Microlensing Study Says Every Star in the Milky Way has Planets


How common are planets in the Milky Way? A new study using gravitational microlensing suggests that every star in our night sky has at least one planet circling it. “We used to think that the Earth might be unique in our galaxy,” said Daniel Kubas, a co-lead author of a paper that appears this week in the journal Nature. “But now it seems that there are literally billions of planets with masses similar to Earth orbiting stars in the Milky Way.”

Over the past 16 years, astronomers have detected more than 3,035 exoplanets – 2,326 candidates and 709 confirmed planets orbiting other stars. Most of these extrasolar planets have been discovered using the radial velocity method (detecting the effect of the gravitational pull of the planet on its host star) or the transit method (catching the planet as it passes in front of its star, slightly dimming it.) Those two methods usually tend to find large planets that are relatively close to their parent star.

But another method, gravitational microlensing — where the light from the background star is amplified by the gravity of the foreground star, which then acts as a magnifying glass — is able to find planets over a wide range of mass that are further away from their stars.

Gravitational microlensing method requires that you have two stars that lie on a straight line in relation to us here on Earth. Then the light from the background star is amplified by the gravity of the foreground star, which thus acts as a magnifying glass.

An international team of astronomers used the technique of gravitational microlensing in six-year search that surveyed millions of stars. “We conclude that stars are orbited by planets as a rule, rather than the exception,” the team wrote in their paper.

“We have searched for evidence for exoplanets in six years of microlensing observations,” said lead author Arnaud Cassan from the Institut de Astrophysique in Paris. “Remarkably, these data show that planets are more common than stars in our galaxy. We also found that lighter planets, such as super-Earths or cool Neptunes, must be more common than heavier ones.”

The Milky Way above the dome of the Danish 1.54-metre telescope at ESO's La Silla Observatory in Chile. The central part of the Milky Way is visible behind the dome of the ESO 3.6-metre telescope in the distance. On the right the Magellanic Clouds can be seen. This telescope was a major contributor to the PLANET project to search for exoplanets using microlensing. The picture was taken using a normal digital camera with a total exposure time of 15 minutes. Credit: ESO/Z. Bardon

The astronomers surveyed millions of stars looking for microlensing events, and 3,247 such events in 2002-2007 were spotted in data from the European Southern Observatory’s PLANET and OGLE searches. The precise alignment needed for microlensing is very unlikely, and statistical results were inferred from detections and non-detections on a representative subset of 440 light curves.

Three exoplanets were actually detected: a super-Earth and planets with masses comparable to Neptune and Jupiter. The team said that by microlensing standards, this is an impressive haul, and that in detecting three planets, they were either incredibly lucky despite huge odds against them, or planets are so abundant in the Milky Way that it was almost inevitable.

The astronomers then combined information about the three positive exoplanet detections with seven additional detections from earlier work, as well as the huge numbers of non-detections in the six years’ worth of data (non-detections are just as important for the statistical analysis and are much more numerous, the team said.) The conclusion was that one in six of the stars studied hosts a planet of similar mass to Jupiter, half have Neptune-mass planets and two thirds have super-Earths.

This works out to about 100 billion exoplanets in our galaxy.

The survey was sensitive to planets between 75 million kilometers and 1.5 billion kilometers from their stars (in the Solar System this range would include all the planets from Venus to Saturn) and with masses ranging from five times the Earth up to ten times Jupiter.

This also shows that microlensing is a viable way to find exoplanets. Astronomers hope to use other methods in the future to find even more planets.

“I have a list of 17 different ways to find exoplanets and only five have been used so far,” said Virginia Trimble from the University of California, Irvine and the Las Cumbres Observatory, providing commentary at the American Astronomical Scoeity meeting this week, “I expect we’ll be finding many more planets in the future.”

Sources: Nature, ESO, AAS briefing

26 Replies to “Microlensing Study Says Every Star in the Milky Way has Planets”

  1. I have written a document – a theory on why life occurs in the universe, the natural balance and perhaps an answer to the meaning of life. Find it at internetname.ws to read it. And if you find it plausible – share it with the world. This is important.

    1. BAM!:

      “[With Mjölnir, Thor] would be able to strike as firmly as he wanted, whatever his aim, and the hammer would never fail,”

    2. There is a lot of such theories. Not enough tests to pick one in particular.*

      * Though apparently in the last years many researchers like Morowitz and Woolf has circled around ATP/AMP cofactors as the route to the RNA world.

      ATP function is also a root in gene phylogenies, so it looks consistent. And phylogenies is one important test.

  2. Given our own system has 8 planets, it’s tempting to think 8 is probably an average.

    If 8 is well above average or well below average, then it’s either a freak circumstance or intelligent life is somehow dependant on the number of planets (which is true to at least some extent: intelligent life as we know it is going to require a minimum of one planet, and some argue quite compellingly that the presence other planets, Jupiter in particular, are necessary for Earth to have become habitable)

    1. First, the average could be as low as four and yet eight planets could be fairly common place — we would need to see the statistical distribution (the bell curve) before we could come to any conclusion as to how close to the norm our solar system comes.

      Second, while there are theories that point to the necessity of there being a Jupiter-sized planet in a solar system before life finds a home, there is still almost no data to back it up. Already most of what we thought we knew about the evolution of solar systems has been overturned by the data coming from Kepler and other sources, and I would be shocked if there weren’t a whole load more surprises in store.

      The great thing about Kepler and the other present and future missions and surveys of exoplanets is that eventually we will have thousands, if not tens or hundreds of thousands of examples to study. Then, and only then, will we have a good understanding of just how special–or not–our own system is.

      1. Good point, I wasn’t thinking very hard when I wrote that 😛 The bell curve could peak anywhere (even zero, though this latest result suggests at least 1), and the slope could be steep or shallow.

        True that there is “almost no data”, but I find the theories quite compelling. I seem to recall seeing an animation, some time ago, that described the migration of large planets early in the solar system’s history, and which showed how Neptune might have migrated outward (due to an orbital resonance between Jupiter and Saturn, if I recall correctly) and kicked countless comets into the inner solar system, which would have then imbued Earth with water.

    2. While there seem to be a nice distribution of planet sizes, the diversity of systems points to many routes. The systems that have many planets are mostly found by Kepler, and by necessity of the lingering observation time bias they are still relatively packed. Maybe 8, or the original 9 that recent Nice models adaptations puts as the preferred model for our own system’s formation, is way below the median.

      We won’t know until, and if, Kepler gets a few years extended mission. If then even, see tacitus’ comment.

      The Late Heavy Bombardment comes and goes, but remaining is that most realistic rates are survivable. In that case, why would a system having more late impactors than ours have less of habitable planets?

      We don’t know how rapid abiogenesis is, it may well happen in between impacts in such systems. Why, the impactors would both feed renewed heat energy and nutrients to the planet for that to happen. You could even wait in a long and too dense tail of impactors before it kicks in.

      No, I think the question is still too loosely constrained. Life seems easy enough to get from chemical evolution, many potential pathways have been offered. Linguistically intelligent life _is_ believed to be a one off freak by most biologists, a trait akin to the elephant trunk.

      However, the one off potential is not yet testable, plenty of complex traits have been reinvented a number of times. (Say, photosynthesis or eyes.) Nor do we have a rarity of planet systems where evolution potentially can play out and hit that trait again.

  3. I am probably dense, but I do not understand the logic of the argumentation used in the article. They write they surveyed 3,247 micro-lensing events (or actually only 440?), where they found three planets (the other seven were from earlier works). And from this they conclude that planets are more abundant than stars in our galaxy? Somehow it does not add up for me. Not that I doubt the abundance of planets in our galaxy, but basing such claim on the discovery of three planets in 3000 events (or 440?) seems a bit odd to me.

    Perhaps I misunderstood something, and should better read the original work. Can anyone here put some more light on it?

    1. No, it’s not explained very well in this post. From reading the paper, it sounds like the authors calculated the odds of a planet producing a measurable difference from the lensing effect of a star without planets (given its mass, orbital size, orientation, position, etc.), and found that it’s unlikely they would have detected the three that they did if one or more planets is not the norm for a given star. In other words, it’s a bit of a long shot for things to be just right for a planet to produce a detectable lensing effect, so the fact that they got the three that they did means that planets must be very common.

      1. This is what I gather as well. The planet is the gravitational microlens. The occurrence of the caustic spike in the image of a back ground star must be due to a dark body, such as a planet. The occurrence of these is used as a prior probability in probably Bayes formula, as seen on page 3 of the paper.


    2. I think that 440 of those events were usable and then if you find at least 3 planets, that’s statistically very significant because the method is limited.

  4. One medium-sized Spiral Galaxy of 100 billion organized suns, among a vast web-network of organized Galaxy filaments, turning, spinning and twisting through the expanding space of receding time.

    A pattern-formation not unlike the complex of neuron pathways along connected filaments in the human brain of 100 billion organized cells, animating intelligent mind.

    If this sampling has any validity of true star-world number, be they 4, 7, or 8 in family systems, then “100 billion exoplanets in our galaxy” of familiar and unfamiliar kind.

    Though this curious recurrence of the 100 BILLION number, through vastly increasing scales of size, may be of mirage-like significance ( a nice round figure ), I wonder if it is not a recurring numerical-pattern, reappearing through increased measures of dimension ( with some underlying relationship? ).

    On the final mega-scale, if you were to subtract dwarf “galaxies”, might the final figure of large, or super galaxies – once again – display the sum of 100 billion?

    ( Well, perhaps not: there may be 200 billion stars in the MW – and a trillion true “galaxies” in observable space, if not its entire inflating realm of place. )

    On question of estimation, this was interesting, in comparing number of stars to shoreline grains of sand:

    “If we count the number of grains in a small representative volume of sand,
    by multiplication we can estimate the number of grains on the whole beach.”
    – ESO on number of stars

    It would only be an estimation: Can man count the number of stars in the sky, or the grains of sand on a seashore – could he ever capture the sum-total of star-worlds in the Cosmos, in orbital revolutions of year and planetary-turns of day?

    Just one sand grain of thought that may lie buried on an untrod beach.

  5. According to a similar article I read in Science Daily, the study concluded that on average there is at least one planet per star in the Mikly Way, not that every star has a planet.

  6. ““We used to think that the Earth might be unique in our galaxy,” said Daniel Kubas”

    As amazing as the microlensing technique admittedly is, this statement is simply ridiculous. I can think of no one who seriously suggested that this star system might be unique, hosted by a star invisible to the human naked eye merely 50 light years away.

    1. I think by “We” he means humanity as a whole, not just his team (or the scientific world). Of course, humanity has never been in agreement on cosmic plurality; there certainly has been plenty of people who seriously think the Earth is unique – scientists and non-scientists – from the ancient Greeks all the way up to the modern day. It’s easy to disagree with them, but you can’t deny they exist.

  7. 3 questions!
    Probably, in order to detect planets using the micro-lensing method it will miss planets circling stars because of the angle of attack from our point of view… or not?
    If planets do not pass through the stars light how can we detect them?
    Of course this does not include planets which we can actually see through a telescope, right?

  8. Ok, does anybody know the other 12 ways of detecting exoplanets, apart from microlensing, spectral analysis, direct observation and transits?

      1. “I have a list of 17 different ways to find exoplanets and only five have been used so far,” said Virginia Trimble from the University of California, Irvine and the Las Cumbres Observatory, providing commentary at the American Astronomical Scoeity meeting this week, “I expect we’ll be finding many more planets in the future.”

      2. I think in that case one has to contact Trimble for her explanation.

        I have a list of 17 different ways to cook pasta, but only 1 or 2 have been used so far. (O.o)

      3. Thanks! I don’t know why I assumed there wasn’t a paper to search for; there is nearly always a paper. I guess the “list” meme threw me off.

        I think my previous comment’s joke came out poorly. I didn’t mean they were unuseful, merely not the daily dish right now.

        Come to think of it, since the discoveries goes exponentially, those methods will have heavy competition right now. But sooner or later the strength of having complementary techniques will be telling.

Comments are closed.