Strange New Worlds: The Search for Alien Planets and Life Beyond Our Solar System - Ray Jayawardhana (2011)

Chapter 6. Blurring Boundaries

Neither Stars nor Planets

Over the past two decades, astronomers have uncovered a surprising variety of worlds in the outer realms of our solar system and beyond. Of the thousands of icy bodies circling the Sun beyond the orbit of Neptune, the biggest few—all found in the twenty-frst century—resemble Pluto in many ways. For a while there was talk of inducting a tenth planet, but Pluto was demoted instead, stirring a public debate that got emotional at times. The revelations at the other end of the planetary-mass range are perhaps even more profound. At the same time as the extrasolar planets, astronomers have discovered a whole new class of objects called brown dwarfs that span the mass gap between stars and planets. They are probably born like stars but end up with characteristics similar to Jupiter. Just as Pluto’s counterparts did at the low-mass end, brown dwarfs are challenging our defnition of what constitutes a planet at the high end. But there is no reason to despair at this cosmic identity crisis. In this case, the current confusion signifes the stunning progress we have made in unraveling brave new worlds, which don’t ft neatly into our old narrow pigeonholes.

a controversial demotion

Gatherings of astronomers rarely make front-page news. But the 2006 general assembly of the International Astronomical Union in Prague was different. A media storm was already brewing in the weeks leading up to it. That’s because of one single item on the conference agenda—a vote on a new defnition of the word “planet.” After months of consultation and deliberation, an IAU-appointed committee had put forth a carefully worded proposal. Yet it failed by a vote of 18 to 50 in a straw poll on August 18. What followed were several days of alternative proposals, town meetings, and “secret” negotiations. Finally, a revised version of the new defnition was adopted on August 24, with a considerable majority of those assembled in favor. Given all the commotion around the world, it’s noteworthy that only 434 of the more than 2,500 astronomers attending the general assembly showed up for the vote.

The new defnition sounds innocuous enough: A planet is a celestial body that (a) is in orbit around the Sun, (b) has suffcient mass for its self-gravity to overcome rigid body forces so that it assumes a hydrostatic equilibrium (nearly round) shape, and (c) has cleared the neighborhood around its orbit. The cause of all the frenzy was the implication from part (c) that Pluto, considered a planet for more than seventy-fve years, is no longer. In fact, the IAU made it explicit: Pluto is a dwarf planet by the above defnition and is recognized as the prototype of a new category of trans-Neptunian objects.

The reactions to this apparent demotion ranged from the emotional to the humorous, from the somber to the silly. Schoolchildren wrote letters to prominent astronomers calling them heartless or worse. Cartoonists and late-night comedians poked fun at Pluto being sent to the cosmic doghouse. The self-styled “Friends of Pluto,” including the widow and son of its discoverer Clyde Tombaugh, protested in Las Cruces, New Mexico, carrying placards that read “Size Doesn’t Matter.” The state representative for Dona Ana County, Tom-baugh’s longtime home, introduced a resolution in the New Mexico legislature affrming Pluto’s planethood. A bipartisan resolution in the Golden State went further, expressing concern for the “psychological harm to some Californians who question their place in the universe and worry about the instability of universal constants” and complaining about this “unfunded mandate” to revise school textbooks. Alan Stern, principal investigator for NASA’s New Horizons spacecraft on the way to Pluto, told a reporter that he is “embarrassed for astronomy.”

Personally, I found it diffcult to get worked up either in defense of the new “planet” defnition or in opposition to it. After all, the need for revision had been building up for years—since 1992, to be exact. That’s when David Jewitt and Jane Luu, two astronomers using the University of Hawaii’s 2.2-meter telescope on Mauna Kea, discovered a faint slow-moving object dubbed 1992 QB1. Follow-up observations confrmed it as a member of the Kuiper Belt, a population of icy bodies beyond Neptune, frst hypothesized almost ffty years earlier as a reservoir of short-period comets. Since then, astronomers have identifed over a thousand other such bodies, and it became increasingly clear that Pluto belongs to the same population. With the discovery of several large Kuiper Belt objects in recent years, some more than half the size of Pluto and harboring moons just like it does, Pluto’s special status was under threat. The issue came to a head in 2005, when Michael Brown at Caltech and his colleagues found a body, later (fttingly) named Eris after the Greek goddess of discord, estimated to be not only bigger than Pluto but also more massive. Now the scientifc community had little choice: they had to either elevate Eris (and others like it) to planet status or drop Pluto. If they chose the frst option, the ranks of solar system planets might swell into the dozens pretty quickly. So it made sense to demote Pluto, for the sake of consistency and simplicity.

As I did several television and radio interviews in Toronto in August 2006, I tried to put a positive spin on things rather than mourn Pluto’s demise. I talked about the stunning new fnds in the outer reaches of our solar system that had prompted the revision; I explained that instead of “losing” Pluto, the solar system has “gained” a whole new class of dwarf worlds. In one interview, I mentioned the ongoing debate over where to draw the line at the high end of the planetary-mass scale—an issue the IAU had sidestepped for the moment—and argued that instead of lamenting these identity crises, we should celebrate the dramatic advances that reveal a remarkable diversity of worlds both in our solar system and beyond.

Bridging the Gap

The distinction between stars and planets used to be simple: stars shine by their own light, generating energy through nuclear fusion, whereas planets do not. But, since the mid-1990s, astronomers have discovered a new class of objects, called brown dwarfs, that blur these boundaries.


Figure 6.1. Comparison of the sizes of the Sun, a red dwarf star, a brown dwarf, Jupiter, and the Earth.

These lilliputian bodies inhabit the mass gap between dwarf stars and giant planets and share some characteristics with both. Astronomers often lump together brown dwarfs and giant planets as “substellar objects.” Weighing in at less than 80 Jupiters (or about 8 percent of the Sun’s mass), brown dwarfs are not massive enough to burn hydrogen steadily. But those with masses above thirteen times that of Jupiter are hefty enough to briefy ignite deuterium, a heavier form of hydrogen, in their cores when young. Once the minute supply of deuterium is exhausted, they cool down over time, just like planets. In contrast to planets, however, most brown dwarfs are found in isolation, though a few do orbit stars. This ambivalence between stellar and planetary qualities extends to other facets. For example, brown dwarfs start out big, with radii as large as low-mass stars, but over time they shrink down to smaller than Jupiter. Strangely, more-massive brown dwarfs end up with smaller radii than less-massive ones. That’s because their size is set by the balance between the inward pull of gravity and the outward quantum pressure exerted by densely packed electrons inside—a quantity that behaves rather differently from normal gas pressure holding up stars. What’s more, the atmospheres of young, hot brown dwarfs closely resemble those of stars like the Sun; but, as they cool with age, weather phenomena such as clouds and rain, usually seen in planets, develop in these atmospheres. In essence, brown dwarfs are star-like when young but become “planet-esque” as they get older.

In fact, some cool brown dwarfs display strong evidence of changing weather patterns. In 2009, Etienne Artigau of the Gemini Observatory and his colleagues reported that one brown dwarf varied in brightness by up to 7 percent every 2.5 hours. The depth and profle of this variation changed over time and nearly disappeared by the following year. The best explanation is that this object spins on its axis with a 2.5-hour period, bringing brighter and darker patches of its atmosphere into our view; those patches themselves evolve over time, much like clouds in the Earth’s atmosphere. Theorists expect tiny dust grains in the brown dwarf’s atmosphere can clump together to form opaque clouds that block light from regions below. To produce the brightness variations seen in this object, its cloud deck must have large holes. Jacqueline Radigan, a PhD student working with me at the University of Toronto, is searching for weather on a sample of about ffty nearby objects to determine how common such weather patterns are and to understand their characteristics and origin. She has already discovered the biggest brightness changes ever seen in a cool brown dwarf—30 percent variations over a rotation period of about eight hours. The size and shape of the variations are such that a giant cyclone system, akin to Jupiter’s Great Red Spot, could be responsible. As she put it, “we might be looking at a massive storm raging on this brown dwarf.” Similar storms could also form on extrasolar giant planets.


Figure 6.2. The big changes in the brightness of the brown dwarf 2M2139+0220 may be due to a giant storm raging in its atmosphere. Credit: J. Radigan (University of Toronto) et al.

The Long Chase

Brown dwarfs were predicted to exist long before they were found. A young Indian-born astrophysicist named Shiv Kumar, with a newly minted PhD from the University of Michigan, frst discussed them in August 1962 at a meeting of the American Astronomical Society in New Haven, Connecticut. Based on theoretical calculations regarding the internal structure of stars, he reported that if any stars were to form with masses below a certain critical value, their core temperatures and densities would be too low for hydrogen fusion. He called them black dwarfs and correctly estimated the threshold to be about 8 percent of a solar mass. A paper on his fndings appeared in the Astrophysical Journal the following year.

The moniker we use today comes from Jill Tarter, better known for her pioneering efforts in the search for extraterrestrial intelligence (SETI) and often considered the role model for Jodi Foster’s character in the movie Contact.The only child of a former pro-football player, Tarter grew up a tomboy in upstate New York. She adored her father, who passed away when she was twelve. After completing a degree in engineering physics at Cornell University, she stayed on to take graduate courses. The chance enrollment in a class on star formation inspired her to switch to astronomy. It was while she was a graduate student at the University of California at Berkeley in the early 1970s that Tarter coined the term “brown dwarf.” “Some people think it’s an awful name,” she told a journalist recently, “but we couldn’t get a good idea of the object’s color temperature. Since brown isn’t a color, we named it that.” For better or for worse, the name has stuck.

Observational searches for brown dwarfs started in earnest in the mid-1980s, with the advent of infrared detectors for common astronomical use. Most early surveys targeted nearby stars in the hopes of fnding substellar companions in their midst. By 1988, two interesting candidates had emerged. One was a faint red companion to a white dwarf, the dense cinder of a burned-out star, with the catalog number GD 165. Two astronomers from the University of California at Los Angeles, Eric Becklin and Ben Zuckerman, found it in an infrared imaging survey. Its inferred mass is close to the boundary between stars and brown dwarfs, but it’s not possible to tell which category the object belongs to. The other candidate was around the solar-type star HD 114762. David Latham of the Harvard-Smithsonian Center for Astrophysics and his colleagues found it in a Doppler survey: the parent star’s wobble implied an unseen companion at least eleven times as massive as Jupiter. Unfortunately, the Doppler method can only give a lower limit for the companion’s mass, so it could be either a star with very low mass or a brown dwarf.

Other astronomers targeted young star clusters, picking out the faintest and reddest (thus coolest) objects in them. Several research teams announced brown dwarf “discoveries,” which did not hold up after closer scrutiny. Some turned out to be red giant stars in the same line of sight, cool because they are bloated and faint because they are far away, thousands of light-years behind the cluster. Others were low-mass stars that didn’t belong to the cluster: they were mistaken for brown dwarfs because of their dimness, which was in fact due to older ages rather than very low masses. These retracted claims fostered skepticism among many researchers who came to view all new announcements with suspicion. For a while, brown dwarf hunting wasn’t the most reputable of astronomical pursuits.

Gibor Basri, a tenured professor at Berkeley, got into the game nonetheless. Born in New York City to immigrant parents from Iraq and Jamaica who had met at the International House while attending Columbia University, Basri grew up in Fort Collins, Colorado. His father was a physics professor at Colorado State University, while his mother taught dance. As a boy, he devoured science fction books. So it was quite a thrill for him, at fourteen, to meet the legendary writer Arthur C. Clarke in Sri Lanka, where his father was on a Fulbright lectureship. (For a year Basri attended the same school as I did—Royal College in Colombo—and I also met Clarke at age fourteen; I wasn’t much of a sci-f fan though.) He decided to study physics like his father when he grew up. But, during college, Basri realized that his true passion lay in astronomy. He studied magnetic activity on stars for his PhD thesis at the University of Colorado, went to Berkeley for a postdoctoral fellowship, and stayed on as a faculty member.

Basri was focusing on young stars and their accretion disks when a paper by three Spanish astronomers based in the Canary Islands caught his attention in 1992. The researchers, led by Rafael Rebolo, proposed a clever new method—the lithium test—to tell brown dwarfs apart from low-mass stars. Even the puniest stars burn up this fragile element fairly quickly, whereas objects below about 60 Jupiter masses never get hot enough for lithium fusion. Thus, if lithium were present in an object older than 100 million years, it would have to be substellar. Basri realized that he could exploit the newly commissioned Keck I 10-meter telescope to take high-quality spectra of faint brown-dwarf candidates to look for this telltale sign. For the frst couple of years, his team’s efforts met with repeated failure.

Bumper Crop

All of that changed in 1995, not only for Basri but for the entire feld, as a result of multiple discoveries made almost simultaneously. John Stauffer, then at the Harvard-Smithsonian Center for Astrophysics, found a faint Pleiades member, dubbed PPl 15 (for ffteenth candidate in the Palomar Pleiades survey), which he passed on to Basri and his colleagues for confrmation. In a Keck spectrum, Basri’s team identifed the signature of lithium, proof that PPl 15 is indeed a brown dwarf. Meanwhile, in a deep survey of the 120-million-year-old Pleiades star cluster, perhaps better known as the Seven Sisters, Rebolo’s group in the Canary Islands found an object so faint and so cool compared with its stellar neighbors that it had to be a brown dwarf. They called it Teide 1, after a Spanish observatory, and inferred a mass just under sixty times Jupiter’s. Later the Canary Islands researchers teamed up with Basri to confrm the presence of lithium in Teide 1 and in a third candidate dubbed Calar 3, after another Spanish observatory.

A search targeting nearby low-mass stars also struck gold at the same time. Astronomers from Caltech and Johns Hopkins University were using the Palomar 1.5-meter telescope in southern California, equipped with a coronagraph that blocked most of a star’s light, allowing them to detect faint companions nearby. They had observed several brown dwarf candidates in 1993 and took a second set of images a year later. If a candidate is a true companion gravitationally bound to the star, the two should move across the sky in tandem, as measured against the background of much more distant stars. One of the confrmed companions was a thousand times fainter than its primary Gliese 229, which itself is a low-mass star, thus almost certainly substellar. But the team kept their fnding under wraps until they had an infrared spectrum of it in hand.

Finally they announced the discovery at the Ninth Cambridge Workshop on Cool Stars, Stellar Systems and the Sun held in Florence, Italy, in October 1995— the same event at which Michel Mayor from Geneva Observatory reported the frst extrasolar planet around 51 Pegasi. Their paper in Nature, led by Tadashi Naka-jima, Benjamin Oppenheimer, and Shrinivas Kulkarni, came out the following month. The substellar nature of Gliese 299B, as the companion was designated, was indisputable not only because it was so faint, implying a mass only thirty to forty times that of Jupiter, but also because its spectrum contained the telltale absorption features of methane. The methane molecule is common in the atmosphere of giant planets but does not form in stars, because they are too hot. Most astronomers consider Gliese 299B to be the frst defnitive brown dwarf discovered, in part because it was the coolest and least massive of the initial crop. At an age of a few billion years, its surface temperature is about 630 degrees Celsius, much cooler than the least luminous stars (at 1500 degrees Celsius). The Pleiades objects, on the other hand, are a bit heftier at 50–70 Jupiter masses, and a lot hotter with surface temperatures of 2300–2500 degrees Celsius, because they are much younger and have not had time to cool down.

Other discoveries followed suit in the next few years. Rebolo’s team identifed several other brown dwarfs in the Pleiades as well as a 25-Jupiter-mass companion to a nearby young star. Maria Teresa Ruiz of the University of Chile and her colleagues found the frst free-foating brown dwarf in the solar neighborhood—named Kelu-1, from a word for “red” in the language of the Mapu-che people native to her country. These “feld” brown dwarfs are located all over, intermingled with stars, so it takes red-sensitive surveys of large swathes of the sky to fnd them. With the Two Micron All Sky Survey (2MASS), which used a telescope each in Arizona and Chile for mapping essentially the entire sky in the near-infrared, astronomers found hundreds of brown dwarfs. The Sloan Digital Sky Survey, which imaged a fraction of the northern sky at optical wavelengths, added to the count, including objects cool enough to show signatures of methane in their midst. Ongoing surveys are looking for even cooler dwarfs—those with ammonia in their atmospheres and not much hotter than Jupiter.


Figure 6.3. Spectrum of Gliese 229B (top) shows features due to molecules like water and methane in its atmosphere, also seen in Jupiter’s spectrum (bottom). Courtesy: B. R. Oppenheimer (American Museum of Natural History)

Current observations suggest that brown dwarfs are common, perhaps a third or a quarter as numerous as stars. Thus our Galaxy alone would contain tens of billions of them. Since they are faint and hard to detect, some astronomers had speculated initially that brown dwarfs could be an important constituent of the “dark matter” that dominates the Galactic mass budget. That doesn’t appear to be the case: even a hundred billion of these lightweights would not add up to much.

Shrouded Origins

As their ranks have grown, so has our interest in unraveling the origin of these peculiar objects. One obvious clue is that brown dwarfs are common as isolated, free-foating objects both in young star clusters and in the feld, but relatively rare as companions to stars. For example, Doppler velocity surveys have revealed a “brown dwarf desert” within a few astronomical units of Sun-like stars, even though they should be easier to detect than less massive planets. That fnding hints at two different formation mechanisms for giant planets and brown dwarfs.

Astronomers think it is more likely that brown dwarfs form in a manner similar to stars, from the denser parts of interstellar clouds, known as “cores.” There is a problem, however. For a core to start contracting under its own weight, it must have suffcient mass to overcome the outward force of gas pressure. That minimum mass is roughly the mass of the Sun. It’s possible that a core breaks up into several smaller fragments during its collapse. But these stellar embryos continue to grow as material rains down on them, so what prevents them from becoming full-fedged stars? Theorists struggle to explain their arrested development.

In 2001, Bo Reipurth of the University of Hawaii and Cathie Clarke of Cambridge University suggested that brown dwarfs are the victims of sibling rivalry. In their scenario, multiple stellar embryos in a core compete to accrete matter, and the one that grows slowest is at the others’ mercy. Gravitational interactions eject it from the core, cutting it off from the gas reservoir and stunting its growth. Computer simulations by Matthew Bate at the University of Exeter and his collaborators indeed show that very-low-mass objects are often kicked out of nascent stellar systems. Interestingly, in Bate’s simulations, the ejected embryos can originate in two different ways—either directly from the breakup of dense cloud cores or from fragmentation of massive protostel-lar disks around existing embryos. That latter route is reminiscent of the disk instability model for giant planet formation (see chapter 2).

Meanwhile, Paolo Padoan (then) at the University of California at San Diego and others have proposed an alternative. Some cores can be much smaller than a solar mass, they argue, because turbulent motions within the molecular cloud can trigger gravitational collapse. In other words, a small core that would not collapse on its own might be induced to do so when compressed by turbulence. Brown dwarfs can then form directly from ultra-low-mass cores. This scenario eliminates the need to invoke ejection to stop embryos from growing, because the mass reservoir itself is limited.

The two theoretical models predict somewhat different properties for brown dwarfs. In the turbulence scenario, whatever is true of low-mass stars should also be true of brown dwarfs. On the other hand, if the ejection hypothesis holds, brown dwarfs would rarely come in pairs, unlike stars of higher mass. That is because such binaries—except perhaps the most tightly bound ones—are likely to be torn apart as they are kicked out of the birth cocoons. One might also expect disks around brown dwarfs to be pruned by close interactions within a multiple system and not live very long.

These differences offer ways to distinguish between the two formation mechanisms. For example, we can search for disks around newborn substellar objects with infrared observations. Since dust grains in a disk absorb light of the central brown dwarf and reemit the energy at longer wavelengths, objects with disks appear brighter in the infrared than those without. My colleagues and I, among others, have surveyed large numbers of brown dwarfs in nearby star-forming regions and clusters, looking for this infrared excess. We fnd that disks are in fact ubiquitous around brown dwarfs at the age of a few million years. Indeed, in a young cluster of a given age, the percentage of brown dwarfs with disks is comparable to the percentage of stars with disks. In short, brown dwarfs harbor disks as often as stars do. Using some of the largest ground-based telescopes as well as the Spitzer space telescope, several groups, including ours, have shown that brown dwarf disks manifest properties similar to those of stellar disks, and that they live for at least as long, if not longer.

What’s more, we have also found that brown dwarfs accrete material from surrounding disks in the same manner as their stellar cousins. Using high-resolution optical spectrographs on the Keck and Magellan telescopes, our team and others have detected broad, asymmetric emission lines of hydrogen in the spectra of many brown dwarfs at very young ages. These lines come from high-velocity gas plunging in from the disk’s inner edge, channeled in by magnetic felds. Lines of ionized calcium and excited helium, also seen in the spectra, point to high temperatures generated when gas crashes on to the brown dwarf’s surface. Still other spectral lines indicate material fowing out, in jets and winds, probably the result of magnetic felds finging out some of the infall-ing gas—as is seen frequently in young stars. The rate of mass accretion from the disk is ten to a hundred times lower in brown dwarfs than in solar-mass stars, but at least in some cases, the material may continue to trickle in for up to 10 million years—another piece of evidence for long-lived disks around these substellar objects.

Recently astronomers managed to observe the shadow of one brown dwarf’s disk oriented edge-on, confrming that the disk is at least 30 AU in radius, about the size of Neptune’s orbit around the Sun. In another case, an even larger dusty ring is seen to girdle a pair of young very-low-mass objects, most likely brown dwarfs. The ubiquity of the disks, their long lifetimes, and large sizes all argue against pruning during ejection and in favor of the turbulence model.

For a while, the evidence from brown dwarf binaries seemed to point in the opposite direction, however. Early surveys of nearby brown dwarfs found only tight pairs, separated by less than about 15 AU, in agreement with the ejection scenario. More recently, astronomers have identifed a number of very-low-mass wide binaries, some separated by several hundred AU, both in young star-forming regions and in the solar neighborhood. Their existence contradicts ejection through dynamical encounters, which would surely have torn these loosely bound pairs apart.

It’s still possible that some brown dwarfs are ejected from their natal clouds. But the current evidence suggests that the majority are born in situ. That is good news for planet formation: if close encounters are rare in newborn star clusters, stellar disks would not often be disrupted either and would live long enough for planet building.

Bottom Scratching

The famous twin Keck telescopes, sitting pretty atop the summit of Mauna Kea on the Big Island of Hawaii, got a new neighbor in 1999. Subaru, named after the Pleiades star cluster (not the car company), is the prized possession of the National Astronomical Observatory of Japan (NAOJ). Unlike the Keck telescopes, whose 10-meter primary mirrors are each composed of thirty-six separate segments, Subaru uses one enormous 8.3-meter mirror to gather light from the cosmos. Built at a cost of some 370 million dollars (U.S.), Subaru helped Japan leapfrog to the forefront of optical and infrared astronomy.

Subaru’s “dome” stands out against those of the other telescopes on Mauna Kea because of its unusual shape: it’s not a dome at all. Instead, the telescope building is cylindrical—a design based on wind-tunnel tests and computer simulations—to suppress local atmospheric turbulence. Sixteen remote-controlled ventilators around the building provide additional help to improve the air fow above the primary. The unique dome design is just one of many novel features that makes Subaru perhaps the most hi-tech telescope on the ground. Other features include an active support system that maintains the mirror shape to high precision, a tracking mechanism that achieves extreme accuracy by using magnetic drives, many different observational instruments installed at four foci, and a robotic auto-exchanger system for switching instruments and secondary mirrors.

Subaru’s giant primary mirror is barely 20 centimeters (8 inches) thick and has more or less the same proportions as a contact lens. The mirror’s shape is automatically adjusted 100 times a second using 261 small actuators—motors with sensors—that push and pull it from behind to counteract atmospheric turbulence, to minimize the twinkling of the stars. The mirror was polished for three years at the Corning Glass Works in New York State. It was trucked to Pittsburgh, loaded on a barge, and foated down the Ohio River to the Mississippi, then down that river to New Orleans. The mirror crossed the Gulf of Mexico on a larger ship, went through the Panama Canal, and traversed the Pacifc to Hawaii where it was placed on a trailer and hauled up the mountain. It arrived at the observatory in November 1998 after a six-week journey.

From a platform high above, I once watched with more than a little trepidation as a giant robotic arm, moving at a snail’s pace, carried a secondary mirror out from a storeroom on one side of the building to a focus above the primary. Even though the primary mirror was covered at the time, it was hard not to imagine the disastrous result of accidentally dropping something heavy on it.

One of Subaru’s strengths is its large feld of view relative to other telescopes of similar size. A single picture taken with its workhorse optical camera, mounted at the prime focus, covers a patch of the sky roughly the apparent size of the Moon. That might not seem particularly big, but it is eighty times bigger than the area visible to the Advanced Camera for Surveys on the Hubble space telescope. As a result, Subaru is well suited for surveys of the sky in search of faint objects, such as extremely remote galaxies or the least massive brown dwarfs. In collaboration with Motohide Tamura of the NAOJ, my team at the University of Toronto is using Subaru, along with other telescopes, to address a basic question: what is the lowest-mass object that can form the same way as our Sun?

Some astronomers draw the line between brown dwarfs and planets at thirteen times the mass of Jupiter, because objects falling above it can fuse deuterium whereas those below cannot. The distinction makes some physical sense, but uncertainties in the mass estimate for a particular object could place it in either camp, causing confusion. Other researchers favor def-nitions based on formation rather than mass: objects born in disks around stars, either through gravitational fragmentation or through core accretion, are planets while those formed in a star-like scenario, from a contracting cloud, are brown dwarfs. Observations provide some support for this proposal. Planet surveys fnd few companions at the high mass end (say, above 10 Jupiter masses), consistent with the limited supply of material in protoplanetary disks. But how can we be sure about the formation history of a particular object? After all, planets formed in a disk could be banished through close encounters with their siblings and end up as free-foaters. Besides, more-massive stars have more-massive disks, so bigger planets can form in their midst, blurring the divide (see chapter 7).

Perhaps the most extreme example of an object that defes classifcation is COROT-3b, found in a transit search with the French space agency’s COROT satellite in 2008, around a star only slightly larger than the Sun. The companion takes just over four days to circle the star, which makes its orbit similar to those of numerous extrasolar hot Jupiters and unlike those of any known brown dwarfs. But, at a mass of over twenty times that of Jupiter, COROT-3b would be a true behemoth compared with planets and more akin to brown dwarfs from that perspective. Was it born like a stellar companion together with its primary star, or did it grow out of a disk later? Did it form farther out and migrate inward to its current star-hugging orbit? How? We are far from defnitive answers.

In any case, there is no reason to think that the star formation process stops suddenly at the deuterium-burning limit. In the ejection scenario for brown dwarf formation, if an embryo is kicked out suffciently early, it could end up as a planet-mass object. In the turbulence model, chaotic gas motions can trigger the collapse of cores barely a few times more massive than Jupiter. In fact, over the past decade, astronomers have identifed a number of isolated brown dwarfs with estimated masses below the deuterium-burning threshold.

The frst crop came from the Canary Islands group in 2000, in a paper published in Science with Maria Rosa Zapatero Osorio as the lead author. The team had carried out a deep imaging survey of a portion of the sigma Orionis star cluster, with the tender age of a few million years and located in relative proximity about 1,200 light-years from Earth. The researchers identifed eighteen objects so faint and so red, compared with previously known stars and brown dwarfs in the same region, that they would have estimated masses in the range of fve to ffteen times that of Jupiter. Spectra of three of them confrmed low atmospheric temperatures.

The announcement caused a commotion in the media as well as in the scientifc community. “Unidentifed Floating Objects: Not Quite Stars or Planets,” read the New York Times headline. The Washington Post declared, “Discovery of Objects Stirs ‘Planet’ Debate: Team’s Labeling Is Immediately Disputed.” Some astronomers were skeptical that the objects, especially those for which spectra were not yet in hand, are in fact cluster members rather than background stars. Others pointed out that the theoretical models used to infer masses are uncertain, especially at the youngest ages and the lowest masses. Still others were irked by the team’s use of phrases like “isolated giant planets” and “rogue planets” to describe the objects; these choices were PR-driven, they impugned. I remember passionate arguments over this at a workshop on the origins of stars and planets held in April 2001 in Garching, Germany.

Within a year, the Canary Islands team obtained spectra of their other candidates and confrmed that all but one likely belong to the cluster and thus have masses in or close to the planetary regime. Meanwhile, another research group led by Phil Lucas at the University of Hertfordshire in England reported about a dozen candidates below the deuterium-burning threshold among the newborn stars in the Trapezium cluster in the Orion Nebula. Gradually, astronomers have come to accept the existence of ultra-low-mass brown dwarfs, but to this day there is no consensus about what to call them. Various researchers have coined terms such as sub–brown dwarf, isolated planetary mass object, and free-foating planet. Some of my colleagues and I have used the word “planemo,” short for planetary mass object, introduced by Gibor Basri. I like the moniker because it refers to the similarity in mass to giant planets, while also making a distinction. Plus, there is precedent in astronomy for invented words like it—pulsar for pulsating radio star, quasar for quasi-stellar object, for instance.

Planets orbiting “Planets”?

Over the years, astronomers have found planetary-mass objects, or “planemos,” in other star-forming regions. But we still do not know how common they are and how low in mass they come. Some researchers wonder whether the least massive objects could have been ejected after forming in disks around protostars. It is to address these questions that we have undertaken surveys of several nearby young regions with Subaru and other telescopes. Our project, called SONYC for Substellar Objects in Nearby Young Clusters, aims to fnd them in a systematic way. We target star-forming regions because newborn brown dwarfs are much brighter than older ones, as they glow from converting gravitational energy into heat during contraction. That makes it possible for us to identify objects down to a few Jupiter masses in our deep optical and infrared images. Besides, the observed characteristics of newborns—whether they harbor disks and come in pairs, for example—might give us clues to their origin.

Once we identify candidate brown dwarfs and planetary mass objects, from how bright they appear in different flters in comparison with stars, we try to take spectra to confrm their nature. Therein lies the challenge. These objects are so faint that even with the world’s largest telescopes, we can barely make out the absorption features due to water vapor, for instance, once their light is spread into the constituent wavelengths. These telltale signatures and the overall “shape” of the spectrum permit us to be sure we are looking at a true lightweight in the cluster rather than an old red dwarf star in front of it.

Not surprisingly, ours is not the only team of astronomers after the same quarry. For example, a network of European astronomers has undertaken a large, multi-faceted program to investigate the origin of stellar and substellar masses. Their overall goal is to measure the so-called initial mass function—the number of objects born in different mass bins, from tens of solar masses to well below 1 percent of the Sun’s mass, which corresponds to several Jupiter masses—and to determine whether it varies from region to region. We, and other groups, focus on the bottom end of this mass distribution. By surveying a small portion of the sigma Orionis region to extreme depth and extrapolating to the cluster as a whole, the Canary Islands researchers have reported that there might be as many planemos as higher-mass brown dwarfs there. That does not appear to be the case in the Trapezium, where less than 10 percent of the cluster members have planetary masses, according to Phil Lucas and his collaborators. Our fndings so far suggest that planemos are even less common in a third region, the NGC 1333 cluster in Perseus. As we survey other regions, we hope to determine whether the number of planemos relative to more-massive brown dwarfs and stars depends on the birth environment—whether denser clusters harbor fewer of them, for instance.

Once we fnd and confrm ultra-low-mass objects, we try to determine their properties as best as we can. For that purpose, deep images of the same clusters taken with the Spitzer space telescope come in handy. If the objects we identify are also seen with Spitzer, we can tell whether they exhibit excess emission at mid-infrared wavelengths from dusty disks. So far, the evidence points to planemos harboring disks as often as brown dwarfs and Sun-like stars. Of course, the disks become progressively less massive around lower-mass objects.

Since planetary systems form out of disks, this fnd-ing raises an intriguing question: could there be planets around objects that are themselves not much heftier than Jupiter? Theorists argue that giant planets should be rare around low-mass stars, let alone brown dwarfs, and current observations seem to agree (see chapters 2 and 4). Substellar disks do not contain enough material to make gas giants, though such companions might form by other means (see chapter 7). In any case, there is no reason to think that asteroids and comets, or even Earth-size planets, could not form in disks around brown dwarfs and planemos. In fact, Spitzer observations reveal signs of growth and chemical processing of dust grains in some brown dwarf disks, perhaps marking the frst tentative steps toward planet building. The Atacama Large Millimeter Array, about to start operations on a high (and dry) mountain plateau in the north of Chile, might reveal gaps and holes in these disks carved out by planetary bodies—signatures often seen in their stellar counterparts. If they exist, Earths around brown dwarfs and planemos would circle suns that aren’t suns at all, committing them to an eternal freeze and making them less than hospitable to life as we know it.