Faint Echoes, Distant Stars

by Ben Bova

  An independent study of Barnard’s Star was undertaken by George Gatewood of the University of Pittsburgh’s Allegheny Observatory and Heinrich Eichhorn of the University of South Florida using an automated measuring system developed at the U.S. Naval Observatory in Washington, D.C. In 1973, they reported that they could find no detectable perturbations in the proper motion of Barnard’s Star.

  The “planet” was a mistake, although van de Kamp insisted on its reality to his death in 1995.

  PULSAR PLANETS?

  Pulsars are collapsed stars that emit radio pulses that were once thought possibly to be signals from intelligent aliens (see Chapter 18).

  In 1991, Andrew Lyne and his colleagues at Britain’s Jodrell Bank radio telescope reported that they had detected a planet orbiting the pulsar PSR1829-10.25

  This was a shock. Pulsars are the collapsed, incredibly condensed cadavers of massive stars that have blown up in tremendous supernova explosions. Any planets orbiting such a star should have been blasted into vapor by the billion-degree-hot plasma of the supernova. Yet here was a planet circling the burned-out hulk of an old supernova.

  Could planets survive a supernova blast? Or could new planets form out of the tangled, swirling cloud of ionized gas that the supernova hurls into space?

  Or was PSR1829-10’s planet another mistake?

  The planet was detected through a slight shift in the frequency of the radio pulses coming from the pulsar, a shift that happened regularly every six months. The Jodrell Bank group deduced that the shift in radio frequency was caused by a solid body—a planet—cutting through the radio beam.

  But it wasn’t. The astronomers had overlooked the fact that the Earth revolves around the Sun. The six-month shift was caused by the Earth’s motion, not a pulsar planet. With enormous embarrassment but manly courage, the Jodrell Bank team withdrew their claim early in 1992.

  In the meantime, though, Alexander Wolsczan of the Arecibo Observatory and Dale Frail of the National Radio Astronomy Observatory in New Mexico announced that they had found evidence for two planets orbiting the pulsar PSR1257+12. Their masses were a few times larger than Earth’s, and their orbital periods were sixty-seven and ninety-eight days, which put them at a distance from the pulsar of slightly less than 1 AU.

  Their discovery was confirmed. The first extrasolar planets had at last been discovered. But how could they possibly be considered as potential abodes for any kind of life? If they had existed before the supernova explosion that formed the pulsar, their surfaces must have been blown into vapor right down to the metallic cores of the planets. Since the supernova blast they had been exposed to sterilizing levels of X-ray and gamma radiation from the pulsar.

  To astrobiologists, the pulsar planets seemed to be a cruel “good news-bad news” joke: Yes, we have found planets, but they can’t be anything more than burned-out cinders.

  Yet there was an encouraging fact underlying the discovery: A star beyond the Sun harbored planets. No matter that the star had exploded and collapsed down to a pulsar. Once it was a true star and planets circled around it. To astronomers, that meant that other stars should harbor planets, too.

  NEW TECHNIQUE, NEW DISCOVERIES

  So the quest continued. By the 1990s, astronomers had developed a new technique for finding extrasolar planets. As we have seen, the astrometric technique depends on measuring slight perturbations in a star’s proper motion, that is, its motion across our field of view.

  Using spectrographic analysis, though, it is possible to make much more sensitive measurements of a star’s radial velocity: its motion toward or away from the observer.

  The spectrograph breaks up the light from a star into a spectrum of its component colors, just as a glass prism breaks up sunlight into a rainbow of colors. By studying stars’ spectra, astronomers have been able to determine their temperatures, their chemical compositions, and—thanks to the Doppler effect—their motions through space along our line of sight.

  By taking spectra of a star and looking for the Doppler shifts in the starlight caused by the star’s radial velocity, astronomers can measure extremely small motions of approach and recession.

  The Doppler-Fizeau Effect

  PLANETS OF SUN-LIKE STARS

  In October 1995, Michael Mayor and Didier Queloz of the Geneva Observatory in Switzerland announced the discovery of a planet orbiting the Sun-like star 51 Pegasi, which lies forty-two light-years from the Sun. Their discovery was quickly confirmed by Geoffrey W. Marcy and R. Paul Butler of San Francisco State University and the University of California at Berkeley.

  A planet orbiting a normal star very much like our Sun had at last been found. That was the good news.

  You can see the star 51 Pegasi on a clear, dark night. It is just visible to the naked eye, slightly in front of the leading (westward) edge of the Great Square of Pegasus, a favorite sight for stargazers. A yellowish star, 51 Peg is a little cooler than the Sun and slightly dimmer. It is definitely a stable, sun-like star.

  Now the bad news. The planet orbiting 51 Peg is at least half the mass of Jupiter, yet its orbit is only 7 million kilometers from its parent star, less than one-eighth the distance of Mercury from our Sun. It is practically brushing the star’s fiery surface. It orbits around 51 Peg in just 4.2 days. Its surface temperature has been calculated to be 1,300°C.

  How can such a massive planet—bigger than Saturn—exist so close to its star? At first glance, it would seem likely that if you towed Jupiter or Saturn that close to the Sun most of the planet’s deep atmosphere, predominantly composed of hydrogen and helium, would be quickly boiled away into space. After all, the temperature at Jupiter’s “surface”—the tops of its clouds—is on the order of -150°C. Saturn is even colder. Heat it to 1,300°C, and the gas giant would shrivel down to its rocky core.

  Or would it? Calculations have suggested that, once formed, such a massive planet would be able to hold on to most of its material even at that high a temperature. But it could never have formed that close to its star; at its present distance from 51 Peg, a planet could no more have accreted enough material to become a gas giant than dense little Mercury did. Perhaps 51 Peg B, as the planet is dubbed in the official jargon of the astronomers, was formed at a Jupiter-like distance from its star and then somehow (!) spiraled in to its present hot, snugly close orbit.

  So, either 51 Peg B was born at least five times farther from its star than it is now, or it is composed of materials that are very different from the gas giants of our solar system. Either way, the first extrasolar planet to be discovered around a Sun-like star is nothing like the planets of our own solar system.

  There was more good news/bad news to come.

  A BROWN DWARF

  Also in October 1995, Shrinivas Kulkarni and colleagues at the California Institute of Technology in Pasadena announced their discovery of a body some twenty to sixty times more massive than Jupiter orbiting the red dwarf star Gliese 229, which is thirty light-years from us. GL229’s companion orbits the star at a distance of about 44 AUs, slightly farther from GL229 than Pluto is from our Sun.

  GL229 B, as it is officially dubbed, is most likely a brown dwarf, a “failed star” that does not quite have enough mass to initiate the fusion reactions in its core that would have made it a true star. Methane and water vapor have been detected in its spectrum, which implies a surface temperature below 1,315°C. If it were a true star, its surface would be too hot for either methane or water vapor to exist.

  Calculations have shown that a body must be seventy-five to eighty times more massive than Jupiter in order to sustain nuclear fusion reactions in its core and become a true star. GL229 B is too small to be a star but very likely is a brown dwarf. Brown dwarf stars should radiate in the infrared, giving off heat from their slow contraction. GL229 B was detected by infrared sensors and photographed both in the infrared and in visible light by the Hubble Space Telescope.

  Of all the extrasolar planets discovered, GL229 B was the first to be ac
tually seen, thanks mainly to its distance from its parent star and its own smoldering infrared emission. In 2002, Michael Liu of the University of Hawaii and his colleagues used the 8.1-meter Gemini North telescope in Hawaii to photograph a brown dwarf orbiting the star 15 Sagittae.

  GOLDILOCKS . . .

  Marcy and Butler, meanwhile, had been patiently monitoring the motions of 120 stars for more than seven years with a spectrograph mounted on a 3.04-meter telescope at Lick Observatory in California. When Mayor and Queloz made their 51 Peg announcement, Marcy and Butler checked their data and confirmed the discovery of the planet.

  Then they searched frantically through their own records of the first 60 of the 120 stars they had been monitoring, hoping to find evidence for another planet.

  They found several.

  Most exciting of all was the discovery of a planet that has been called “Goldilocks” because its orbit is not too close and not too far from its star for it to be at just the right temperature to bear liquid water on its surface. In other words, its orbit is totally within its star’s thermally habitable zone.

  There is only one other world in the universe that we know bears liquid water on its surface, and that planet—Earth—also bears life.

  Does Goldilocks?

  Goldilocks orbits the star 70 Virginis, a star that is just a tad hotter than the Sun but slightly dimmer. Although Goldilocks (or 70 Vir B) is at least 6.5 times more massive than Jupiter, it orbits closer to its star than Venus does to the Sun. Its “year” takes only 117 days. While its highly elliptical orbit is entirely within the habitable zone where liquid water could exist on its surface, its high mass led some astronomers to theorize that Goldilocks is not truly a planet but another brown dwarf. And the fact that liquid water could exist on Goldilocks’ surface does not necessarily mean that it does.

  . . . AND THE BIG BEAR

  The second planet Marcy and Butler found orbits 47 Ursae Majoris (the Big Bear) and is most like the kind of planet to be found in our own solar system. The star is visible to the naked eye below the “bowl” of the Big Dipper (which is a part of the Big Bear constellation).

  The mass of the planet orbiting 47 Ursae Majoris is about 2.3 times that of Jupiter, and its distance from its star is some 300 million kilometers, corresponding to an orbit between Mars and Jupiter. Its orbital period (year) is about three Earth years.

  The Big Bear planet would not look terribly out of place in our solar system; its orbit is in the area where our Asteroid Belt exists, and, unlike most of the other extrasolar planets detected so far, its orbit is almost exactly circular, like the planets of our own solar system.

  Moreover, in 2001, the Marcy-Butler team announced that they had found a second planet orbiting 47 Ursae Majoris with a mass and orbit similar to Jupiter’s. This second planet also has a nearly circular orbit, which makes 47 UM the first star discovered to harbor a planetary system that resembles our own. Perhaps there are smaller, Earth-like planets orbiting 47 UM closer than the two Jupiter-sized worlds.

  More recently, in 2003, an Anglo-Australian team announced the discovery of a Jupiter-sized planet in a nearly circular Jupiter-sized orbit around the star HD70642, which lies ninety light-years from Earth. The significance of this discovery is that there are no “hot Jupiters” close to the star, which means that smaller, Earth-sized planets could be orbiting there. HD70642’s system appears to be the most like our own solar system, to date.

  Later that same year, Steinn Sigurdsson of Pennsylvania State University and his colleagues reported detecting a planet some 2.5 times more massive than Jupiter orbiting a white dwarf star and its pulsar companion some 7,200 light-years from Earth. Not only is this planet the farthest yet discovered, but its existence at such aged stars shows that planet formation must have been going on many billions of years ago.

  VAN DE KAMP VINDICATED?

  It gets even more interesting.

  In June 1996, Gatewood, at the Allegheny Observatory, tentatively announced the discovery of two planets orbiting Lalande 21185 at 8.1 light-years’ distance, the fourth closest star to the Sun. A dim red dwarf star that is racing across our field of view, Lalande 21185 was one of those stars that van de Kamp claimed as possessing “unseen companions” some fifty years earlier.

  Like van de Kamp, Gatewood used the astrometric technique for measuring the star’s proper motion, although he had the benefit of modern electronics as he tracked the star’s motion from 1988 with a multichannel photoelectric detector.

  Gatewood calls his findings tentative because the more massive of the two planets needs thirty years to complete its orbit around the red dwarf star, and he has less than ten years of data on hand. But he believes that Lalande 21185 has a planet of almost Jupiter’s mass orbiting at 2.5 AUs, which would be between the orbits of Mars and Jupiter in our system. Its “year” is about six Earth years.

  At 10 AUs’ distance is a second planet of about 1.5 times Jupiter’s mass; this is the one that orbits Lalande 21185 in thirty years. In our solar system, Saturn’s orbit is 9.54 AUs from the Sun; its “year” is 29.46 Earth years.

  When informed of the Lalande 21185 discovery, tentative though it may be, theorist Alan P. Boss of the Carnegie Institution of Washington, D.C., said, “This is the place to look for an Earth-like planet.” Butler and Marcy have detected no evidence of a wobble in Lalande 21185’s motion, but they say this is inconclusive, since positive confirmation—or refutation—of Gatewood’s discovery will take thirty years of patient observation.

  There is a double irony in Gatewood’s discovery. Not only has he revisited one of van de Kamp’s claims, but it was Gatewood who helped demolish van de Kamp’s work by identifying the “instrumental effects” that threw such fatal doubt on van de Kamp’s studies of Barnard’s Star.

  UPSILON ANDROMEDAE’S THREE PLANETS

  In 1999, Butler and colleagues at the Anglo-Australian Observatory announced the discovery of three Jupiter-sized planets orbiting Upsilon Andromedae, a star that is about three times brighter than the Sun and some forty light-years distant. The innermost planet is about 70 percent of Jupiter’s mass and orbits a mere 0.06 of an Astronomical Unit from its star, practically brushing its surface at a distance of only 9 million kilometers. The next closest is in a slightly eccentric (noncircular) orbit at about 0.8 AU; it is at least twice as massive as Jupiter. The outermost planet is at least four times Jupiter’s mass and in a very eccentric orbit that averages 2.5 AUs from the star—about half the size of Jupiter’s orbit.

  HOT JUPITERS

  By the close of the twentieth century, astronomers had discovered dozens of extrasolar planets, far more than the nine planets of our own solar system. And more discoveries were coming in every week, it seemed. As of early 2003, more than 100 extrasolar planets have been found.

  Moreover, dozens of other stars were found to be surrounded by dust clouds: accretion disks in which planetary formation is undoubtedly taking place. The first such disk was found at the star Beta Pictoris in 1983. In late 2001, Aki Roberge of Johns Hopkins University reported that the star 51 Ophiuchi is surrounded by an accretion disk of planetesimals and dust; 51 Ophiuchi is a very young star, perhaps no more than 300,000 years have elapsed since it began to shine. This discovery shows that stars form accretion disks very early in their history and start the process that leads to the formation of planetary systems.

  It became abundantly clear that our solar system is not unique: Stars form planetary systems quite naturally. That is the good news.

  However, there is also bad news. Almost all the planets discovered to date are very massive and orbit their parent stars in elongated, highly elliptical trajectories. This has led some astronomers to doubt that they are truly planets; instead, they might be brown dwarf “failed stars.”

  Few of the extrasolar planetary systems resemble our own solar system very closely. For example, the star 55 Cancri possesses at least two planets and very possibly more. One of them is a gas giant orbiting a mere 0.
1 AU from the star; it whips around 55 Cancri in just 14.6 days. The other, though, is roughly the same distance from its star as Jupiter is from the Sun, in a slightly elliptical orbit 5.5 AUs from the star. It is more than four times more massive than Jupiter. The data indicate that another Saturn-sized planet may be orbiting farther out from 55 Cancri. Despite the gas giant snuggled closer to its star than Mercury is to the Sun, the 55 Cancri system is more like our own solar system than any other yet discovered.

  The extrasolar planets discovered so far are all of Jovian size or bigger and most of them orbit very close to their stars. They were soon dubbed hot Jupiters: Jupiter-sized planets orbiting a hair’s breadth away from the seething surfaces of their parent stars.26

  The only way to explain the hot Jupiters was to assume that they had formed out of the star’s accretion disk at distances similar to the gas giants of our own solar system, then had been forced into their present tight, blistering-hot orbits. Computer simulations showed that planetary systems could behave like a giant arcade video game early in their histories, with planets ping-ponging back and forth, some of them bouncing into orbits that skim incredibly close to their star, others slingshotting out of the system altogether to wander off into interstellar space.

  This is a far different situation from our own placid, stable solar system. Or is it? Our solar system was certainly more violent and unstable in its youth. We have seen that Earth’s Moon is probably the result of a pulverizing collision between the Earth and a Mars-sized planetesimal. Did Jupiter-sized worlds slingshot through the solar system eons ago? There is some evidence that the orbits of the gas giant planets were influenced early on by close encounters with planetesimals and then migrated to their present stable positions.