Nader Haghighipour. American Scientist. Volume 96, Issue 4. Jul/Aug 2008.
Imagine that Earth had two suns-an arrangement that might make for some spectacular sunrises and sunsets, among other oddities. Planets with multiple suns have long been standard fare in science-fiction stories and movies. So fans of this genre should be pleased to learn that over the past few years investigators have indeed found planets in such star systems, albeit planets that are much larger than Earth. These discoveries took many of us in the astronomical-research community by surprise.
Our astonishment didn’t stem from any lack of interest in the topic. Indeed, people have been thinking about planets around other stars for centuries. Way back in 1584, Giordano Bruno, a Dominican monk, suggested that “innumerable suns exist; innumerable earths revolve around these suns in a manner similar to the way the seven planets revolve around our sun.” He was tried by the Inquisition and burned at the stake for making the assertion.
Others were later more free to pursue the idea, and some began looking for planets in other parts of the universe more than 150 years ago, but not until 1992 did these efforts bear fruit. By studying the time interval between radio pulses received from an object known as PSR B 1257+12, Aleksander Wolszczan of the Pennsylvania State University and Dale A. Frail of the National Radio Astronomy Observatory identified cyclic back-and-forth movements and surmised from them that this pulsar has three planetary-mass bodies around it.
Wolszczan and Frail’s discovery was groundbreaking, yet because the central star is a pulsar, it was clear that these distant worlds must be bathed in lethal radiation, leaving no prospect that they could be hospitable to life. But in 1995 Michel Mayor and Didier Queloz of Observatoire de Genève uncovered a planet circling a normal, Sunlike star, one known as 55 Pegasi. Astronomers have now detected more than 270 planetary systems, many of which are so different from our solar system that the current theories of planet formation are unable to explain how they came to be.
Around the Sun, small, rocky planets occupy close-in orbits, and gaseous giant planets are situated at larger distances. But many of the newly discovered planetary systems contain Jupiter-like objects roasting in orbits that are smaller than that of Mercury. Also, Earth and the other planets of the solar system travel around the Sun in near circles, whereas the orbits of many extrasolar planets are highly elliptical.
Among the other surprises that astronomers have turned up is the existence of planets in binary star systems, places where two stars are gravitationally bound together and orbit around their common center of mass. Observations indicate that most stars-at least the normal ones-are located in binaries or in clusters of three or more. (Pulsars are well-known exceptions to this rule.) So at first blush it may appear only reasonable that planets would be found in dual-star systems. What’s more, investigators have uncovered planet-forming clouds of dust and gas in and around some of these systems. So why is the recent discovery of binaries with planets so astonishing? Because for many years theorists had been telling observational astronomers that planets cannot form in such environments. And curiously enough, their reasons for saying so seemed perfectly sound.
In general, planets are a side effect of the process that creates stars, which starts with the collapse of a molecular cloud (a diffuse blob of gas and dust) and culminates when the temperature and pressure at the center become sufficient to support the nuclear fusion reactions that give stars their light. This radiation and the wind of gas given off from the newly ignited star clear away most vestiges of the original molecular cloud. All that remains is a thin disk, aligned with the star’s equator. This platter of gas and dust, the birthplace of planets, is known as a nebula (not to be confused with the term “planetary nebula,” which means something else entirely to astronomers).
The formation of planets takes place in steps. First, dust particles in the disk gently approach one another and stick together to form larger aggregates, which in turn collide and sweep up smaller material until they grow to several kilometers in size. These objects, known as planetesimals, smash into one another and in the process form bodies as big as the Moon or even Mars, which astronomers call protoplanets or planetary embryos.
At larger distances from the star, where temperatures are lower and the nebula retains some of its original gas, planetary embryos may grow much larger and form the cores of giant planets, which eventually become so massive that they attract a large portion of the gas from their surroundings and develop gaseous envelopes. (Gas giants, such as Jupiter and Saturn, are believed to have formed in this way.) At closer distances, where the nebula has lost a large portion of its gas, planetary embryos continue to collide and grow for several hundred million years, eventually forming a small number of rocky, terrestrial planets.
For such steps to take place, the nebula must, of course, contain a sufficient amount of stuff. Traditionally, theoreticians believed that the minimum nebular mass needed to form a planetary system similar to our own is approximately 1 percent of the mass of the Sun. Computational simulations showed, however, that this condition might not be fulfilled around binary stars.
Some binary stars loop around each other in highly elliptical orbits; in those cases, the distance between the two stars varies considerably over time. When the spacing between them is small, the gravitational tug of one star disturbs the nebular material surrounding the other, causing the orbits of the various bits and pieces to become elliptical. With each close approach of the two stars, the degree of ellipticity increases until the nebular material is thrown completely out of the gravitational field of its host star. In this way, one star in a binary system depletes the nebular disk around the other star. And what remains may not be enough to build up into planets.
A stellar companion can thwart the formation of planets in a second way, too. The tendency of one star to increase the ellipticity of the orbits of objects around the other causes asteroid-sized chunks to approach one another with large relative speeds. The resulting collisions can be so severe that, instead of sticking together to create something bigger, things fragment.
Knowledge of such effects had, for many years, lulled theoreticians into thinking that planets form only around single stars-or around stars so far separated from any stellar companion that they can be considered isolated. And based on the Copernican principle (that Earth holds no special status in the universe), astronomers surmised that the process of planet formation necessarily had to produce analogues of our solar system. That is, they expected to find giant planets situated at large distances and smaller ones close in, with everything moving in approximately circular orbits. They were wrong on all counts.
Waking Up to Two Suns
The theoretical issues that seemed to rule out planets in binary systems arise only if the two stars experience relatively close approaches. Computer simulations have shown the critical distance to be approximately 80 to 100 astronomical units (one astronomical unit, or AU, is the distance between Earth and the Sun, about 150,000,000 kilometers). Although there has never been any theoretical prohibition against finding planets around binaries of greater separation, the culture of seeking “planetary systems similar to our own” had been so strong that it constituted the foundation for how planet-hunting astronomers carried out their observational campaigns. For these reasons, the candidates for planet searches were routinely chosen from lists of single stars or of widely separated binaries.
Today, among the stars known to have planets, 25 percent are members of binary systems. However, except for a few examples, the separations of these binaries are between 250 and 6,000 AU. Such large separations allow astronomers to apply the same techniques they use to find planets around single stars (principally the radial-velocity method, which uses Doppler shifts to detect the back-and-forth wobble that a planet induces in the star it orbits) to the detection of planets in binary systems.
Despite the focus of most planet hunters on single stars, very early on some Canadian astronomers stumbled on a planet within a closely spaced binary system. Although it was widely recognized as a planet only in 2003, it is arguably the first one ever detected around another star, because evidence for it was collected initially even before Wolszczan and Frail had their first inklings of pulsar planets.
The story goes like this: In 1988, Bruce Campbell of the University of Victoria, along with Gordon A. H. Walker and Stevenson Yang (both then at the University of British Columbia), measured variations in the radial velocity of the star gamma Cephei A and concluded that it hosts a Jupiter-sized planet. Gamma Cephei A, a 3 billion-year-old star of approximately 1.6 solar masses, has a small stellar companion, gamma Cephei B, a brown-dwarf star. Their separation is 18.5 AU, roughly the distance from the Sun to Uranus.
Finding a planet similar to Jupiter orbiting gamma Cephei A was surprising, to say the least. Indeed, it was so surprising that it did not withstand the skepticism of its own discoverers. In 1992, Walker and his colleagues announced that what they had originally thought to be an oscillatory variation of the radial velocity of gamma Cephei A was just mundane activity in a part of that star’s atmosphere called the chromosphere. They thus retracted what would have been the unveiling of not just the first planet in a binary system, but the first extrasohr planet ever!
It took another decade of monitoring gamma Cephei A to conclude that the original detection of a planet around this star was, in fact, legitimate. In 2003, Artie P. Hatzes of the Thüringer Landessternwarte Tautenburg, along with Campbell, Walker, Yang and four other colleagues, published a paper in the Astrophysical Journal showing that gamma Cephei A does, in fact, host an object that is about twice the mass of Jupiter.
The discovery of this and other planets in binary star systems with modest separations has spurred much research. Currently there are three such systems known: gamma Cephei, GJ 86 and HD 41004. Whereas the first two provide clear examples of a dual-star system with a planet orbiting one star, the configuration of HD 41004 is so unusual that it is hard to know how exactly to classify it. What was initially thought to be a planet orbiting the smaller star (a body dubbed HD 41004 Bb) is now thought to be itself a small star. But this triple-star system does have a planet: HD 41004 Ab, which orbits the largest of the three stars.
In addition to this peculiar find, observers have discovered binary stars with moderate separations that have retained enough of their original nebulae to trigger the formation of planets at some point in the future. Luis F. Rodriguez of the Universidad Nacional Autónoma de México and several colleagues uncovered one example in 1998, a binary star system embedded within a molecular cloud known as L1551. The stars of this system are 45 AU apart, and each has a disk that is about 20 AU wide and contains 6 percent of the mass of the Sun-wide and massive enough one day to produce both gas giants and terrestrial planets.
The recognition of planets and protoplanetary disks in moderately close binary star systems has presented astronomers with many questions. In particular, how exactly do planets form in such environments? And might the conditions on some of these planets be amenable to supporting life?
Whether potentially habitable worlds can exist in a given binary system depends fundamentally on each star’s ability to provide its planets (and planet-forming materials) with stable orbits. When interstellar separations are small, the influence of one star on the bodies circling the other is substantial. This effect becomes even more pronounced when the orbits of the stars around their center of mass are highly elliptical.
For a planet to maintain a stable orbit around one star of a binary-a so-called “S-type” (satellite type) binary-it has to escape the gravitational influence of the other star. It can do so only if its radial distance from its parent star is much smaller than the separation between the two stars of the binary. In the case of gamma Cephei A, the planet is at a distance of 2.1 AU from the primary star, whereas that star’s binary companion is located almost 10 times farther away. So the planet has a stable orbit. But numerical simulations have shown that a different planet at a substantially larger radial distance would soon be ejected from the system.
What about planets that travel around both stars of a binary (so-called circumbinary planets)? What conditions are required for their long-term health? Such a planet will be secure in its orbit if its distance from the center of mass of the two stars is much larger than the separation between them. Circumbinary planetary systems, also known as P-type (planetary type) binaries, are more likely to be stable when the two stars have small separations, because if they are far apart, the orbit of the planet has to be much wider still. And in that case, the tug of passing stars can destabilize it.
How do S- and P-type binaries form in the first place? The widely accepted core-accretion model described above, in which a planet similar to Jupiter or Saturn is born through the growth of its core and the subsequent attraction of a gaseous envelope, has been able to account for giant planets in binary systems. For example, the application of such numerical models to the stars of gamma Cephei results in a giant planet-but one with an orbit that is slightly different from what has been observed.
Astronomers sometimes invoke an alternative model, called “disk instability,” in which Jupiter-like planets assemble themselves rapidly in a gravitationally unstable nebula. The disk-instability model also seems able to account for how such bodies form in binary systems. However, astrophysitists do not all agree on how to interpret these findings. Indeed, how planets form in and around binary star systems is still a topic of ongoing observational and theoretical research.
Life Under Two Suns
Whatever the impediments to understanding their formation, it is now abundantly clear that planets exist around some modestly separated binaries. Might such worlds support life forms similar to the ones we know about? In other words, might they be habitable?
For astronomers to classify a planet as habitable, it must satisfy certain physical requirements. Because all life needs liquid water (at least as far as anyone knows), the planet must orbit its central star at distances where the star’s illumination is sufficient to prevent water from freezing and at the same time is not so intense as to cause all of it to vaporize. The annular region around a star where an Earthlike planet receives this much light is called, naturally enough, the habitable zone. Clearly, the orbit of a habitable planet cannot deviate too much from a circle, because a highly elliptical orbit would cause the planet to spend long periods outside this zone of comfort.
The location and width of the habitable zone are functions of a star’s size and brightness, and also vary with the planet’s ability to support an atmosphere, which can boost the temperature at the surface considerably through the greenhouse effect. Having an atmosphere also helps to even out variations in temperature from place to place and from day to night.
For a planet to maintain an atmosphere, it has to be sufficiently massive that its gravitational attraction does not allow gas molecules to escape too quickly. That requirement in turn implies that a habitable planet cannot be too small (Mercury, for example, isn’t big enough to hold an atmosphere for very long). But it cannot be too large either: A planet the size of Jupiter will have a very thick atmosphere, which may not permit enough illumination to shine through and will not allow the solid surface to cool off. So the surface will therefore be hot and under high pressure-two properties that are unfavorable for development of life.
Such large, gaseous planets would, however, be expected to have moons, which might well be able to support life. The possible habitability of such moons is a sexy subject in astronomy. In our solar system, Europa, one of the larger moons of Jupiter, has attracted a tremendous amount of attention in this regard. Within the community of astronomers focused on extrasolar planets, the habitability of the moons of giant planets is also a very hot topic. Many of us are now forging techniques to detect such distant satellite worlds.
Another important feature of a habitable planet is its ability to undergo plate tectonics. This requirement may seem odd: Why, after all, should extraterrestrial life need a world with earthquakes and volcanoes? The short answer is that by recycling carbonate and silicate rock through the planet’s interior, plate-tectonic motions act to stabilize the atmosphere over geologic time. Without them a planet would be prone to experiencing either a runaway greenhouse effect or a globe-encircling glaciation. And for a planet to undergo plate tectonics, it has to have a mass between 0.5 and 10 Earth masses. (Although a size in that range appears to be necessary, it isn’t sufficient to ensure that plate tectonics operates-Venus being a nearby counterexample.)
The understanding of what constitutes a habitable planet is, obviously, shaped by theories about how Earth came to be. Astronomers in general accept the notion that Earth and other terrestrial planets of the solar system formed through collisional growth of protoplanets. Because the only interaction among these objects was through their mutual gravitational attraction, and because gravity is naturally a weak force, it took a long time (several hundred million years) for terrestrial planets to emerge.
One important question, which is still not fully answered and which constitutes one of the major scientific mysteries of planetary science, concerns how water appeared on Earth. When Mercury, Venus, Earth and Mars were first forming, the inner solar system was subject to intense radiation from the Sun, making this region too hot for water to exist. Why then does our planet have so much water?
Some astronomers believe that water was delivered to Earth some time after its formation, carried in by comets, which originated in the outer parts of the solar system where water in the form of ice was (and is) abundant. This explanation suffers from a major drawback, however: Measurements of the ratio of deuterium to ordinary hydrogen in Earth’s ocean waters do not match the ratio found in comets.
A better understanding of what actually happened comes from numerical modeling of the process, which suggests that the water on Earth was incorporated while the planet formed. It probably came from small protoplanets that initially resided in the outer region of the asteroid belt-a wide reservoir between Mars and Jupiter containing many kilometer-sized bodies left over from the original solar nebula. These objects were far enough from the Sun to retain appreciable amounts of ice. Eventually, they found their way to the inner part of the solar system, where they crashed into Earth, adding enough water to cover most of the planet with several kilometers of ocean.
Why did this icy material move from the outer part of the disk inward toward Earth? The cause of that life-giving migration was the largest planet of our solar system, Jupiter, which had many dose approaches with the protoplanets circling in the outer asteroid belt. The resulting gravitational interaction sent many of them, along with the water they contained, headed toward Earth.
A similar phenomenon might have given birth to Earthlike worlds in extrasolar planetary systems. Numerical modeling of the process shows that the key for our solar system was the small ellipticity of Jupiter’s orbit. In Systems where the giant planet moves in a near-circular orbit, inward water delivery through this mechanism is quite efficient. However, if the ellipticity of the giant planet’s orbit is large, close approaches between this object and various protoplanetary bodies cause many of them to be ejected from the system, leaving the final set of terrestrial planets with little or no water.
So to form Earth-sized planets with substantial amounts of water in die habitable zone of a binary star system, a few conditions have to be satisfied. First, the separation of the binary needs to be at least moderate (say, 20 to 40 AU). Also, the two stars should move in low-ellipticiry orbits about their common center of mass. And having a giant planet with an almost-circular orbit located outside the habitable zone, although not strictly necessary, speeds the delivery of water where it is needed for life.
Discovering Sister Worlds
The detection of Jupiter-like planets in binary star systems with moderate separations is very challenging. The most widely used method, the radialvelocity technique, may not have the necessary precision to separate the effects of a planet from those of a nearby stellar companion. A better approach may be to use transit photometry, which measures the diminution in the brightness of a star when a planet passes in front of it. But like the radialvelocity method, transit photometry has trouble distinguishing the effects of a planetary object from those of a companion star. At the moment, neither technique is capable of discerning Earth-sized objects around single stars, so finding such bodies in dual-star systems will be even more vexing.
Despite the many challenges, planet-hunting astronomers have been busy trying to improve the situation. Several international groups are hoping to develop detection methods that are primarily for the purpose of identifying planets in binary star systems. Contributors to a coordinated exoplanet search in Italy, for example, have applied the radial-velocity method to binaries using data obtained from the Italian Telescopic Nazionale Galileo, which is located on La Palma, one of the Canary Islands. Doing so requires more calibration measurements than are normally taken when single stars are observed. Interpreting the collected spectra (a computationally challenging step in any radial-velocity search) becomes that much more difficult as well.
These Italian investigators plan to use photometry, too, to look for S-type planets in binary star systems. The basic strategy is to examine eclipsing binaries-systems in which one star periodically passes in front of the other. By carefully timing these eclipses, it should be possible to sense the presence of planets orbiting around one star, which make it wobble slightly as it orbits the other.
Also, since 2006 a satellite called CoRoT (shorthand for Convection, Rotation and planetary Transits) has been put to the task of finding extrasolar planets through transit photometry. And NASA’s upcoming Kepler mission, which is scheduled to begin in 2009, will use a space-based telescope for the specific purpose of detecting watery, Earthlike planets around other stars. At least some astronomers will surely want to use these powerful space telescopes to look for S-type planets in eclipsing binaries. And with such tools, it’s only a matter time before a habitable world with two suns makes the final leap out of the realm of science fiction and into reality.