Origins: Fourteen Billion Years of Cosmic Evolution (18 page)

They could deduce still more about the planet. Moving at a particular distance from its star, a planet’s gravity will pull on the star with a force that depends on the planet’s mass. More massive planets exert greater force, and these forces make the star dance more rapidly. Once they knew the planet-star distances, the team could then include the
masses
of the planets in the list of planetary characteristics that they had determined through careful observation and deduction.

This deduction of a planet’s mass by observing the star’s dance comes with a disclaimer. Astronomers have no way to tell whether they are studying a dancing star from a direction that happens to coincide exactly with the plane of the planet’s orbit, or from a direction directly above the plane of the orbit (in which case they will measure a zero velocity for the star), or (in almost all cases) from a direction neither exactly along the plane nor directly perpendicular to it. The plane of the planet’s orbit around the star coincides with the plane of the star’s motion in response to the planet’s gravity. We therefore observe the full orbital speeds only if our line of sight to the star happens to be the same as the plane of the planet’s orbit around the star. To imagine a loosely analogous situation, put yourself at a baseball game, able to measure the speed of the pitched ball as it comes toward you or moves away, but not the speed with which the ball crosses your field of vision. If you are a talent scout, the best place for you to sit is behind home plate, in direct line with the baseball’s motion. But if you observe the game from the first or third baselines, the ball thrown by the pitcher will, for the most part, neither approach you nor recede from you, so your measurement of the ball’s speed along your line of sight will be nearly zero.

Because the Doppler effect reveals only the speed with which a star moves toward us or away from us, but not how rapidly the star crosses our line of sight, we usually cannot tell how nearly our line of sight to the star lies in the plane of the star’s orbit. This fact implies that the masses that we deduce for exosolar planets are all
minimum
masses; they will prove to be the planets’ actual masses only in those cases when we do observe the star along its orbital plane. On the average, the actual mass of an exosolar planet equals twice the minimum mass deduced from observing the star’s motions, but we have no way to know which exosolar planet masses lie above this average ratio, and which below.

In addition to deducing the planet’s orbital period and orbital size, as well as the planet’s minimum mass, astrophysicists who study star dances by the Doppler effect have one more success: they can determine the shape of the planet’s orbit. Some of these orbits, like those of Venus and Neptune around the Sun, have an almost perfect circularity; but others, like the orbits of Mercury, Mars, and Pluto, have significant elongation, with the planet traveling much closer to the Sun at some points along its orbit than at others. Because a planet moves more rapidly when it is closer to its star, the star changes its velocity more rapidly at those times. If astronomers observe a star that changes its velocity at a constant rate throughout its cyclical period, they conclude that these changes arise from a planet moving in a circular orbit. If, on the other hand, they find that the changes sometimes occur more rapidly and sometimes more slowly, they deduce that the planet has a noncircular orbit, and can find the amount of the orbital elongation—the amount by which the orbit deviates from circularity—by measuring the different rates at which the star changes its velocity throughout the orbital cycle.

Thus, in a triumph of accurate observations coupled with their powers of deduction, astrophysicists who study exosolar planets can provide four key properties of any planet that they find: the planet’s orbital period; its average distance from its star; its minimum mass; and its orbital elongation. Astrophysicists achieve all this by capturing the colors of light from stars that lie hundreds of trillions of miles from the solar system, and by measuring those changes with a precision better than one part in a million—a high point in our attempts to probe the heavens in a search for Earth’s cousins.

Only one problem remains. Many of the exosolar planets discovered during the past decade orbit their stars at distances much smaller than any of the distances between the Sun to its planets. This issue looms larger because all the exosolar planets so far detected have masses comparable to that of Jupiter, a giant planet that orbits the Sun at more than five times the Earth-Sun distance. Let us take a moment to examine the facts, before we admire the astrophysicists’ explanations of how these planets may have come to occupy orbits so much smaller than those familiar to us in our own planetary system.

Whenever we use the star dance method to search for planets around other stars, we must remain aware of the biases built into this method. First, planets close to their stars take much less time to orbit than do planets far from their stars. Since astrophysicists have limited amounts of time with which to observe the universe, they will naturally discover planets moving in, for example, six-month periods far more quickly than they can detect planets that take a dozen years for each orbit. In both cases, the astrophysicists must wait through at least a couple of orbits to be certain that they have detected a repeatable pattern of the changes in the stars’ velocities. To find planets with orbital periods comparable to Jupiter’s twelve years could therefore consume much of an individual’s professional career.

Second, a planet will exert more gravitational force on its host star when close rather than when far. These greater forces make the star dance more rapidly, producing larger Doppler shifts in their spectra. Since we can detect larger shifts more easily than smaller ones, the closer-in planets attract more attention, and do so more rapidly, than the farther-out planets do. At all distances, however, an exosolar planet must have a mass roughly comparable to Jupiter’s (318 times Earth’s) to be detected by the Doppler shift method. Planets with significantly less mass cannot make their stars dance with a speed that rises above the threshold of detectability by today’s technology.

In hindsight, then, no surprise should have accompanied the news that the first exosolar planets to be discovered all have masses comparable to Jupiter’s, and all orbit close to their stars. The surprise lay in just how close many of these planets turned out to be—so close that they take not several months or years to complete each orbit, as the Sun’s planets do, but only a few days. Astrophysicists have now found more than a dozen planets that complete each orbit in less than a week, with the record holder sweeping out each orbit in just over two and a half days. This planet, which orbits the Sun-like star known as HD73256, has a mass at least 1.9 times Jupiter’s mass, and moves in a slightly elongated orbit at an average distance from its star equal to only 3.7 percent of the Earth-Sun distance. In other words, this giant planet possesses more than 600 times Earth’s mass at a distance from its star less than one tenth that of Mercury.

Mercury consists of rock and metal, baked to temperatures of many hundred degrees on the side that happens to face the Sun. In contrast, Jupiter and the Sun’s other giant planets (Saturn, Uranus, and Neptune) are enormous balls of gas, surrounding solid cores that include only a few percent of each planet’s mass. All theories of planet formation imply that a planet with a mass comparable to Jupiter’s cannot be solid, like Mercury, Venus, and Earth, because the primordial cloud that formed planets contained too little of the stuff that can solidify to make a planet with more than a few dozen times the mass of Earth. The conclusion follows, as one more step in the great detective story that has given us exosolar planets, that all exosolar planets so far discovered (since they have masses comparable to Jupiter’s), must likewise be great balls of gas.

Two questions immediately arise from this startling conclusion: How did these Jupiter-like planets ever come to orbit so close to their stars, and why doesn’t their gas quickly evaporate under the intense heat? The second question has a relatively easy answer: The planets’ enormous masses can retain even light gases heated to temperatures of hundreds of degrees, simply because the planets’ gravitational forces can overcome the tendency of the atoms and molecules in the gas to escape into space. In the most extreme cases, however, this contest tips only narrowly in favor of gravity, and the planets lie just outside the distance at which their stars’ heat would indeed evaporate their gases.

The first question, of how giant planets came to orbit so close to Sun-like stars, brings us to the fundamental issue of how planets formed. As we have seen in Chapter 11, theorists have worked hard to achieve some understanding of the planet-formation process in our solar system. They conclude that the Sun’s planets accumulated themselves into being, growing from smaller clumps of matter into larger ones within a pancake-shaped cloud of gas and dust. Within this flattened, rotating mass of matter that surrounded the Sun, individual concentrations of matter formed, first at random, but then, because they had a density greater than average, by winning the gravitational tug-of-war among particles. In the final stages of this process, Earth and the other solid planets survived an intense bombardment from the last of the giant chunks of material.

As this agglutinative process unfolded, the Sun began to shine, evaporating the lightest elements, such as hydrogen and helium, from its immediate neighborhood, and leaving its four inner planets (Mercury, Venus, Earth, and Mars) composed almost entirely of heavier elements such as carbon, oxygen, silicon, aluminum, and iron. In contrast, each of the clumps of matter that formed at five to thirty times Earth’s distance from the Sun remained sufficiently cool to retain much of the hydrogen and helium in its vicinity. Because these two lightest elements are also the most abundant, this retentive ability produced four giant planets, each with many times Earth’s mass.

Pluto belongs neither to the class of rocky, inner planets nor to the group of outer gas-giant planets. Instead, Pluto, still uninspected by spacecraft from Earth, resembles a giant comet, made of a mixture of rock and ice. Comets, which typically have diameters of 5 to 50 miles rather than Pluto’s 2,000 miles, rank among the first sizable clumps of matter to form within the early solar system; they are rivaled in age by the oldest meteorites, which are fragments of rock, metal, or rock-and-metal mixtures that happen to have struck Earth’s surface and to have been recognized by those who know how to tell a meteorite from a garden-variety rock.

Thus the planets built themselves from matter much like that in comets and meteorites, with the giant planets using their solid cores to attract and retain a much larger amount of gas. Radioactive dating of the minerals in meteorites have shown that the oldest of them have ages of 4.55 billion years, significantly greater than the oldest rocks found on the Moon (4.2 billion years) or on Earth (just less than 4 billion years). The birth of the solar system, which therefore occurred about 4.55 billion
B.C.
, quite naturally led to the segregation of planetary worlds into two groups: the relatively small, solid inner planets and the much larger, more massive, mainly gaseous giant planets. The four inner planets orbit the Sun at distances of 0.37 to 1.52 times the Earth-Sun distance, while the four giants remain at the much greater distances, ranging from 5.2 to 30 times the Earth-Sun distance, which allowed them to be giants.

This description of how the Sun’s planets formed makes such good sense that it almost seems a shame that we have now found so many examples of objects with masses similar to Jupiter’s, moving in orbit around their stars at distances much less than Mercury’s distance from the Sun. Indeed, because the first exosolar planets to be discovered all had such small distances from their stars, for a time it appeared as though our solar system might prove the exception, rather than the model of planetary systems, as theorists had implicitly assumed in the days when they had nothing else on which to base their conclusions. Understanding the bias imposed by the relative ease of discovering close-in planets gave them some reassurance, and before long they had observed for sufficiently long times, and with sufficient accuracy, to detect gas-giant planets at much greater distances from their stars.

Today, the list of exosolar planets, ordered by distance from the star to the planet, begins with the entry described above, of a planet that takes only 2.5 days to perform each orbit, and extends, through well over a hundred entries, to the star 55 Cancri, where a planet with at least four times the mass of Jupiter takes 13.7 years for each orbit. Astrophysicists can calculate from the orbital period that this planet has a distance from its star equal to 5.9 times the Sun-Earth distance, or 1.14 times the distance from the Sun to Jupiter. The planet ranks as the first to be found with a distance from its star greater than the Sun-Jupiter distance, and therefore seems to provide a planetary system roughly comparable to our solar system, at least so far as the star and its largest planet are concerned.

However, this is not quite so. The planet that orbits 55 Cancri at 5.9 times the Earth-Sun distance represents not the first but the
third
to be discovered in orbit around this star. By now, astronomers have accumulated sufficient data, and have grown so skilled at interpreting their Doppler shift observations, that they can disentangle the complex star dance produced by two or more planets. Each of these planets attempts to impose a dance in its own rhythm, with a repetitive period equal to the span of the planet’s orbit around the star. By observing for a sufficiently long time, and by employing computer programs that fear no calculation, planet hunters can tease from comingled dances the basic steps induced by each orbiting world. In the case of 55 Cancri, a modest star visible in the constellation called the Crab, they had already found two closer-in planets, with orbital periods of 42 days and 89 days and minimum masses of 0.84 and 0.21 Jupiter masses, respectively. The planet with a minimum mass equal to “only” 0.21 Jupiter masses (67 Earth masses) ranks among the least massive yet detected; but the record low mass for an exosolar planet has now fallen to 35 Earth masses—still so many times greater than Earth’s that we should not hold our breath in anticipation that astronomers will soon find Earth’s cousins.