Lonely Planets

by David Grinspoon

  Queloz were confident that they had something, but they were reluc-

  tant to announce their discovery. They wanted to be careful because for

  many decades astronomers have announced discoveries of extrasolar

  planets (those orbiting other stars), only to be proven wrong. Such false

  alarms have harmed many careers. The apparent strangeness of their

  new planet only increased their caution. In March they had to suspend

  operations as 51 Peg disappeared behind the Sun. They planned a set of

  observations for early July, when the star would reemerge. They calcu-

  lated exactly the velocity 51 Peg should exhibit when it reappeared.

  This would be the final test. If it had the right velocity in July, then

  there was no other explanation and they would announce to the world

  that they had found an extrasolar planet.

  In the first week of July, Mayor and Queloz returned to the observa-

  tory. This time they brought their families, and some champagne and

  cake, along to the telescope dome. Late on the first evening Pegasus

  rose over the horizon and they pointed the telescope at 51 Peg. A few

  minutes later they had their answer: the velocity matched their predic-

  tion perfectly. They uncorked the bottles: 51 Peg is being swung to and

  fro by a planet-size orbiting companion.*

  Mayor and Queloz did not go public with this finding until they found

  out that the report they had submitted to Nature was accepted for pub-

  lication. In October, Mayor made the stunning announcement at an

  astronomy conference in Florence. After that, word traveled quickly. It

  soon reached the Swiss team’s greatest competitors, Geoff Marcy and

  Paul Butler at San Francisco State University. Marcy’s reaction, upon

  hearing of the discovery, was mixed. He had been searching for extraso-

  *Mayor has proposed the name Epicurus for this new planet. As of this writing its official name is still “the planet orbiting 51 Pegasi.”

  210

  L o n e l y P l a n e t s

  Image unavailable for

  electronic edition

  lar planets, using a spectroscopic technique similar to that of the Swiss

  team, for eight years. He had been scooped in a major way. He could

  have made this discovery had he looked at the right star and analyzed

  the data.* On the other hand, he was thrilled at the idea of a solid planet

  detection. He was also skeptical, given the history of false alarms about

  extrasolar planets. Just five days after Mayor’s announcement, Marcy

  and Butler went to the Lick Observatory, in the mountains east of San

  Jose, to observe 51 Peg. Within a few nights they had gathered enough

  data to confirm the Swiss discovery. The planet was real.

  What happened next was strange. Marcy and Butler issued a press

  release that began: “Astronomers at the University of California at

  Berkeley and San Francisco State University have confirmed a recent

  *Marcy and Butler had actually excluded 51 Peg from their search because of an error in a star catalog, which misclassified this star as an unstable red giant, rather than the stable Sun-like star that it is.

  Enter the Exoplanets

  211

  report of a planet orbiting a nearby star—the first time a reported dis-

  covery of a planet around a normal star has withstood scrutiny.”

  The next morning the astronomers began receiving a barrage of

  phone calls from every major media outlet in the world. The day after

  that, their discovery was front-page news on papers all around the

  globe. Later that week ABC canceled their planned Nightline and

  Marcy appeared as a guest, talking about the new planet.

  Back in Switzerland, Mayor and Queloz were fuming. In the press

  reports announcing that American astronomers had discovered a planet,

  some mentioned that this object had also been observed by a Swiss team,

  and some left out this detail. This was not the fault of Marcy and Butler,

  who, whenever they had the opportunity, were careful to mention that

  the discovery had actually been made by the Swiss. But Mayor and

  Queloz were forbidden from talking to the press by the editors of

  Nature. Those are the rules: if you want to be published in Science or Nature, you have to refrain from publishing your results elsewhere so

  that the journal can have the scoop. This created an absurd situation

  where the actual discoverers suffered under a gag rule while those who

  had confirmed their detection got the glory and became media stars.

  After the first discovery, several new extrasolar planets (or exoplan-

  ets) were soon found by using the same spectroscopic technique. Marcy

  and Butler made many of these detections, and despite the media confu-

  sion surrounding the first discovery, their reputation as leading planet

  discoverers is completely deserved. In fact, they did not even have to go

  back to a telescope to discover more planets. They found them by

  “data mining,” simply by analyzing data that was sitting in their hard

  disks from the past several years of observations. They already had evi-

  dence of more planets but had not found them because they had not

  bothered to look for the signature of planets with very short orbital

  periods, since everybody knows that giant planets cannot exist that

  close to their star. Often in science the really exciting discoveries are

  made when what everybody knows turns out to be wrong. In this case,

  once Marcy and Butler realized that giant planets can be on tiny orbits,

  they looked through old data and found the signature of several giants

  with orbital periods of just a few days. By 1996, they had found four

  more extrasolar planets.

  Since then, in eight years we’ve found more than one hundred planets

  orbiting other stars. The number of known extrasolar planets now

  212

  L o n e l y P l a n e t s

  dwarfs the nine planets in our own system. Reports of new exoplanets

  have been pouring in so rapidly that individual detections have ceased

  to be big news, even as the expanding collection of new worlds contin-

  ues to delight astronomers. For the first half dozen or so planet detec-

  tions, each was given big headlines. After that, as planet detections

  have increased, newspaper headlines about extrasolar planets have

  decreased, but persisted.

  The 103rd new planet is just not as exciting as the third, and not only

  the press has become jaded. Scientists, too, have gotten used to hearing

  about new planets. By 2002, it had reached the point where, at least on

  a busy day, when you see the headline “15 New Planets Announced at

  Conference,” you might not even click on it to read more. Only head-

  lines proclaiming new kinds of planets are guaranteed a click.

  What the scientists know is that nondetections are important, too.

  Every time we observe a star and find no telltale wobble, it puts new

  limits on the number and type of planets that may be out there. The

  nondetections do not make headlines (“Scientists Discover Nothing”),

  but cumulatively they are as interesting as the detections. When we’re

  wondering how unusual our own system is, and how this changes our

  estimates of the prevalence of life in the universe, it is t
he pattern that’s important—the overall distribution of planets of different types orbiting stars of different types. This pattern will gradually be revealed over

  many years, and not in one headline-making discovery.

  Planet finders are still holding press conferences, but only when they

  find new planets or planetary systems that stand out from the growing

  herd. In most cases, this involves systems that are, in some way, more

  like our own solar system than any other previously found.

  A glance through the headlines of articles about extrasolar planets

  written by New York Times science writer John Noble Wilford reveals

  this progression. By 1998, individual planet detections no longer mer-

  ited their own articles until June 26, when “New Planet Detected

  around a Star 15 Light-Years Away” announced one closer to the Sun

  than any others. On April 16, 1999, “At Long Last, Another Sun with

  a Family of Planets” described the first discovery of a system where

  three planets orbited the same star. On March 30, 2000, “Two

  Relatively Small Planets Are Found” heralded the discovery of two

  planets with masses slightly smaller than Saturn’s. Most previous dis-

  coveries had revealed planets much larger than Jupiter and Saturn. By

  June 14, 2002, Wilford was proclaiming, “Astronomers Detect Signs

  Enter the Exoplanets

  213

  That Jupiter Has Distant Cousin.” This article described the first detec-

  tion of a planet whose mass and orbit are similar to that of Jupiter. We

  can expect that this parade of “firsts” will continue indefinitely, as

  improving observing technology, and more time to look, turns up sys-

  tems that are closer and closer to our own in various ways.

  T H E F I R S T H U N D R E D

  In part, this great emphasis on seeking systems with characteristics

  closely resembling ours arises because the first one hundred planets dis-

  covered are almost all completely unlike anything we had expected.

  This collection certainly does not resemble the planets of our solar sys-

  tem. Like the planet that Mayor and Queloz found orbiting 51 Peg,

  these are mostly “hot super-Jupiters”: giant planets, mostly more mas-

  sive than Jupiter, locked in tight little orbits, many of them closer to

  their home stars than Mercury is to our Sun. Not only don’t these plan-

  ets resemble the ones around here, but the architecture of all of these

  newly found planetary systems seems to be fundamentally different

  from that of our own system. They are built on a different plan.

  Maybe we shouldn’t be surprised. There are obvious “observational

  selections” at work here—aspects of the search techniques that bias our

  findings. The first ones you discover are the easiest ones to see, and

  they are not going to be representative of all that is out there. When

  you look at a distant redwood forest, you first see sequoias and miss

  the undergrowth, but that doesn’t mean that the only inhabitants of the

  forest are the giant trees that dominate your initial view.*

  Keep in mind that we don’t “see” these worlds at all. Instead, we infer

  their presence by the gravitational sway they induce in their central star.

  It’s like watching a hula-hoop contest from across a wide valley. We’re

  too far away to see the thin little hoops, but why else would those big

  hips be shaking like that? What we actually notice is a big boogying star,

  doin’ da gravitational bump with a slight, unseen companion.

  We don’t know much about these new planets, but we can tell a few

  things. We can guess their weight by the size of the stellar wobbles they

  cause. Giant planets are easiest to detect because, with their massive

  *It’s like pot smokers—the ones who call a lot of attention to themselves are a minority who tend to fit a certain negative stereotype. This creates a skewed demographic sampling for those who don’t look closely.

  214

  L o n e l y P l a n e t s

  gravitational pull, they really get their stars moving. So, of course the

  first ones found are preferentially “super-Jupiters.” Most of them are

  more massive than any planets in our own system.

  It’s also easier to find planets on orbits that hug their stars closely.

  The undulations they induce are both larger and faster. Kepler’s third

  law ensures that close-in planets will orbit faster than their more dis-

  tant siblings anyplace where gravity follows the laws of gravity.* Stars

  with close-in planets wobble faster, so you don’t have to observe for

  nearly as long to bag the planet. Who’s got the time to wait around to

  observe the slower and smaller effects of a more distant Jupiter that

  takes twelve years to orbit?

  Once we do find a planet, this same law allows us to deduce, from its

  orbital period, its distance from its star, which allows us to estimate its

  climate. The planet orbiting just 5 million miles from 51 Peg must have

  a surface temperature of nearly two thousand degrees Fahrenheit,

  which makes even Venus and Mercury seem temperate by comparison.

  Given these artificial selection factors, the first hundred extrasolar

  planets we’ve found are naturally dominated by hot super-Jupiters, that

  is, by giant planets orbiting close to their star. We’re beginning to get

  used to this population of oddball planets, but initially they came as a

  complete surprise because there weren’t supposed to be worlds like that.

  Six years before the first extrasolar planet was discovered, I finished

  grad school, where I learned the rules for building planets and solar sys-

  tems. Back then, the expectation was that planetary systems elsewhere

  would, in their overall basic architecture, resemble ours. They would have

  a few gas giants out in the cold outer regions and, in the inner hot zone, a

  few rock balls, possibly one or two in the zone with watery swimming

  holes and maybe even something in there doing the backstroke.

  Given the characteristics of the first hundred new planets, these

  expectations were clearly too narrow. Most of the giant planets discov-

  ered are orbiting close in, where theory told us it would be too hot for

  gas giants to form. More recently, several planets have been discovered

  with orbital distances and masses more closely resembling those of

  Jupiter and Saturn. However, unlike Jupiter and Saturn, many of these

  are not on well-behaved, near-circular orbits. Many follow wild ellipti-

  cal orbits that would probably not allow for Earth-like planets to share

  their planetary system.

  *And, dude, that’s, like, everywhere.

  Enter the Exoplanets

  215

  Okay, so the first hundred new planets do not conform to our old

  expectations. What can we learn from this, other than (surprise, sur-

  prise) that we were naive? How can this ensemble of worlds help our

  effort to contextualize Earth and Earth life?

  At this point, “Rare Earth” thoughts may start popping into your

  head: What if we live in a completely deviant solar system, and our

  presence here indicates that such an unusual location is required for

  something like us to come along?

  The extent to which the earlies
t exoplanets found defied our expecta-

  tions and differed from our own system has been seized upon by some

  to draw conclusions about our planet’s uniqueness. In describing the

  extrasolar planets discovered up until 2002, the authors of Rare Earth,

  Peter Ward and Don Brownlee, wrote in their next book, The Life and

  Death of Planet Earth, that “astronomers were surprised to discover

  that most solar systems do not have the larger planets far from the Sun

  in well-behaved, nearly circular orbits as ours does.” They used this to

  support their contention that systems, and planets, like our own are

  rare and unusual. Yet this statement is flawed. We simply don’t have

  the data to draw such a conclusion. We cannot yet say anything about

  what “most solar systems” are like.

  We are seeing an astounding variety in exoplanets—and much more

  variation in the basic architecture of planetary systems than we ever

  thought likely. From this, we are tempted to conclude that ours is not a

  garden-variety solar system, but we don’t know this. What we can say is

  that the galactic garden has gone to seed and is overgrown with a sur-

  prisingly diverse bouquet of wild suns and their unruly planets. This cer-

  tainly doesn’t tell us that there aren’t many other systems out there that

  closely resemble ours. Far from it. In fact, more recently we’ve found a

  few that do seem to resemble our own. It appears that they are not the

  majority, but even this appearance may turn out to be due to the obser-

  vational biases that make certain kinds of planets much easier to find.

  We have now looked around at a few hundred stars and found that

  around 5 percent have planets that we can detect with our current tech-

  nology. At this point, systems like our own are still difficult to detect,

  and 90 percent of solar systems could look exactly like ours. We won’t

  know definitively how typical our own planetary system is until we

  take a more thorough census of the planets in our stellar neighborhood.

  All we now know for sure is that nature is, as always, more imagina-

  tive than we are, and planetary systems are much less uniform in con-

  216

  L o n e l y P l a n e t s

  struction than we expected. We’re learning that the galaxy is not con-

  structed like a Soviet-era Russian (or modern-American) housing devel-