In Chapter 1 we asserted that one facet which makes scientific discover possible is the possession of an intuitive feeling for nature, particularly for its quantitative aspects. The Sun, being the closest star to us, has fashioned and refined that intuitive feeling that astronomers possess for the nature of stars when they approach their study. Because other stars are immensely far away, astronomers, extrapolating and feeling their way forward by analogy with the Sun, have pursued a study of stellar surface features even when they could not directly observe their surfaces. Hence, the Sun has been a vital bridge to the world of stars and the variety of phenomena that must be occurring in their outer layers. Even today in the largest telescopes, stars are still basically just points of light. However, over the last twenty years innovative technology has made it possible to study the outer layers of other stars at a level of detail that is still crude compared to what we can achieve for the Sun, but is, nevertheless, a dramatic improvement over earlier achievements. Let us begin this chapter by surveying the Sun's surface layers and then extending that discussion to what we have learned about other stars. Table 15.1 contains some of the important attributes of the Sun as a star to keep in mind throughout our discussion.
[Table 15.1]
15.1. Fine Structure in the Solar Photosphere
The solar photosphere is a transition layer from the invisible interior to the external environment surrounding the Sun. Through this 500-km thick layer temperature declines outward by several thousand Kelvins. The average temperature, or surface temperature as it is known, is about 6000 K as derived from the observed luminosity by using the blackbody radiation laws. The reason that the solar photosphere approximates a blackbody is because it quickly becomes opaque as we probe deeper into it. Photospheres of other stars also approximate blackbodies to a greater or lesser extent than the Sun. However, when we look closely at the solar photosphere, we find a far more detailed structure than one might presume from something as bland as the blackbody radiation laws.
Even in photographs of the solar photosphere taken in white light a number of features are evident. We have already noted that the photosphere darkens toward the limb; near the limb we can also see bright patches called faculae. And in high-resolution white-light photographs, one can see that the entire disk is covered at all times by small, bright features separated by dark lanes called granules (Figure 15.1b).
[Figure 15.1]
Remarkably clear pictures of the solar surface in narrow wavelength ranges have been made by a telescope mounted in the Skylab station. Fine details on the surface are not blurred by the Earth's atmosphere at the altitude that the station was orbiting. Such high-resolution photographic studies reveal a potpourri of bright granules with dark intergranular lanes; these give the surface the honeycombed appearance visible in Figure 15.1b. Time sequences of photographs show granules forming, disappearing, and re-forming in cycles lasting several minutes. At any given time the whole photosphere is broken up into better than 4 million granules, each occupying roughly 1 million km2 of the surface. Obviously the photosphere is not a uniform layer of gas; the temperature of the photosphere must vary not only in depth but also laterally across the face of the Sun.
The granules are cells of gas with characteristic diameters of 1000 km and lifetimes of several minutes. From the bright center of the granule to the darker intergranular region, the brightness variation corresponds to a temperature difference of about 200 K. Photospheric granules are a form of convection resulting from the upwelling of unstable convective elements from the hydrogen convection zone below the photosphere. As evidence of this convective exchange, spectral lines, such as shown in Figure 15.1c, from the bright centers are Doppler-shifted toward the blue (coming toward us) and those from the dark intergranular regions are shifted toward the red (moving away from us). These Doppler shifts indicate that the bright centers are hot rising gas moving at a few tenths of a kilometer per second that radiates its excess energy and then forms the cool sinking gas in the dark intergranular lanes.
In 1960 vertical oscillatory motions were detected in and above the solar granulation which possess a period of almost exactly 5 minutes with velocities of about 0.5 km/s. Thus the layers above the hydrogen convection zone are moving up and down with respect to the mean position of the photosphere and low chromosphere. The typical excursion is on the order of 50 to 100 km. The motion seems to be organized over a few thousand kilometers and has been reported to cover areas as large as 50,000 km, with roughly two-thirds of the solar surface experiencing oscillations at any given moment. In 1984, the Sun's closest stellar neighbor, Alpha Centauri, was also shown to be also undergoing 5-minute oscillations. It now appears that the 5-minute oscillation is but one extreme in a range of oscillations, with a 160-minute oscillation as the other extreme. Thus the Sun quivers much like a bowl of gelatin; the consequences of these oscillations will be taken up in Section 17.3.
Dark features on the Sun have been reported for at least 2000 years. Several sightings per century are contained in ancient Chinese records. Although not the first sightings recorded in Europe, Galileo's telescopic observations in 1610 provided the first details on sunspots, the most conspicuous of a number of transient phenomena to be found in the solar atmosphere.
A typical sunspot has a cellular structure with a dark center, the umbra, surrounded by a grayish filamentary region, the penumbra. Although sunspots emit radiation, the umbra looks dark because it is seen against an even brighter photospheric background, whose temperature is some 1800 K higher. The umbra is about one-fourth as bright as the photosphere and the penumbra about three-fourths as bright. A large sunspot group is shown in Figure 15.2.
[Figure 15.2]
Sunspots develop in a matter of hours as small pores in the intergranular region of the photosphere. They grow rapidly, and they generally form in clusters, marking a sunspot group, whose orientation is approximately parallel to the solar equator. Each end of the group is often dominated by a large spot surrounded by smaller spots. The very largest groups may cover up to one-fifth the solar diameter. Sometimes a sunspot group persists for several months, but a typical lifetime is about 1 week. A typical large spot in a group is some 10,000 km across; exceptional ones are 50,000 km in diameter, or about four times the diameter of the Earth. In a week or so this large spot builds to its maximum diameter; then its size slowly declines. Individual spots in a sunspot group undergo slow changes from day to day while they maintain their association.
More than a century ago it was discovered that sunspots come and go in a roughly 11 year cycle. The sunspot number plotted in Figure 15.3 shows many cycles of this ll year variation. From the figure it is obvious that the heights of successive maxima are unequal, and the interval between successive peaks or troughs is not constant; the 11-year period is a very rough average. Each sunspot cycle opens with spots forming at latitudes around 35oN and 35oS, and as the cycle progresses, spots form closer to the equator in both hemispheres as shown in Figure 15.4. The maximum number of spots form when sunspots are forming at latitudes around 25o in both hemispheres. When the last spots of a cycle are forming near 5oN or 5oS, a few spots again form at latitudes around 35oN or 35oS herald the beginning of a new cycle.
[Figure 15.3]
[Figure 15.4]
Immense arching, curved features are observe around sunspot groups (Figure 15.8d). They are structures of gas whose shape is determined by curved magnetic field lines, since sunspots are known to be the centers of intense magnetic fields. Astronomers know this because of the Zeeman effect (Section 12.5). It was first identified in the absorption lines of sunspot spectra by George Ellery Hale (1868-1938) at the Mount Wilson Observatory in 1908. In the Zeeman effect, the strength of the magnetic field can be determined from the separation of components of absorption lines, while the direction of the field is shown by the sense of polarization of these components (Figure 15.5).
[Figure 15.5]
The leading spots of a spot group (in the forward direction of the Sun's rotation) are opposite in polarity from the following spots. Opposite polarities are like those of the north and south ends of a bar magnet. The unit of measure of the intensity or strength of the magnetic field is called a gauss. The measured field strength in sunspots exceeds that of the Earth's field by several thousand times, being several thousand gauss. In the Sun's northern and southern hemispheres, polarities of leading and following spots are also opposite to each other. That is, the polarity of the leading spot in the northern hemisphere may be northseeking in one sunspot cycle, whereas that for the southern hemisphere is southseeking. Then in the next sunspot cycle the leading spot will be southseeking in the northern and northseeking in the southern hemisphere. Thus magnetic field polarity reverses in both hemispheres in succeeding sunspot cycles.
As a whole, the Sun does not have a general magnetic field like that of the Earth. But by averaging over small localized and intense magnetic fields, one gets the impression of a general field a few times stronger than the Earth's field, or several gauss. Magnetic measurements have been made for the Sun fairly regularly over the last 25 years. From which it has been found that magnetic fields in the polar regions reverse polarity near the time of sunspot maxima. Thus what appears to be a general field is probably the accumulation of surface fields that have drifted into the polar regions.
[Box - Origin of Sunspots, Figure 15.6]
15.2. Chromosphere of the Sun
During a total eclipse of the Sun, when the Moon has just covered the photosphere, a thin (about 2000 km) pinkish fringe of light, called the chromosphere, appears beyond the Moon's edge (Figure 14.7b). The chromosphere gets its reddish hue from the emission of radiation in the red-colored alpha line of the Balmer series of hydrogen. Projecting from the chromosphere here and there are rosy arches and loops of gas called prominences, which may extend 100,000 km or more into the overlying corona.
For the few seconds of an eclipse when only the chromosphere is exposed, the photosphere's normal absorption spectrum is no longer visible, and we see an emission spectrum called the flash spectrum (Figure 15.7). It is the spectrum of light originating in the chromosphere. Many of the emission lines match the wavelengths of the absorption lines, but among the exceptions is a bright yellow line produced by helium. Why do helium emission lines appear in the chromospheric flash spectrum but not in the photospheric absorption spectrum? The reason is the chromosphere's higher temperature--up to 30,000 K at the highest level--and lower density. Neutral helium can be excited to emit radiation only when the temperature is greater than 10,000 K. And the appearance of ionized helium lines requires temperatures in excess of 20,000 K. From this we conclude that the temperature must rise rapidly from the top of the photosphere up through the chromosphere.
[Figure 15.7]
Chromospheric events can be monitored, even when there is no eclipse, by photographing the chromosphere in monochromatic light. A single-wavelength photographic device, the spectroheliograph, can make pictures of the solar disk in the residual light of an absorption line, such as the red hydrogen alpha line or the violet K line of singly ionized calcium (Figure 15.8). Such a picture, called a spectroheliogram, is shown for red hydrogen alpha in Figure 15.8b and for the violet calcium K line in Figure 15.8c.
[Figure 15.8]
The reason for choosing the residual light of a strong absorption line is that most of the photons from the photosphere have been absorbed and the remaining ones are those coming from the low chromosphere. Photons in the continuous spectrum come from the bottom of the photosphere, whereas photons in stronger and stronger absorption lines come from higher and higher up in the photosphere and the low chromosphere. By choosing absorption lines of different strengths, we can photograph different levels of the Sun's atmosphere.
Photographs of the chromosphere show us a very different view of the Sun from what we see in the photosphere. Bright patches in the chromosphere, called plages, are seen to overlie photospheric sunspot groups (Figure 15.8d). Hotter and probably more dense than the normal chromosphere, with a spatially averaged magnetic field of a few hundred gauss, plages are typically 10 times larger than the sunspot group lying below them. Plages are nearly always found above regions in the photosphere in which a strong magnetic field exists, and they always appear before the spots form. Their usual life span is about 40 to 50 days, during which several spot groups or none at all may form. The plage areas do not look the same in the red hydrogen alpha line (Figure 15.8b) and in the violet calcium K line (Figure 15.8c), but they obviously mark the same general region of the chromosphere.
In a violet calcium K line spectroheliogram one sees a network of bright gas surrounding dark cells in addition to the larger plages. This chromospheric network constitutes the boundaries of large-scale convective cells, known as supergranules, which are seen as Doppler shifts in photospheric pictures such as Figure 15.9. (Supergranules derive their name because of their resemblance to convective motions and the fact that they are typically an order of magnitude larger than granules; beyond their names, granules and supergranules have little in common.) Apparently the chromospheric network is the locus of very intense and highly localized magnetic fields that are concentrated on the boundaries of the supergranule cells by the motions of gas.
[Figure 15.9]
In short-exposure spectroheliograms, the chromosphere appears stippled with a myriad of jetlike spikes of gas, called spicules (Figure 15.10). Spicules rise rapidly from the chromospheric network, attaining typical heights of 10,000 km, and then they fade away or collapse in several minutes (Figure 15.11). At any instant, 250,000 of them may cover a few percent of the Sun's surface. They too outline the boundaries of the supergranule cells. Figure 15.10 is a red hydrogen alpha line spectroheliogram containing short dark features, like blades of grass, that are thought to be spicules outlining the bright interiors of supergranule cells. Skylab photographs suggest that the chromosphere may have granulelike features much larger than those observed in the photosphere.
[Figure 15.10]
[Figure 15.11]
Solar flares are perhaps the most complex of the Sun's transient phenomena. They vary in size, brightness, and behavior, and flaring activity is most common when sunspots are most numerous. A solar flare may suddenly erupt as an intensely bright area in a chromospheric plage (Figure 15.12) that will send material hurdling outward into space. Emitting radiation strongly throughout the electromagnetic spectrum, flares rise to great brilliance in several minutes and then fade in half an hour to several hours.
[Figure 15.12]
The current view of solar flares is that they result from the sudden release of energy stored in coronal magnetic fields above sunspot groups. That is, the radiant and thermal energy, which for a large flare may amount to as much as 1032 erg, comes from energy stored in twisted magnetic field lines. If we think of magnetic field lines as being like the rubber band of a toy airplane, then the storage of energy is analogous to winding the rubber band. There seems to be a limit to how much twisting field lines can tolerate. When that limit is reached, it is only a matter of seconds to minutes before the energy stored in the field is released as thermal energy of motion, just as releasing the rubber band turns the propeller of the toy airplane.
In flares, free electrons are accelerated up to velocities of about half the speed of light. As these energetic electrons collide with ambient gas, they share their kinetic energy and heat the gas to a few thousand degrees in the chromosphere and to as much as 20 million K in the low corona. This heating phase may last from seconds to minutes and is responsible for the X-ray, ultraviolet, and visible radiation emitted by flares. Some high-energy electrons pass out through the corona, where they excite successive layers to emit radio-frequency radiation.
A major flare is the most energetic of all solar events, equaling
in power a 100 million hydrogen bombs with a yield of a 100 megatons
each. The ultraviolet and X-ray radiation arriving at the Earth
can disturb the ionosphere, persisting sometimes for hours. Over
the few days following the flare outburst, subatomic particles
may spiral into the Earth's polar regions, causing brilliant auroral
displays and radio blackouts. All these events, which do not
necessarily occur with every flare, partially distort the Earth's
magnetosphere, generate geomagnetic storms, and induce severe
electrical power surges. Solar flares are one of the few astronomical
events that directly disturb our terrestrial environment.
15.3. Corona of the Sun
The corona is that region of the solar atmosphere lying above the chromosphere. As seen in Figure 15.13, taken during a total eclipse, it is the large halo of white, glowing gas extending out a few solar radii (millions of kilometers) beyond the dark limb of the Moon. At times when an eclipse is not in progress, specially designed refracting telescopes, called coronagraphs, that block out light from the photosphere are used to observe the corona.
[Figure 15.13]
Compared with the many hours of almost continuous surveillance accumulated in satellite studies over the last 20 years, eclipse studies have yielded not more than a few hours of observations (an eclipse seldom lasts longer than a few minutes). Even coronagraphic studies, which have greatly increased observing time, cannot match the time coverage and resolution of satellite observations from above the atmosphere. In February of 1980, the launching of the Solar Maximum Mission spacecraft, one of the most sophisticated and complex satellites ever built, provided the opportunity to keep watch on the Sun during its period of maximum surface activity. However, in November of that year, Solar Max lost its attitude-control system, so that the craft could no longer point its eight instruments at interesting portions of the solar surface. Then in April of 1984, the crew of the Space Shuttle Challenger were able to wrestle Solar Max into the cargo bay of Challenger for repair (chapter opening). This first successful repair of a satellite in orbit opens a new era for space observatories, in that in the future they will be repaired or serviced in orbit or even returned to the ground for modification. Properly attended, Solar Max may well last into the 1990s or almost a complete solar cycle of 11 years.
Approximately 30 emission lines have been identified in the visible part of the coronal spectrum, and many hundreds of emission lines are known in the ultraviolet and X-ray spectrum. They originate in highly excited ions of familiar elements, such as iron, from which several to as many as 15 electrons have been stripped in the corona's extremely hot, tenuous gases. (It takes temperatures from many hundreds of thousands up to several million degrees to sustain such a degree of ionization.)
From millimeter to meter wavelengths there is a wide spectral window in the Earth's atmosphere that lets in radio radiation. The Sun, when quiet and undisturbed, normally emits thermal (blackbody) radiation, which is characteristic of a million-degree corona. When the Sun is disturbed, as when solar flares are occurring, nonthermal radio emission is added to the thermal component, and it can be quite intense.
There are several lines of evidence, besides the coronal spectrum, that confirms a rise in temperature through the chromosphere into the corona. Since heat flows from high to low temperature regions, then clearly energy must be pumped by some mechanism from the low-temperature photosphere to the high-temperature corona. For a number of years astronomers thought that the corona's high temperature resulted from energy carried into the corona by mechanical waves starting in the turbulent hydrogen convection zone below the photosphere.
As evidence grew that the magnetic fields of the photosphere and chromosphere were highly localized and very intense, it seemed hard to ignore the possibility that these magnetic fields extending up into the corona were part of the coronal heating process. When X-ray pictures showed that the corona was divided into active regions and hole regions primarily because of the structures of their magnetic fields, it became readily apparent that most, if not all, of the heating involves magnetic fields. The heating is produced by the direct dissipation of the energy stored in magnetic fields into thermal energy in the coronal gas. The lower parts of the chromosphere, however, are still thought to be heated by mechanical waves.
Eclipse pictures of the corona provide evidence for the importance of magnetic fields in structuring the corona. In white-light photographs one can see that the corona is irregular and structured (Figure 15.13). Beautiful, long streamers extend outward in the Sun's equatorial regions. Near sunspot maximum the corona is nearly circular, with streamers radiating out in all directions. Near sunspot minimum, the corona extends farther out in the equatorial region and terminates rather abruptly, with short, thin plumes curving out of polar areas.
Because coronal gases are almost transparent, we often are looking through several structures at once in eclipse pictures, which blurs the details. This is why direct photographs in X-ray and extreme ultraviolet wavelengths, where we look down on top of coronal structures, are so valuable to the study of the corona. To photograph the corona directly, we must observe in the 10 to 900 A (X-ray to extreme ultraviolet) wavelength region, where radiation from the much hotter corona overwhelms the short-wavelength radiation of the photosphere (Figure 15.14). But X-ray or ultraviolet pictures must be taken from space because the Earth's atmosphere absorbs these short wavelengths. In X-ray and ultraviolet pictures the corona appears highly inhomogeneous and generally asymmetrical, and it varies over time on both short and long time scales.
[Figure 15.14]
There are three different types of structural regions that collectively characterize the entire solar corona: Coronal holes and coronal active regions are two of them, and they are reasonably well defined in terms of observational characteristics; the third is the coronal quiet regions, which are not well defined. Coronal holes (Figure 15.14) are regions of slightly lower temperatures and significantly lower densities, with magnetic fields of about 10 gauss whose lines open out into interplanetary space. Rays and polar plumes extend out of them away from the Sun. Coronal holes are also thought to be the source of most of the subatomic particles in the solar wind. Coronal active regions, however, are extremely different from coronal holes. They consist of loop structures, they are somewhat hotter and much denser regions of the corona, and their magnetic fields of about 100 gauss having field lines that loop back into the Sun instead of extending outward as in the case of coronal holes. Between these two extremes are the ill-defined coronal quiet regions, which appear to be something in between. There are also many small, bright points visible in X-ray pictures, such as can be seen in Figure 15.14.
With time-lapse photography of the corona one sees spectacular motions of towering masses of luminous gas, called prominences, such as the one shown in Figure 15.15. Projected against the solar disk, they are the dark, threadlike filaments seen in Figure 15.8b. Their forms vary from almost stationary, quiescent arches and graceful loops to rapidly moving surges.
[Figure 15.15]
The typical prominence is hundreds of thousands of kilometers long and extends several tens of thousands of kilometers above the photosphere. It consists of gas cooler and denser than that in the corona around it. In the more active prominences, gas may rise at rates of hundreds to thousands of kilometers per second, which is sufficient to escape from the Sun. Frequently, however, matter appears to rain down from the corona in great luminous masses. Apparently, magnetic fields hold up these huge walls of gas against the Sun's gravitational pull. We can see matter flowing along the body of a prominence, following the curving and looping magnetic lines of force, as in Figure 15.15.
The mean lifetime of large quiescent prominences is about two
to three rotations of the Sun. During sunspot maximum, 20 filaments
may appear on the disk; during sunspot minimum, there are typically
about 4. Prominences always appear to be associated with a plage
or sunspot group. In fact, the large quiescent prominences tend
to form along the division between regions of different magnetic
polarity in plages. The polarity of each side of the plage region
is also the same as that of the sunspots in the photosphere underneath
it.
15.4. Activity Cycle of the Sun
During a sunspot cycle, the general level of all activity in the solar atmosphere follows the number of sunspots. Thus a sunspot group seen in white-light photospheric pictures is just the most visible indicator of a large disturbed region in the solar atmosphere called an active region. Such regions can be up to several hundred thousand kilometers in extent. The various transient phenomena that are part of the active region are summarized in Table 15.2 and are illustrated in Figure 15.16. The common bond among these visible features is magnetic fields. The field appears first, followed by a facula in the photosphere and a plage in the chromosphere. This in turn can be followed by a sunspot group, flare activity, and prominences. The precise behavior is somewhat different for each active region, but there is little doubt that the phenomena in different layers of the solar atmosphere are related. The solar cycle of activity (a succession of active regions over a 22-year period) is fundamentally a magnetic cycle. It is a many-year variation in the quantity of magnetic field that emerges in the solar atmosphere.
[Table 15.2]
[Figure 15.16]
The Sun's transient activity produces a variety of transient effects here on Earth, which follows a rough cycle in unison with the Sun's cycle. An example of this relation is the increase in auroral activity in the Earth's atmosphere during sunspot maxima and its decrease during minima. Changes in solar activity also affect Earth's weather and climate, but how and over what time scales these effects occur we do not clearly understand. In recent years some interesting discoveries have been made about the constancy of the Sun's activity and its relation to the Earth.
Approximately 1.4 million erg of radiant energy fall on each square centimeter of the Earth every second. The theory of stellar evolution, discussed in Chapters 17 through 19, suggests that the Sun's luminosity must have increased by perhaps 30 percent since it began its existence some 4.6 billion years ago. However, as recently as 1975, paleoclimatic evidence concerning long-term temperature variations on Earth showed that any change has been less than 3 percent in the last 1 million years. One can argue that if the solar luminosity had been as much as 25 percent less than the present value, the oceans would have frozen, preventing biological evolution. This suggests that if the solar luminosity has increased significantly since Earth's formation, something about the Earth has compensated for such a change.
Solar Max measurements provide evidence that sunspots and faculae, granulation, and solar oscillations are responsible for as much as 0.4 percent changes in the solar constant over a time scale of a week. And since 1980, the solar constant has steadily decreased by 0.02 percent per year. What longer-term changes exist in the solar constant, if any, are important to our knowledge of the Sun and the Earth.
Another fundamental question is the degree to which the 22-year cycle of solar transient activity repeats itself over long stretches of time. In 1893, E. Walter Maunder (1851-1928), a British astronomer, found in European historical records that very few sunspots were seen in the period from 1645 to 1715, now known as the Maunder Minimum. Within the last several years, Maunder's work has been confirmed and extended by astronomers worldwide. In addition to the absence of sunspots during the Maunder Minimum, very few aurora were observed in Europe, and during eclipses, the corona was also absent or very weak. Virtually no sunspots were reported in Asia during the Maunder Minimum, even though naked-eye sunspot-sighting reports exist there from as early as 28 B.C. Finally, measurements of the amount of carbon 14 (C146), a radioactive isotope of carbon, in tree rings show that during the Maunder Minimum there was an excess of this isotope in the Earth's atmosphere. We believe that the high-energy subatomic particles called cosmic rays, moving randomly through the Galaxy, collide with the nucleus of nitrogen atoms (N147) in our atmosphere, converting it to carbon 14. When the Sun is very active, the interplanetary magnetic field is strong, and Galactic cosmic rays are deflected away from the Earth. Thus high levels of carbon-14 in the Earth's atmosphere correspond to low levels of solar activity.
Historical research has shown a correlation among carbon 14 abundance in the tree rings, winter severity, Galactic cosmic-ray activity, and solar activity, as shown in Figure 15.17. Periods of colder climate appear to coincide with low levels of solar activity. Evidence exists that at least a dozen similar periods of minimal solar activity, lasting from 50 to 200 years, have occurred over the last 8000 years. Answers as to why the Sun should experienced decreases in transient activity are to be found in changes in magnetic activity which are probably caused by changes in the pattern of convection underneath the photosphere. Consequently the interior of the Sun may not be as constant in its behavior as once thought.
[Figure 15.17]
15.5. Stellar Activity
In recent years astronomers have come to realize that the hydrogen convection zone underneath the Sun's photosphere is responsible for its differential rotation. In turn, the differential rotation produces the 22-year magnetic cycle and all its associated activity seen in the photosphere, chromosphere, and corona are consequences of magnetic fields. Therefore, we should expect surface activity on other stars that also possess convective regions near their surfaces. Mathematical analyses of convection in stars suggests that stars of spectral types F, G, K, and M should possess a subsurface convection analogous to that of the Sun. Also, these analyses indicate that in stars of spectral types O, B, and A convection should occur near their centers and, consequently, should have no directly observable effects. Evidence has indeed been found for the existence of surface activity in other late-type stars analogous to the Sun.
Knowing that the Sun's photosphere possesses an intricate structure of granules, sunspots, and faculae, it was assumed several decades ago that if the surfaces of stars were directly observable one would see similar phenomena. But since they can not, in general, be observed directly, it has only been in recent years that technology, primarily computers, computerized data handling, and radiation detectors, have made it possible to seriously address the question of fine structure in stellar photospheres.
The means for such studies are the profiles of absorption lines in a star's spectra. The spectrum in Figure 15.1c is of light from a very small area on the solar disk. Imagine what would happen if light from many small regions across the solar disk were added together. Instead of the zigzag-shaped absorption line in Figure 15.1c, the absorption line profile from many areas on a star's disk would be smeared-out and thus be wider. Analyzing the profile of this smeared-out line by the variation in brightness over the narrow range of wavelengths of the profile has revealed that some areas of stellar disks are cooler than others.
The results of such studies is tentative evidence for granules, comparable in size to solar granules, for the G2 V star Alpha Centauri and the F5 V star Procyon. In addition, evidence has been found for bright patches comparable to solar faculae and dark spots possibly comparable to sunspots. There is still much to do both to confirm this evidence and to extend the studies to many other stars. However it seems that, as we fully expected, the Sun is not unique in possessing fine structure in its photosphere.
Among stars, only the Sun is close enough for astronomers to observe activity on its surface directly. The technique used to detect stellar activity cycles is to determine the fraction of a star's observable disk covered by chromospheric plages from the residual radiation in the H and K lines of singly ionized calcium coming from chromospheric plages. For example, when the Sun is more active, a greater fraction of its face is covered by plages (about 20 percent at maximum activity) than when it is less active (0 percent at minimum activity).
For example, spectroscopic evidence for chromospheres has been found in all main-sequence spectral classes from F to M as anticipated. Giant stars in this same spectral class range also display spectroscopic characteristics indicative of chromospheres. Among supergiants, however, chromospheres appear to be present primarily in those of spectral class G and later. For stars of earlier spectral classes (O, B, and A), no such spectroscopic evidence for chromospheres has been found as was forecast.
The evidence for stellar coronae comes from X-ray emissions detected by the Einstein Orbiting Observatory. The great penetrating power of X-rays makes possible the observation of X-ray objects through the interstellar haze to be discussed in the next chapter. But the Earth's atmosphere is also not transparent to X-rays, so that X-ray observations must be made from satellites moving well above the atmosphere.
The rich variety of known X-ray sources, numbering over 3000, was unanticipated, and more sources are constantly being discovered. To our human eyes our Galaxy is a galaxy of stars, whereas to our radio "eyes" it is a gaseous and diffuse looking object as we will discuss in Chapter 20. In general, discrete radio sources are exotic objects or energetic events, certainly not ordinary stars. The production of X-ray photons requires hundreds to thousands of times more energy per photon than does that of visible photons. Therefore, like discrete radio sources, it was thought that the vast majority of X-ray sources should be exotic objects, such as black holes (discussed in Chapter 19), or energetic events, such as interstellar shock waves plowing into gigantic clouds of gas and dust. Indeed these objects are seen, but the vast majority of observed X-ray sources are stars. Such X-ray stars not only occur along the main sequence but also include much of the giant branch and many blue and red supergiants. For example, about 30 stars in the Hyades open cluster (Section 14.4) are emitting X-rays. Other examples are the red supergiants Betelgeuse (M1 I) and Antares (M2 I), the bright giant Canopus (F0 II), and the three main-sequence A stars Sirius, Vega, and Altair. From spectral class O to A, X-ray emission seems to decrease, while there is no recognizable behavior for spectral classes G to M.
In these X-ray observations, there is dramatic evidence for hot coronae in early-type stars as well as in spectral classes F through M. For example, Figure 15.18 shows X-ray images of the nearest star system, Alpha Centauri, in which the X-ray emission is quite evident.
[Figure 15.18]
Measurements of stellar activity have been made for about 400 main-sequence stars within 80 ly of the Sun. In the study, 91 stars were followed, some as long as 14 years, to determine whether or not they exhibit a cycle of activity similar to the 11-year solar cycle. A sizable fraction of the stars studied did indeed exhibit cyclic behavior analogous to the Sun's. The time periods, where fully measured, vary from about 7 years to almost twice that period. Hence the Sun does not appear to be unusual with its 11-year cycle of activity. Although these results will be added to in the future, they suggest that subsurface convection, differential rotation, and magnetic fields occur in main-sequence stars of spectral classes F through M, and the consequent surface activity they spawn seems to divide the stars studied into two almost distinct groups (Figure 15.18). As we will discuss later, stars are not all the same age; some stars are currently being born, while other stars are dying. The two groups of stars defined by the level of their surface activity seem to be different age groups. And the conclusion from our studies to date is that surface activity declines as stars age.
[Figure 15.19]
The Sun is losing part of its mass through the action of the solar wind. As point out above, the solar wind is just another aspect of surface activity, so that the loss of mass by stars might also vary with their ages. We might conclude that for late-type main-sequence stars, mass loss by stellar winds also declines in importance during a star's life just as the other forms of surface activity seem to decline. The loss of mass over a stars live not only should have consequences for the star but also for the interstellar medium into which this matter is flowing. Let us now move on to look at the nature of the interstellar medium.