Eyewitness to Stellar Evolution.
The late November night presented a quiet sky with the stars shining serenely, their steady light belying any propensity for change. I was working on a program to study planetary nebulae with a 1-meter (40-inch) telescope. In the course of this project I had made a hobby of compiling a detailed light curve for the 9th-magnitude "symbiotic star" AG Draconis. About once a week, Midwest weather permitting, I would measure its brightness in 11 colors. Symbiotics are double stars composed of an aging giant bound to a smaller, hotter star, usually a tiny, ultradense white dwarf. Hot matter streams violently from the big star into a disk around the small one. The result is a star whose spectrum is that of a cool giant coupled with weird high-temperature features - an unusual object, one well worth following.
The high northerly declination of AG Draconis, 66[degree sign], made it circumpolar, so I could track it through every season. For more than two years the star had been varying slowly in synchrony with its binary orbit. I had known that it was capable of rare, sudden outbursts - which in fact were once studied by the editor in chief of this magazine. But it never occurred to me that I would witness an eruption. That happened only to other, luckier people.
It was late, and AG was already way over in the northwest, forcing me to stretch from the catwalk to the telescope across a 20-foot drop to line up the star in the finder eyepiece. But it seemed that I must have made a mistake and gotten the wrong star. The photometer counts were much too high. Another, more dangerous stretch yielded a better view of the familiar field - but with my star strangely bright! It finally dawned on me that I was seeing it in an outburst. There it was, a mini-nova, one of the most exciting sights I could imagine.
It was a vivid reminder that stars do not always stay the same, even if in this case binary action was involved and the star, or rather the double, would return to nearly its original condition.
Truly permanent change in a star, astronomy books often suggest, is not to be expected in human lifetimes. Stellar evolution takes place over such long ages that from the human standpoint, the Sun and stars seem to stay the same forever. Even our variable-star friends need to be watched carefully to sense any alteration in their patterns - Delta Cephei pulsing every 5.37 days, Algol dipping every 2.87 days, Mira returning to Cetus about once a year. Only when a rare explosive event temporarily alters a familiar constellation might we ponder the awesome processes of stellar evolution that can forever transform one kind of star into another.
Certain types of changes, however, can truly betoken stellar evolution in action - showing us signs of the gradual aging of the universe. What might a dedicated observer see? Can we actually witness this action, without waiting for such obvious and sudden events as the violent explosions of supernovae, in which a star destroys itself completely?
Most stars reside on the "main sequence," living quietly in the long, stable phase of their lives when they generate energy by fusing hydrogen into helium deep in their cores. The main sequence gets its name from its place on the so-called Hertzsprung-Russell diagram, which has been a fundamental tool of astrophysics ever since it was invented around 1911. The "H-R diagram" plots the intrinsic brightnesses of stars against their temperatures (or colors or spectral types); an example is shown on the facing page. Running diagonally across the diagram, the main sequence includes an extremely wide range of stellar characters - from brilliant hot, blue monsters of spectral type O at top left to the dim, barely visible embers of red dwarfs, type M, at bottom right.
The huge variety of main-sequence stars results merely from differences in their masses. The range is from nearly 100 times the Sun's mass for the rare, hot dazzlers that shine far across the Milky Way galaxy, down to 0.08 solar mass for the dimmest red dwarfs, none of which can be seen with the naked eye, even those next door. A star spends most of its life on the main sequence until its hydrogen fuel begins to run out, but some stars stay there much longer than others. The most massive and luminous stars remain on the main sequence for a few million years; the smallest dwarfs will hoard their fuel for trillions of years.
Although main-sequence stars are amazingly stable (a fact that has allowed astronomers to evolve on Earth), they do change. The Sun, for example, has increased its brightness by about 30 percent since its birth 4.6 billion years ago. Such slow changes are not noticeable during human history or even across the most commonly studied spans of geologic time. So we pay stellar evolution little heed. The Sun will be the same tomorrow and a million years from now.
Scattered here and there, however, are stars that do not fit onto the main sequence: stars that are too bright and large, or too dim and small, for their masses. Cool giant stars can be as big as the inner solar system, and some supergiants could engulf much of the outer solar system as well. White dwarfs, so dim that only a few are in reach of amateur telescopes, are hardly larger than Earth. Neutron stars, invisible to all but the largest optical telescopes, are the size of a town. Yet all of these have masses not very different from the Sun.
Nearly all these non-main-sequence types result from the stellar aging process. Its extremely slow pace prevents us from seeing any one star move through its life cycle from birth to death. Instead, armed with the knowledge of how stars work and how they are structured inside, astronomers can string together the different star types to form freeze-frame sequences of complete life cycles - just as people in a large crowd could be sorted to show all the stages of human lifetimes.
When a main-sequence star finally begins to run low on hydrogen, its helium core shrinks and brightens, forcing the star's outer parts to balloon enormously. This giant phase lasts only about a tenth as long as the star's earlier residence on the main sequence. The core grows hot and dense enough to begin converting its helium - the dead "ash" of hydrogen fusion - into carbon and oxygen, releasing energy anew. Eventually the star ejects its outer envelope to produce a planetary nebula, leaving the core exposed as a new white dwarf.
In a massive enough giant the story becomes more complicated. The core goes on to fuse carbon into heavier and heavier elements - neon, magnesium, silicon, and others, and finally iron. An iron core cannot produce any further energy by fusion no matter how hot and dense it becomes. Unable to support itself, it collapses violently to set off a supernova outburst. The collapsed core may remain in the form of a tiny neutron star detectable as a spinning pulsar. Or, if it is massive enough, it punches an infinitely deep gravitational well in the fabric of space-time and leaves behind a black hole - a star never to be seen again.
Here and there, however, in the immense, slow pageant that moves a star from birth to death, there are quicksteps - crises when a star suddenly increases its pace of evolution. If we search carefully enough we can find these events happening before our eyes. We want to look for variability that results not from repeated cycles but from one-way changes that betoken the aging process itself. Let's watch.
At the Beginning: T Tauri and FU Orionis
Stars are born from the condensation of dusty clouds in interstellar space. As a massive cloud contracts under its own gravity it breaks up into small blobs. The central part of each blob goes on collapsing until it heats up and begins glowing, while its outer part flattens into an orbiting disk. The disk (the last remains of which may condense into planets) pours matter inward, feeding the growing star. This infall is highly irregular and chaotic. The result is a flickering T Tauri star. It is wildly bright in the infrared because of all the surrounding warm dust and in the ultraviolet because of matter crashing onto its surface.
The variations we see in T Tauri stars are not evolutionary changes as such, only instabilities in the disk. However, the disk occasionally becomes really unstable and begins to dump huge quantities of matter onto the waiting star below. Sixty years ago, a modest 16th-magnitude T Tauri star (or so we believe) now named FU Orionis brightened by 6 magnitudes over a period of a year. Its rise was easily visible from week to week. FU Orionis has remained bright ever since. V1057 Cygni did the same thing starting in 1970 (but has since faded somewhat).
During an "FU Orionis" phase a star can accumulate another percent or so of what will be its final mass. Perhaps another T Tauri star will flare up within your lifetime; if so, you will see it changing forever. If T Tauri itself ever performs such an act it should reach naked-eye brightness.
Delta Cephei's pulsations are regular enough to use as a clock. The pattern of variation repeats almost perfectly as it changes from visual magnitude 3.6 to 4.3 and back in slightly more than 5 days. Cepheids abound; several, including Eta Aquilae and Zeta Geminorum, can be followed with the naked eye. These are midtemperature supergiants, aged stars that began life with higher masses than they have now and are currently caught in an unstable state.
Inside such a star, radiation gets absorbed by a layer in which helium is becoming ionized (electrons are stripped from helium atoms). The absorption of energy raises the layer's heat and pressure, causing the star to expand. The expansion in turn cools the layer and lowers the pressure, so the star contracts again and begins the cycle anew. The star cannot find a stable state.
The Cepheid phenomenon occurs only for giants in a particular state of evolution and having a narrow range of surface temperatures (around 5,000[degree sign] to 6,000[degree sign] Kelvin). The "instability strip" on the Hertzsprung-Russell diagram is sharply defined; it is sketched in on page 43. All stars that begin life with about 5 to 20 solar masses eventually pass through the instability strip, some more than once. Given the pace of stellar evolution, we could wait a very long time for Delta Cephei to change its habits. But what if a star were balanced precariously at the edge, or between two types, of instability?
Polaris is the least-recognized bright Cepheid; perhaps its position at the celestial pole implies constancy. In fact its variations have always been slight, a tenth of a magnitude at most, with a period of 3.97 days. Nevertheless Polaris seems to be a star whose actual evolution is in some way visible. Over the course of this century its variations have gradually faded almost away. By the 1990s Polaris was varying by only a few hundredths of a magnitude - though Doppler measures show the star continuing to pulse slightly with only a small increase in its 3.97-day period (January issue, page 18). Is Polaris ceasing to be a Cepheid? Will it stop pulsing altogether?
Polaris is not alone in its ability to change. The huge globular clusters of our galaxy's halo contain Cepheid-like (but much lower-mass) RR Lyrae stars, which pulse with periods of a day or less. Their oscillations can be understood by analogy with the vibrations produced by musical instruments. The strongest frequency of a musical note played on a violin is its fundamental, which sets the pitch. Higher frequencies, or overtones, add quality, or timbre, and help distinguish a violin from an oboe. One outlandish RR Lyrae star was seen to switch from the fundamental to a higher overtone of pulsation practically overnight, the result of some slight change in its overall size and density. Whatever these stars are doing, we are clearly seeing stellar evolution in action.
Planetary nebulae abound in our galaxy; several are bright showpieces in amateur telescopes. Yet they are remarkably ephemeral. Expanding at 20 kilometers per second, they disperse within a mere 50,000 years or so. We know of so many because their central stars are among the hottest in the galaxy, lighting the nebulae brightly enough with their ultraviolet output for the nebulae to be seen across thousands of light-years.
As a planetary nebula expands, its central star at first heats to well over 100,000[degree sign] K while maintaining a constant luminosity - typically several thousand times that of the Sun. The star then cools, and as its temperature declines to 50,000[degree sign] K or so its luminosity drops to a few tens of Suns. As the nebula disperses into the interstellar gloom, the star settles down into life as a white dwarf. Its fate is to continue cooling forever.
Although 50,000 years is about 10 times longer than recorded human history, it is short enough that some astronomers believe they have detected slow, long-term changes in some planetary-nebula central stars. More rarely, changes happen in remarkably short spans. FG Sagittae is the central star of a faint, large planetary. From 1900 to 1970 FG Sagittae brightened gradually from magnitude 13.5 to 9.5. In 1955 it was classified as a hot supergiant, but by 1983 it had cooled to a temperature similar to that of the Sun. In 1992 it began a series of irregular, precipitous drops in brightness, perhaps as thick, dark dust clouds condensed in its atmosphere.
Stranger still were the changes in FG Sagittae's spectrum. Stars near the solar temperature show numerous narrow, dark absorption lines in their spectra. Many of these lines are the signatures of common metals such as iron, chromium, titanium, and vanadium. Accompanying the other changes, FG Sagittae's spectrum exhibited increasing amounts of barium, zirconium, yttrium, and several "rare earths," including such underappreciated elements as cerium, praseodymium, neodymium, promethium, samarium, and gadolinium.
The rare-earth elements are truly rare: within the Sun there is one atom of cerium, the most common, for every 1.3 million iron atoms. But in FG Sagittae they dominate the spectrum. Yttrium and barium are enriched well beyond 30 times normal. At the same time the common "iron-peak" elements, which include nickel, have nearly faded away. The unusual elements are produced within a star's core by the slow capture of neutrons - a process that changes one element to the next-heaviest to the next in the periodic table of elements. We are seeing the products of these nuclear reactions being dredged up to the surface of the star by convection.
Maybe FG Sagittae is in an especially rapid evolutionary state, in which a helium-rich layer surrounding a dead carbon core suddenly erupted in a frenzy of nuclear fusion. The result may be a backtrack of evolution in which a white dwarf reinflates to become a giant! Its newly lowered surface gravity will cause more matter to be lost, and astronomers in the distant future may see a small planetary nebula develop within the larger one.
Such strange double planetaries are indeed known. The central stars of the large planetary nebulae Abell 30 and Abell 78 are surrounded by small blobs of gas that contain no hydrogen - apparently the ejecta of such "born-again" giants. In FG Sagittae, and in a similar, recently discovered object known as Sakurai's Star (S&T: August 1997, page 20), we may be seeing giants on the rebound - right before our eyes.
Monsters in Their Death Throes
Seemingly modest Rho Cassiopeiae is one of the galaxy's brightest stars, a half million times more luminous than the Sun. Just west of the W of Cassiopeia, it shines at a naked-eye magnitude of 4.5 despite its distance of 8,000 light-years and interstellar absorption. Normally its surface temperature remains about 7,000[degree sign] K, a bit hotter than the Sun. In 1945, however, the star faded to nearly the limit of naked-eye visibility and cooled to only 3,000[degree sign] K. It apparently ejected a cloud of gas that condensed into smoky dust, burying the star within. After two years Rho Cas seemingly returned to normal, but we know that a great quantity of its mass was lost forever and that the star must now be further along its evolutionary path. It will happen again, so keep an eye on the celestial Queen Cassiopeia as she rounds the pole.
A grander version of this event took place in Eta Carinae during the 19th century. Now 6th magnitude, the star itself is buried in a tiny, irregular nebula that the Hubble Space Telescope revealed as a spectacular smoky hourglass, with two lobes flowing in opposite directions away from a star within. Eta Carinae brightened from 4th magnitude during the early 1800s to match Canopus at about magnitude -0.8 in 1848. That year it was second only to Sirius as the brightest star in the night sky. Then within 30 years it plummeted to invisibility.
Like Rho Cassiopeiae, something in the Eta Carinae system ejected what appears to be twice the mass of the Sun. Eta itself is one of the most massive stars known, perhaps as heavy as 100 Suns. There is good evidence that a second star in the system, now invisible, was the one that ejected the nebula (February issue, page 26). This star must have lost 2 or more percent of itself. It cannot possibly be the same as it was a mere 200 years ago. In any case, both the main star of Eta and its presumed eruptive companion are progressing to what will probably be two spectacular supernovae.
Even in Death
When stars like the Sun die they end up as Earth-size white dwarfs. The remains of massive stars are mostly neutron stars. But these classes of "dead" stars are hardly lifeless. Perversity in language being a prime trait of astronomy, they are in their way still very much alive.
As the tale is usually told, white dwarfs turn into "black dwarfs," cold cinders that have radiated their heat away. There are no such things. The cooling time of a white dwarf - massive yet possessing only a very small surface from which to lose heat - is so long that even the oldest of them still shine yellow-orange hot. All white dwarfs are still evolving, albeit slowly. The same is true for pulsars, spinning neutron stars that lose speed as they fling away particles and energy from the edges of their magnetic fields.
Remarkably, such effects can be watched. There are two kinds of white dwarfs, those that have helium surfaces as a result of losing all hydrogen, and those with hydrogen-rich surfaces due to the inward settling of heavier helium. When a hydrogen-rich white dwarf cools through the range from 13,000[degree sign] to 10,000[degree sign] K it begins to quiver and chatter, rather like a complex Cepheid, with many similar periods of only a few minutes each. In a few cases astronomers have been able to watch these oscillations slowly evolve, and from the rate of change they have found the rate at which a white dwarf cools. It is right in accord with theory.
Neutron stars are the best natural clocks in the universe. The one in the Crab Nebula, the result of the supernova of 1054, spins 32 times per second; many others spin roughly once a second. Most remarkable are the "millisecond" pulsars, which have been spun up to incredible rotation speeds by accreting matter in a close binary system. Their spin periods can be extremely precise. And many are seen to be gradually slowing down with age, at precisely measurable rates. For instance, the fastest known pulsar, PSR B1937+21, was turning once every 0.001557806468819794 second as of 0:00 Universal Time December 5, 1988. At that moment the pulsar was slowing down at a rate of one part in 469,608,000 per year. Tracking such precise behavior taxes the best atomic clocks that technology has devised.
Although it may take sophisticated observations to detect these and other changes, the instruments that astronomers use are no more than extensions of the human eye and mind. Through our instruments and our research astronomers, we can see that the "eternal" starry population around us is not static. Although the changes wrought by stellar evolution are almost imperceptibly slow, we can indeed witness them happening as the universe ages before our eyes.
James Kaler, professor of astronomy at the University of Illinois, is the author of Stars and Cosmic Clouds (Scientific American Library) and The Ever-Changing Sky (Cambridge University Press). He explores a "star of the week" on the Web at www.astro.uiuc.edu/~kaler/sow/sow.html.