22% of Sun-like Stars have Earth-sized Planets in the Habitable Zone

How common are planets like Earth? That’s been a question astronomers and dreamers have pondered for decades, and now, thanks to the Kepler spacecraft, they have an answer. One in five Sun-like stars in our galaxy have Earth-sized planets that could host life, according to a recent study of Kepler data.

“What this means is, when you look up at the thousands of stars in the night sky, the nearest sun-like star with an Earth-size planet in its habitable zone is probably only 12 light years away and can be seen with the naked eye. That is amazing,” said UC Berkeley graduate student Erik Petigura, who led the analysis of the Kepler and Keck Observatory data.


The Kepler telescope’s mission was to try and find small rocky planets with the potential for hosting liquid water and perhaps the ingredients needed for biology to take hold. For four years, the space telescope monitored the brightness of more than 150,000 stars, recording a measurement every 30 minutes.


Analysis by UC Berkeley and University of Hawaii astronomers shows that one in five sun-like stars have potentially habitable, Earth-size planets. (Animation by UC Berkeley/UH-Manoa/Illumina Studios)

For a recent focused study, scientists concentrated on 42,000 sun-like stars (G and K type stars), looking for periodic dimmings that occur when a planet transits — or crosses in front of — its host star. A team of scientists from the Kepler mission and the Keck telescope in Hawaii have announced that from that survey, they found 603 planets, 10 of which are Earth size and orbit in the habitable zone, where conditions permit surface liquid water.

Since there are about 200 billion stars in our galaxy, with 40 billion of them like our Sun, noted planet-hunter Geoff Marcy said that gives us about 8.8 billion Earth-size planets in the Milky Way.

But Marcy also cautioned that Earth-size planets in Earth-size orbits are not necessarily hospitable to life, even if they orbit in the habitable zone of a star where the temperature is not too hot and not too cold.

“Some may have thick atmospheres, making it so hot at the surface that DNA-like molecules would not survive. Others may have rocky surfaces that could harbor liquid water suitable for living organisms,” Marcy said. “We don’t know what range of planet types and their environments are suitable for life.”

Analysis of four years of precision measurements from Kepler shows that 22±8% of Sun-like stars have Earth-sized planets in the habitable zone. If these planets are as prevalent locally as they are in Kepler field, then the distance to the nearest one is around 12 light-years.Credit: Petigura/UC Berkeley, Howard/UH-Manoa, Marcy/UC Berkeley.
Analysis of four years of precision measurements from Kepler shows that 22±8% of Sun-like stars have Earth-sized planets in the habitable zone. If these planets are as prevalent locally as they are in Kepler field, then the distance to the nearest one is around 12 light-years.Credit: Petigura/UC Berkeley, Howard/UH-Manoa, Marcy/UC Berkeley.

All of the potentially habitable planets found in their survey are around K stars, which are cooler and slightly smaller than the sun, Petigura said. But the team’s analysis shows that the result for K stars can be extrapolated to G stars like the sun.

The Kepler spacecraft is now crippled because of faulty gyroscopes, but scientists say had Kepler survived for an extended mission, it would have obtained enough data to directly detect a handful of Earth-size planets in the habitable zones of G-type stars.

If the stars in the Kepler field are representative of stars in the solar neighborhood, then the nearest (Earth-size) planet is expected to orbit a star that is less than 12 light-years from Earth and can be seen by the unaided eye. Future instrumentation to image and take spectra of these Earths need only observe a few dozen nearby stars to detect a sample of Earth-size planets residing in the habitable zones of their host stars.

“For NASA, this number – that every fifth star has a planet somewhat like Earth – is really important, because successor missions to Kepler will try to take an actual picture of a planet, and the size of the telescope they have to build depends on how close the nearest Earth-size planets are,” said Andrew Howard, astronomer with the Institute for Astronomy at the University of Hawaii. “An abundance of planets orbiting nearby stars simplifies such follow-up missions.”

Further reading: Institute for Astronomy, University of Hawaii; UC Berkeley; Keck Observatory; NASA; PNAS.

19 Replies to “22% of Sun-like Stars have Earth-sized Planets in the Habitable Zone”

  1. The caveat buried in the middle of the article: “But Marcy also cautioned that Earth-size planets in Earth-size orbits are not necessarily hospitable to life.”

    In addition to the admitted life-hospitable factors which are not known about those planets, here are some more, but not all, factors to be determined:

    Suitable axis rotation rate; appropriate tilt of planet axis for seasons; magnetic field from planet molten core; sufficient planetary plate teutonics; suburban galaxy location providing sufficient actinide radioisotopes for decay heat; suitably sized moon to create needed tides; giant outer gas planets to clean up loose asteroids and comets; sufficient distance from recent supernovas.

  2. As mentioned in other comments, we really DO need to launch another planet finder mission to continue revising the Drake equation. While we’re at it, a mission to look for nearby red dwarf stars with earth like planets orbiting them would be great! Highly evolved red dwarf stars might be the place to go if you were in need of metals and other ‘rare earth’ elements, presumed to exist in such locations. Heck, for all we know, THAT might have happened/be happening today, in our solar system? (Shhh… don’t tell the earthlings we want their gold, uranium, platinum and etc. They might want to be paid for it!)

    1. Red dwarf stars, atleast on the lower end, only fuse hydrogen to helium and then stops. They also have lifespans of 20billion years and upwards, meaning no red dwarf has yet existed long enough to be ‘highly evolved’.
      Havent you confused red dwarf with red giant?

      1. I believe he means that red dwarf stars usually have a higher % of metals than do sun like stars.

        The same can be said of sun like stars having a % of metals than stars which are much larger.

      2. Right you are magnus… but there are SO many of them there’s bound to be exceptions! Like those formed in binaries or dense molecular clouds… 50-75 billion of em in the Milky Way!

  3. Hopefully not take many decades to reach those planets. I would like to be live when it happens. 🙂

  4. Maybe I’m missing something obvious, but those numbers don’t add up. If they surveyed 42,000 stars and concluded that 22% of them had earth sized planets in the habitable zone, then hey must have found 42,000 • .22 = 9240 such planets. But they found only 10. That’s only .023%. Could someone please explain this discrepancy?

    1. They did not conclude 22% from the survey of 42,000 stars. The 22% conclusion was was made from calculations based on data they recorded from 150,000 stars. The 42,000 star survey was a more focused study in which they only located 10 earth-like planets (approx size w/in habitable zone). The data suggests that there should be more such planets among those 42,000 star systems, but this does not mean you will be able to directly detect all of those planets. Keep in mind this article is just summarizing the findings without explaining all the factors that led to the conclusion, so if your only source of information is this article then it can be rather confusing. I recommend looking at the actual study findings for more information (the article linked some of its sources).

      1. Even on the study of 150000 stars the numbers still do not stack up. And why would only 10 ELP’s turn up in the more focussed study?. Is this because there were too many false positives?. The article in no way explains the conclusion about 8.8 billion ELP’s. Prof Brian Cox is referring this article in his facebook page which is why I read it. It always worries me that such extrapolations are being made on relatively small datasets. Our galaxy is not uniform in its structure and planetary systems are incredibly varied, and more stars exist as doublets and triplets than our own singular Sun. I would be very cautious about such stats until we can directly view these ELP’s, esp around red dwarf stars.

    2. I agree. If only 10 planets are Earth like in habitable zone then extrapolating 42000 to 200 billion still only means approx. 47 million ELP’s,NOT 8.8 billion, out by several orders of magnitude. And this article is being endorsed by Prof Brian Cox, no less!. Someone has screwed up here me thinks!…

      1. I think the apparent discrepancy in percentages reported is due to the method of detection used by the Kepler telescope. The transit method requires the planetary system under observation to be edge on. The probability of this condition being met for Earth like orbits is around 0.5%. Hence the proportion of Earth like planets detected will represent a much smaller proportion than those in existence.

  5. I have looked up the most recent data on Kepler from the NASA website and there is something seriously wrong with these numbers in this article. I think someone has been misquoted and the actual numbers of predicted ELP’s are at best in the millions, NOT billions. Kepler data is constantly being modified to exclude, for example, eclipsing binaries, stars with sunspot/flare activity, which is REDUCING the total numbers of ELPs predicted. I also read an updated Drake equation based on Kepler data which shows we may actually be the ONLY civilisation in this epoch transmitting radio signals in the galaxy, which kinda fits with our more subjective experience that we are indeed alone. One other aspect to all this is even if 22% stars have ELP’s, that still means the majority DONT, and therefore finding life as we know it is unlikely….

  6. Hi Steve, yes I agree regarding the edge-on aspect of the Kepler method for planetary detection, but even so this is not explained in the article. Taken at face value the numbers are still wildly optimistic. Big assumptions, like for example there will be homogeneity in the way other star/planetary systems are distributed. Is the plane of other systems in line with the galactic plane for example?. Edge on allowed I calculate no more than 10 million possible ELP’s, with your % adjustment raising that to about 2 billion absolute maximum. Of course these are just extrapolations. No statistician would support such numbers until at least 5% of the galaxy stars had been observed. Prof Cox does beg the question why if so many ELP’s there is no radio chatter across the galaxy, not one jot. That must question this data too…

    1. Hi Ivor. You make some good points there. I must get around to reading the original article. I think surveying 5 percent of the stars in the Galaxy to get a representative sample is too big an ask given the number of stars involved. Do you mean a sample taken from the Galaxy as a whole? Given at least the suggestion that habitable zone planets might be fairly common, I’d concentrate further searches on our local neighbours because if we found any we’d stand a better chance due to their proximity of seeing if biomarkers were present.

      With regard to radio-chatter, I’m not sure if we should expect too much. Even if life is common in the Galaxy, and that’s a big if at the momient, intelligent technological life spewing out radio waves detectable on Earth could be very uncommon or even non-existent, despite the Fermi paradox. I read somewhere that radio chatter from Earth is diminishing as we use more satellite communication, Internet, and mobile phones and, if there are technological civilisations out there, they could have gone through the same development path.

  7. The 5% comes from basic statistical methods of number crunching ie a p value. I agree on the targeted search for ELP’s in our “local” star neighbourhood, say within 100 light years. Finding ELP’s further away seems pretty pointless – its not as though a meaningful telecoms conversation could be had!. As for radio chatter we have been emitting, deliberately and otherwise, quite powerful signals, radar and the like. I guess the question to be answered is ” if we do find an ELP within say 50 light years should we send a specific radio message to that planet?”…..( and then in 100 years time wake up to the Independence Day scenario!)…cheers…

  8. You might wish to watch the Scienceof Dr Who with Brian Cox which was aired on BBC last night. He deals with the Fermi paradox and other issues of ET. Interesting point he makes about technological civilisations, in that any kind of machinery gives off heat, and its this signature of infra red that we are beaming out into space and have done since the industrial revolution. Its not just about radio waves. Apparently a team at Penn State Uni are examining data on IR emmissions from candidate stars/planets for these tell tale signs od technological development. So maybe global warming has a value after all!….cheers…

Comments are closed.