very unusual open cluster, as it has never dispersed over time. Almost every other cluster that formed so many billion years ago has long-since separated into individual stars. We shouldn’t be surprised to find mature planets in this cluster. Whether they would be habitable is another matter.
In the case of M67, its orbit takes it relatively far out of the galactic plane. It therefore spends far less of its time in the disruptive environment of the galactic disc which is probably why it has survived so long.
The paper did not preclude planets in stable orbits in open clusters -as we see them now. I think there was a question mark over the early environment in such clusters, when they were more compact.
I don’t know about M67 specifically. If the stars are in a bound system (are they?), and their orbits regularly take them through a central region where the stellar density is as high as Michael says, then the survivability of HZ planetary orbits is questionable at least. If the stars stay more or less where they are relative to each other, then any planets that survived the early days of the cluster should be OK.
Michael said on January 18, 2014 at 15:00:
“Off topic if there were intelligent alien life in these globular clusters would they look at the black holes that frequent them as potential gods.”
Among the first questions I would ask an ETI is if they have religion in any form similar to ours? Did they too “worship” anything that seemed beyond their understanding in the early development? Or is this a trait unique to humans?
As for a culture flipping out if they suddenly found themselves in a universe with many, many suns instead of just a few, how plausible is that? Yes it makes for a dramatic story but would an entire species go collectively nuts from suddenly seeing a lot of stars in their sky?
Thanks to light pollution we are losing the night sky over Earth with every passing day. When people who hardly ever see the stars visit a region where they have truly dark skies and suddenly see many stars, the reaction is one of awe and wonder, not panic and fear. Or at the least curiosity.
This is the paper I have been referring to:
Habitability in Different Milky Way Stellar Environments: a Stellar Interaction Dynamical Approach
http://arxiv.org/abs/1306.0464
“The SETI Potential of Open Star Clusters”, by Henry Cordova, Vol. 1, No. 4 , SETIQuest, (1995 ).
At that time, I was able to identify about 30 open clusters over a billion years old (time enough for life to evolve).
Many more old OCs have been described in the literature since that time. There is a professional monograph covering much of the same ground I did in my popular SETIQuest article, but in much more detail.
Reprints are available online from Springer, for a price.
Seti in Star Clusters: A Theoretical Approach (2003)
R. De La Fuente Marcos
C. De La Fuente Marcos
http://link.springer.com/article/10.1023%2FA%3A1023345929739#page-1
The De La Fuente Marcos paper is cited by Tartar and Turnbull in their SETI targeting contributions.
http://en.wikipedia.org/wiki/Catalog_of_Nearby_Habitable_Systems
I’ve always had this fantasy that old OCs singled out for SETI searches would someday be called “Cordova Clusters”.
Henry Cordova
hank at qmss dot com
Gerry asks “If I’m doing the math right, Messier 67 is about 4200 cubic ly (10 ly radius) with about 500 stars, that averages to about 8 cubic ly/star, which would imply an average separation of about 2 ly (less toward the center, more toward the periphery)”
And no one has answered properly!
No Gerry – I think you are coming from the world of chemistry where that calculation would be true for cubic packing (where six nearest neighbors pop up simultaneously). If the distribution was a PERFECTLY random one stars per 8 cubic ly and we surrounded a star by a sphere of 2ly then the expected number of neighbors would be 4 NOT 1. The expected number for just one is 1.25, but the distribution is not random so the median closest separation must be less than this, probably far less.
That said, I agree with Andy over the given radius for a cluster being problematic in of itself.
It’s not surprising that planets are found circling stars in open clusters as often as they are around solitary stars. All stars are born in open clusters, that is, molecular clouds are the birthplace of all stars, and it is only after sufficient stars commence fusion reactions that the residual gas and dust is b;own away by the stellar winds of those stars .
Later, the cluster may evaporate into individual stars as it is subject to disruptive tidal forces from the galactic disc. Those stars ejected from the cluster will carry their retinue of planets with them, as did our own sun.
Although most open clusters do break apart after a few million years, those that do survive are therefore likely to have stars with planets. This is precisely what makes old OCs suitable targets for SETI investigators and planet hunters.
The only mechanism I can think of that would favor the disruption of solar systems among cluster stars is close gravitational encounters with other stars in the cluster. Although these objects appear crowded on astrophotographs and through the eyepiece, the spacing of stars within a cluster (approximately 1 light year) may not be close enough for this to happen too often. Has anyone actually done the numbers on this?
Alternatively, it could be argued that the eccentric orbits of many exoplanets are the result of just such interactions, close encounters of these stars with other cluster members soon after they formed. I would suspect that ordered, well-behaved planetary orbits as in our own solar system should be the norm, not the exception, a result of unperturbed system formations. The fact so many high-eccentricity orbits are detected may be evidence of chaotic and turbulent conditions in young clusters, evidence which survives even after the star is ejected or the cluster breaks up.
Then again, perhaps the prevalence of high eccentricity orbits is just a selection effect: high eccentricity orbits are easier to detect.
Clearly, more theoretical work needs to be done.
@Gerry January 16, 2014 at 21:27
‘If I’m doing the math right, Messier 67 is about 4200 cubic ly (10 ly radius) with about 500 stars, that averages to about 8 cubic ly/star, which would imply an average separation of about 2 ly (less toward the center, more toward the periphery).’
Rob Henry is quite right it is not a simple density model, we can see that in the pictures, but a much more complicated one based on factors such as,
1) The masses of the individual stars
2) The number of stars
3) And the age of the system
to name a few.
That is why people are finding it difficult to reply. I would have thought it would be based on an angular momentum/potential energy relationship. A battle between potential energy wanting to collapse it all like a ballerina death hug versus the angular momentum of a drunk on the dance floor trying to throw it apart.
‘Would that be enough separation to preclude frequent gravitational disruptions and life-destroying blasts of radiation from other stars (including all those red giants)?’
This depends on the age and density of the cluster and the mass of their constituent stars, the more massive the star the quicker it burns. I like your thinking though on the red giant stage stars, they can have quite an impact on the local space environment.
Don’t forget the Virial Theorem, a mathematical relationship expressing the distribution of potential and kinetic energy in mechanical systems. I’ve forgotten most of the math involved, but it is used to examine the kinematics and dynamics of star clusters.
http://en.wikipedia.org/wiki/Virial_theorem
The average stellar density IS a term in the equation for encounter rate. Another one is the velocity dispersion, in this case of the cluster members relative to each other.
The case of M67 is discussed specifically in the reference I gave above. This paper agrees with what many here have said, there is no problem with stable planetary orbits in this open cluster.
There is a possibility with globular clusters that I find fascinating: there could be a large population of “delocalised planets”. A great many planets are thought to be ejected or caused to leave their birth system by close encounters in such clusters.
The problem of planets in globular cluster stars is a particularly interesting one. GC’s are old, very old, they appear appear to have formed along with the galaxy itself. Their stellar populations are composed of stars that condensed before metals enriched the interstellar medium, and it is unlikely that earthlike or rocky planets circle stars in GCs. Any planets likely to be found there will be gas giants, primarily hydrogen and helium.
GC stars are very old, and have no doubt created heavy elements in their cores through nucleogenesis, but the clusters themselves are relatively gas- and dust-free–not much of that material has been recycled back to the interstellar medium through supernovae, planetary nebula, stellar winds, or other highly evolved stars. Alas, Asimov’s “Nightfall” scenario is not a likely one.
Henry Cordova, I believe that these ideas about GCs have been revised. A lot of them show significant metallicity and stellar evolution. Also bear in mind that the only correlation on firm footing is that between gas giants and metallicity. There is no such correlation with smaller planets, at least until a very low metallicity is considered. It’s more likely the gas giants don’t exist rather than the terrestrial planets.
I think the Nightfall scenario probably is unlikely with sun-like stars. Planets in their HZs would be in unstable orbits. But I wonder about M-stars with their tight HZs, maybe they are stable.
Thanks, kzb
This new thinking about rocky planets in Globular Clusters is good news for SETI enthusiasts, especially now that red dwarfs have been shown to have their share of earthlike planets in their slim habitable zones.
A GC with a million or so very old stars might be a very good place to look for other civilizations, and cultures arising there there would have plentiful opportunities to travel to nearby stars. Very old GCs may very well be highly populated with intelligent beings.
What might it look like being inside M67:
http://www.slate.com/blogs/bad_astronomy/2014/01/15/exoplanet_planet_found_orbiting_a_solar_twin_in_a_star_cluster.html
To quote:
“I have to wonder, too … what does the sky look like for that planet? Well, the star occupies nearly half of it, so it’s a blazing hellfire of light and heat. Right, so let’s let our minds take a step further out, literally, and suppose there’s an Earth-like planet orbiting that star at a more comfortable distance.”
What The Night Sky Would Look Like From Inside A Globular Cluster
George Dvorsky
Yesterday 3:20pm
Globular clusters are incredibly dense structures often featuring hundreds of thousands of stars packed into a relatively small space. So what would it be like to live inside such a thing? A team of astronomers recently created a simulation to find out.
For their simulation, astronomers William Harris and Jeremy Webb designed a hypothetical Earth-like world orbiting a single main-sequence star — but one located smack-dab in the middle of 47 Tucanae, a globular cluster located 16,700 light-years from Earth. This cluster packs 570,000 stars in an expanse of space that’s a “mere” 120 light-years across.
Harris and Webb sent me a sneak-preview of an article that’s set to appear in the July edition of Astronomy. Here’s what they have to say about the remarkable perspective from an Earth-like world:
At the center, our planet would be surrounded by a few hundred stars per cubic light-year (several thousand stars per cubic pc), which is thousands of times the stellar density of the Sun’s neighborhood in the Milky Way’s suburbs. The typical distance from our hypothetical planet to the closest star, however, still would be substantial — about 0.05 light-year (0.015 pc). In our solar system, this would place it beyond the inner edge of the Oort Cloud of comets.
Unless the closest stars happen to be red giants, none of them would have angular diameters large enough to resolve with the human eye, so all the stars still would appear as points of light. Across the entire sky, inhabitants of our hypothetical world would see 10,000 stars brighter than 1st magnitude — compared with just 29 in Earth’s sky — and more than a thousand brighter than Earth’s most brilliant nighttime star, Sirius. The brightest suns would blaze at apparent magnitudes brighter than –9, or 100 times more luminous than Venus appears from Earth. More than 130,000 stars would shine brighter than 6th magnitude, the naked-eye limit, compared with 6,000 from Earth.
Although it might sound like lots of empty space still exists at the cluster’s center, the prospects for doing astronomy from there would be discouraging. The biggest problem would be the sheer amount of light from all those stars. The cluster’s suns would combine to give an average sky brightness some 20 times brighter than Earth’s night sky at Full Moon (or about 16.7 magnitudes per square arcsecond). In other words, the darkest night our viewers would ever see would be a strange sort of twilight that possesses a kind of grainy texture unlike the uniform sheet of light we see on Earth. The galaxy’s disk — already hard to see from Earth at Full Moon except from isolated locations — would be visible in the background but hard to study. Astronomers on our hypothetical planet likely would favor telescopes with small fields of view and excellent baffling against scattered light.
Full article here:
http://io9.com/what-the-night-sky-would-look-like-from-inside-a-globul-1589324556