SHAKE, RATTLE, AND SHINE
New Methods of Probing the Sun's Interior
by Marcia Bartusiak
Some 5 billion years ago a dense clump of interstellar gas, possibly nudged by a shock wave coursing through space, began to condense and slowly spin. Over time a star was born, a relatively ordinary yellow dwarf situated in the outer fringes of the Milky Way galaxy. Ancient astronomers, who looked up at this star several aeons later from one of its attendant planets, called it the sun, and they were both awed and fascinated by its mysterious brilliance. To these early celestial observers the sun was perfect, an unblemished orb of fire. And despite reports that dark splotches occasionally appeared on the face of the sun, this ancient conception of a flawless solar globe held firm even into the Middle Ages.
This illusion, of course, was shattered in the early seventeenth century when Galileo in Italy, as well as observers in Holland, Germany, and England, pointed a newfangled instrument, called a telescope, at the sun and confirmed that the solar surface was indeed spotted. "For the most part they are of irregular shape, and their shapes
continually change, some quickly and violently, others more slowly and moderately," wrote Galileo of his sighting.
Since then, the science of solar astronomy has largely evolved as an extension of Galileo's first effort. What is known about the sun essentially comes from examination of its outer features, although modern-day instruments, both on the ground and in space, have revealed a solar surface more turbulent and varied than seventeenth-century astronomers could ever have imagined: High-speed streams of solar particles emanate from dark coronal "holes"; solar prominences, immense arches of glowing gas, soar for hundreds of thousands of kilometers above the solar surface; and solar flares, lightning-like cataclysmic explosions, can flash across a region of the sun in a matter of minutes.
Nearly all these effects reflect complicated and tumultuous activities inside the sun itself. But an exact description of what lies beneath the sun's fiery surface has been based more on conjecture than explicit measurement. The sun may be the star closest to Earth, yet at the same time, it is very remote. Its center lies some 700,000 kilometers from its surface, and all that fiery hot material in between acts as an effective shield, keeping solar astronomers from directly viewing the sun's interior. The laws of physics, however, do enable scientists to make some educated guesses on what they would see. Theoretical modeling and computer simulations have established that the sun is powered at its core, the inner 20 percent, by the thermonuclear conversion of hydrogen into helium. The resulting energy slowly makes its way out of the core, first by radiative diffusion, and then, starting about seven-tenths of the way out, by convection as the heated gases physically flow upward within the sun's outer layers. The gases subsequently release their energy at the surface, bursting through like bubbles in a pot of boiling fudge, only to recirculate downward to be heated once again. In this way a regular pattern of convection cells—columns of hot gas rising, cooling off, and then descending—is created within the sun (see Figure 1.1).
Many elements in this description, however, are far from secure. Certainty can arrive only if the sun is probed directly. Given the very nature of the sun—its 1.5-million-kilometer width and scorching temperatures—such an endeavor always seemed like an impossible dream—but no longer. In recent decades, solar astronomers have noticed that the sun quivers and shakes. It continually rings, in fact, like a well-hit gong. These reverberations, which carry information about the sun's deep
interior, are allowing solar observers to begin to examine the sun's hidden layers, much the way seismic tremors rumbling through our planet permit geophysicists to scan the earth's interior. Appropriately enough, the name of this new field is helioseismology.
''Two decades ago few people would have believed that it would have become possible to make measurements of conditions inside stars," says Cambridge University's Douglas Gough, a pioneering theorist in this up-and-coming field. "Yet the advent and the rapid development of helioseismology [have] provided accurate probes that have made seeing inside at least one star a reality."
Although still in its infancy, helioseismology has already challenged and revised several long-held conceptions of the solar interior, such as the depth of the convection zone and the way in which the inner sun rotates. Astronomers expect additional revisions as an international helioseismological network, presently in the process of being established, attempts to measure the solar quivers more accurately than ever before.
The information that is gleaned may affect far more than solar models. Knowledge of the sun's inner composition affects calculations of the age of the universe, as well as the amount of helium forged in the Big Bang. Moreover, knowing exactly how the sun spins internally is important in testing Einstein's theory of general relativity, which is the anchor for most of modern cosmology. The domain of helioseismology is broad and far reaching. Yet like so many other significant findings in astronomy, discovery of the quivering sun was totally unexpected.
THE MUSIC OF THE SUN
In 1960, using the 60-foot tower solar telescope atop Mount Wilson in southern California, Robert Leighton of the California Institute of Technology, together with Robert Noyes (now at the Harvard-Smithsonian Center for Astrophysics) and George Simon (now at the National Solar Observatory in Tucson), set out to measure changes in certain absorption lines in the solar spectrum. The lines were observed to Doppler shift, to move to higher or lower frequencies, as gases at the surface of the sun moved either toward or away from the observers. By measuring this shift the observing team hoped to discern the bobbing motions of individual solar granules, the cells of upwelling and sinking gases that cover the solar surface. A spectral-line feature would move toward the blue end of the spectrum whenever a cell heaved upward; the line would shift toward the red end as the cell dove back down. To the surprise of the Caltech astronomers, these velocity patterns were not chaotic, which is what they had expected, but instead were fairly oscillatory. Like a churning sea, the entire surface of the sun was found to be awash with periodic waves, not discernible to the naked eye, each rising and falling with a period of about 5 minutes. Moving at a speed of nearly 2000 kilometers per hour, any one patch can rise and then fall more than 70 kilometers over the 5-minute cycle.
For a while these pulsations, which continually grow and die away at any given site, were thought to be merely a local phenomenon, possibly eruptions from the roiling convection zone just beneath. However, that assumption began to change in 1970 when Roger Ulrich at the University of California in Los Angeles, and, independently, John Leibacher and Robert Stein, who is now with Michigan State University, provided a more global interpretation. Leibacher, now the director of the National Solar Observatory, says that his insight was due to a bit of theoretical serendipity, a stroke of good fortune that occurred as he was trying to simulate the 5-minute solar oscillation on a computer. "Hard as
I tried, I couldn't get my model to yield the answer that I wanted," he recalls. "Another mode kept overwhelming it. I tried and tried to get rid of what I thought was an error, but nature, or in this case the computer's simulation of nature, would not yield. It gave us the right answer."
Ulrich, Leibacher, and Stein came to realize that the 5-minute oscillation was not a local effect but rather the superposition of millions of acoustic or sound vibrations ringing throughout the sun. Since the sun is a spherical cavity with set dimensions, only particular wavelengths can be trapped inside and resonate, much the way an organ pipe resonates at specific frequencies. At any given spot on the sun, the 5-minute oscillation thus grows and decays as these myriad modes, each with its own period, velocity, and strength, move in and out of phase. It is as if the sun were a symphony orchestra, with all the instruments being raucously played at the same time. It sounds like a cacophony, but when the noise is properly analyzed the separate instrumental tones emerge (although the sounds are far below audible frequencies). All these vibrations combine at times to produce a net oscillation on the solar surface that is thousands of times stronger than any one vibration.
The physics of these solar acoustic waves was already well understood from studies of the earth's atmosphere. Such waves propagate at the speed of sound by means of alternating compression and rarefaction of the solar gas, with pressure as the restoring force. Hence, these waves are also known as p modes. If you squeeze the sun, it rebounds under pressure. The individual p modes have periods ranging from a few minutes to nearly an hour.
Helioseismologists believe that the sun should also exhibit gravity waves, or g modes, just as seen in the earth's atmosphere. In this mode the solar material would be oscillated by the pull of gases of different density upon one another. Here, buoyancy is the principal restoring force. For example, if a mass of gas is displaced downward, entering a denser medium, it is buoyed upward. Once this rising mass becomes heavier than its surrounding medium, though, it falls downward once again, ready to repeat the cycle. Primarily originating in the sun's central regions, these longer-period waves (40 minutes or longer) do not propagate very well through the convection zone and are therefore expected to have extremely small amplitudes at the surface. The unstable convection zone does not support buoyancy oscillations very well. So far, reported sightings of g modes in the sun have not been confirmed. There are also modes, called f modes, that exist at the surface of the sun and travel about much like waves on the ocean. Such waves are essentially surface gravity waves and are virtually compressionless.
Since the sun is three dimensional, each solar acoustic wave is a bit more complicated than a simple wave resonating in an organ pipe. What is known as the "degree" of the wave, a parameter conventionally labeled l by helioseismologists, can loosely be thought of as the total number of horizontal wavelengths that encircle the sun's surface. These wavelengths range from the width of an individual solar granule—a few thousand kilometers, resulting in a high-degree number—to the entire solar circumference, a number approaching unity. Each degree, in turn, can have varied frequencies and overtones, which reflect the variety of resonances possible in the other directions as well. Being three dimensional, the nodes of these standing waves—the regions where nothing moves—are not points but either concentric spheres or planes that slice through the sun parallel and perpendicular to each other. The quantity l, in fact, is actually the total number of nodal planes, both parallel and perpendicular to the solar equator, belonging to each oscillation.
Why should solar acoustic waves exist at all? That is not known with certainty, but helioseismologists have their suspicions. "Something is driving these modes," says solar astronomer Ken Libbrecht of Caltech. "We think it is due to the sun's turbulent convection. The turbulence in the surface layers generates acoustic noise, and if you generate noise within a cavity, then you excite the normal modes of that cavity." What is better understood is the method by which these waves travel around the sun.
Imagine a sound wave diving into the depths of the sun. Both temperature and density increase as the wave, traveling at around 200 kilometers per second, penetrates deeper and deeper, and this causes the wave to refract, or bend, as it travels inward. "The surface of the sun is cold, some 5800 degrees. The center is hot, around 15 million degrees. So, the sound speed actually increases as you go down into the sun, because sound speed increases with temperature," explains Gough. "A wave propagating downward into the sun thus experiences a faster speed deeper in than it does near the surface. As a result, it gets refracted." Eventually, the wave turns completely around and heads back up to the surface, where, because of the sharp drop in density at that boundary, it is reflected downward once again. In this way, the acoustic wave can travel around the sun many times, establishing a standing-wave pattern that lasts for days or weeks.
Acoustic modes above a certain frequency (roughly 5.5 millihertz) cannot be reflected by the sun's photosphere. These higher frequency waves simply move into the sun's chromosphere and dissipate their energy. Thus, there exists a finite number of p modes that can be
trapped inside the sun, about 10 million. Not all are detectable, but a good fraction of these modes are excited to observable amplitudes.
In general, the longer the horizontal length of the wave (in other words, the lower its degree), the deeper its plunge into the solar interior. This is because acoustic waves with longer wavelengths are refracted more gradually and so propagate more steeply into the sun. Shorter wavelengths, on the other hand, stay near the surface. Thus, by studying a wide range of modes, from high to low-degree, solar physicists can effectively "enter" the sun in a step like fashion, peeling away each of the star's layers as if it were an onion (see Figure 1.2). And with a wave's propagation dependent on the temperature, velocity, and density of the medium through which it is traveling, each mode offers valuable clues on the makeup and structure of the solar interior, much the way the resonant tone of a musical instrument provides hints as to the instrument's design—for instance, whether it's shaped like a flute or clarinet. Such effects afford solar astronomers with enormous diagnostic capabilities. For example, waves traveling in the same direction as the sun's fluid material will move a bit faster relative to a fixed observer, shifting their apparent frequency upward. Conversely, waves traveling against the flow will decrease in apparent frequency. Analysis of these splits in frequencies therefore offers a means of mapping the sun's large-scale internal motions.
In 1970, when Ulrich, Leibacher, and Stein first formulated their idea of acoustic modes, it was only one of many possible explanations of the sun's 5-minute oscillation. The notion didn't take firm hold until 5 years later. At that time, in 1975, Franz-Ludwig Deubner of West Germany was able to separate his observations of the 5-minute oscillation into neat differentiated modes. A power spectrum
of his Doppler velocity data, recorded for many hours along a strip of the solar equator, took the form of narrow and strikingly regular bands, a representation of the allowed frequency values for the resonant modes. The sum of these values was the 5-minute oscillation. A theory had at last been transformed into a valuable new tool for solar astronomy (see Figure 1.3).
Over the years, solar observers have refined their techniques for detecting both high- and low-degree modes. For the lowest modes, waves with values from zero to three, whose lengths are comparable to the size of the sun, observers look at the collective Doppler shift of a spectral line averaged over all or much of the solar disk. Since day/night gaps introduce spurious signals that make analysis difficult, investigators at the University of Birmingham in Great Britain and the Observatory of Nice in France established field stations around the globe to obtain an uninterrupted record of the sun's activity. Researchers have also made (and continue to make) long-term observations at the South Pole, where the sun never sets during austral summers.
Modes with degrees in the tens to hundreds, however, do not show up in such globe-spanning Doppler shift data. The wavelengths of these modes are relatively small compared to the size of the sun and so are averaged out. These higher-degree modes are effectively discerned in spectrograms that register very localized Doppler motions across the face of the sun, much like the first Caltech measurements. Ground-based instruments can observe parcels as small as 100 kilometers across, about the width of the state of Texas. When processed, pictures of these parcels, known as velocity images, look like salt and pepper strewn over the solar disk (see Figure 1.4). The dark areas depict the regions on the solar surface that are sinking; the bright spots, conversely, are rising or moving toward the earth. It is these oscillations that are differentiated
into the high-degree components. This method has enabled observers to differentiate modes with degrees up to a few thousand, although atmospheric distortion does play havoc with degrees that measure above 400. "The surface of the sun is seething with movement. There's a background noise of about a kilometer per second," points out Libbrecht. "But, on top of that, there are oscillations as little as a millimeter per second, which we can see. It's quite remarkable."
REWRITING THE TEXTBOOKS
Analysis of these data can be handled in one of two ways. Traditionally, researchers have constructed a set of solar models and then adjusted certain parameters, such as the temperature and density of various solar elements, until they best fit the p modes observed ringing through the sun. More recently, however, theorists have been developing mathematical techniques known collectively as inversion, which extract the solar parameters directly from the modes themselves. Inversion, in essence, converts the solar "tones" into a map of its essential features. This second approach is far more challenging than the first. Supercomputers are often used to deal with tens of thousands of modes at one time.
Interpretation of the p modes began soon after they were discovered. Deubner, for example, reported that the modal frequencies he had uncovered were actually lower than theoretical predictions. These observations led Gough in 1975 to deduce that the sun's convection zone must be deeper than previously estimated. Additional observations prompted Edward Rhodes (now at the University of Southern California), Roger Ulrich, and George Simon to draw the same conclusion. A deeper convection zone could account for the unexpected signal. It had long been assumed that the convection zone's depth was 20 to 25 percent of the solar radius. Calculations by Rhodes and his group determined that it was more like 30 percent. It was the first major solar parameter to be adjusted based on helioseismological data. The convection zone's greater depth means that convection can transport heat from the bowels of the sun more efficiently than once thought.
By profiling the speed of sound throughout the sun, helioseismologists have also been able to profile its composition. As pointed out earlier, the speed of sound generally increases at greater and greater depths as temperatures increase, but that trend changes when the wave approaches the very center of the sun. At that point the speed of the wave actually dips back down a bit, because thermonuclear reactions in
the core have converted about half of the original hydrogen into heavier helium. The speed of sound decreases, gets more sluggish, in denser materials (see Figure 1.5).
By carrying out such diagnostics of the sun's interior, helioseismologists turned the sun into a physics laboratory. One of their experiments, in fact, caught an error in some previous calculations of the sun's properties. In completing their profile of the speed of sound through the sun, helioseismologists discovered a discrepancy. The acoustic observations suggested that the solar sound speed was greater in the sun's mid-regions (roughly halfway in) than theoretical models had expected. Since this increase (nearly 1 percent) was difficult to explain, the helioseismologists began to wonder if solar physicists were underestimating the sun's opacity in that region (opacity being a crucial parameter in the helioseismological calculations). Gough and several colleagues predicted the opacity might differ by as much as 10 to 15 percent. "This suggestion," says Gough, "motivated a group at the Lawrence Livermore National Laboratory to look into the matter with care. And indeed, our prediction was confirmed." The opacity was recomputed, and, with the new parameter plugged in, the speed-of-sound discrepancy disappeared. Theory and observation came into line.
A view of the sun that has changed most dramatically since the birth of helioseismology is the overall profile of the sun's internal rotation. It turned out to differ considerably from the picture previously generated by theorists in their computer simulations. It has long been known, from observations of sunspot movements, that the sun's rate of rotation steadily declines from the solar equator to the poles. The poles complete a circuit in about 36 days, the equator in just 25. (Being a ball of gas, the sun is not constrained to rotate like a rigid body.) "This is poorly understood," notes Libbrecht, "but possibly linked to both convection processes and coriolis forces." Numerical simulations of this process had
led to a model of the sun's differential rotation that was commonly referred to as "constancy on cylinders." The sun in this picture, at least through the convection zone, was supposedly composed of a set of nested cylinders that extended from pole to pole, aligned with the sun's axis of rotation. The inner cylinders, which surfaced at the higher latitudes, rotated more slowly than the outer ones, which met the surface at the more rapidly rotating lower latitudes. This also meant that the angular velocity at a particular latitude should have gradually decreased with depth.
But this picture failed to fit the observations of a number of helioseismologists, including Timothy Brown at the National Center for Atmospheric Research in Colorado and Cherilynn Morrow, then a student at the University of Colorado. Morrow and Brown began to show that the sun's rotation rate at a given latitude actually remains fairly constant down through the convection zone. Past that zone, angular velocities at the poles and equator shift toward the same rate. Halfway into the sun, beyond the convection zone and into the radiative interior, the sun rotates somewhat like a rigid body. These observations confirm the suspicion that the sun's differential rotation at the surface, long a mystery, is somehow generated by convection rather than processes deeper in the interior.
Brown and Morrow's model was sustained and extended by a wealth of new data gathered by Ken Libbrecht. For 6 months in 1986 at Caltech's Big Bear Solar Observatory, located in the center of Southern California's Big Bear Lake, Libbrecht and his students took a Doppler image of the sun each minute, gathering a total of around 70,000 pictures. The team then extracted vibrational modes from these images after some 40 hours of supercomputer time. "We were interested in measuring as many modes as we could," says Libbrecht, "because each mode has its own story to tell about the medium in which it was trapped." Lastly, inversions of these modes, performed by Jorgen Christensen-Dalsgaard of Denmark's Aarhus University and others, mapped the sun's rotation down to a depth of about 450,000 kilometers, 60 percent of the way to the sun's center. Similar sets of images were taken again in 1988, 1989, and 1990 and added to the data base.
"We find that the rotation persists almost independent of radius, down to the base of the convection zone," says Libbrecht. "Then, there is a fairly sharp transition to solid body rotation. This is one of the biggest outstanding questions concerning the sun—why does it rotate in this manner? We couldn't go down to the very core—we went down about
six-tenths of the way—but, dynamically, it seems to make sense that the whole interior is rotating uniformly with about a 27-day period."
These and other measurements, particularly data taken earlier at the South Pole by Thomas Duvall of the NASA Goddard Space Flight Center, John Harvey of the National Solar Observatory, and Martin Pomerantz of the Bartol Research Institute, have served as a valuable check on the theory of general relativity, Einstein's revolutionary view of gravity. A competing theory of general relativity, introduced in the 1960s, had suggested that Einstein's calculation of a general relativistic effect that perturbs the orbit of the planet Mercury might be wrong. Supporters of this alternate theory of general relativity argued that a large portion of the inner sun was spinning much faster than the solar surface, causing the sun's core to flatten. If the solar interior is rotating very fast, Einstein's explanation for Mercury's peculiar orbital behavior would be in jeopardy, since his calculations assumed a fairly spherical sun. "The sun was rotating faster in the past than it is now. The sun is slowing down," points out Gough. "But has the inside of the sun slowed down as well? The discussion hinges on how strongly the inside is coupled to the outside." While helioseismologists are not yet able to make an exact measurement ("We can't totally rule out a fast-spinning nugget at the solar center," notes Libbrecht), the data do strongly suggest that the innermost core, a few percent of the sun's total volume, is not rotating fast enough to squish the sun and disrupt Einstein's theory.
Also affected by the changing profile of the inner sun has been astronomers' understanding of the solar dynamo, the "engine" that drives the ebb and flow of activity over a solar cycle by inducing immense electrical currents and magnetic fields. A decade or so ago, astronomers thought that the dynamo resided in and was driven by the turbulent convection zone as a whole. However, now that angular velocities are seen to remain fairly constant through the convective regions, that idea is now ruled out. In its place, theorists are suggesting that the dynamo occupies a more narrow zone between the bottom of the convection layer and the top of the deep interior, the region of transition where rotation rates change most sharply.
THE CHICKEN OR THE EGG?
Interpretation of helioseismological data can, in some ways, be likened to that old conundrum, "Which came first, the chicken or the egg?" Theorists turn to current models of the sun to differentiate and analyze the various acoustic modes; the modes, in turn, help refine the
standard model of the sun. It's a tricky business, as a number of uncertainties are incorporated in the standard model of the sun. As noted earlier, estimating the sun's opacity, a feat that requires massive computing, can generate uncertainties of several percent or more, and this affects how modal frequencies are calculated from the Doppler information. Convection near the surface, the structure of the sun's atmosphere, nuclear reaction rates, and the resultant element abundances all affect helioseismologists' interpretations of the oscillations. Theorists and observers thus work hand in hand, each responding to the findings of the other in hope of converging on the correct model of the sun.
Helioseismologists are setting an ambitious agenda for themselves. A top priority is understanding the solar cycle, that 11-year period over which sunspot counts and solar flares wax and wane and solar magnetic field strengths build up and decline. Solar astronomers already suspect that the sun's magnetic fields interact with convection, indeed at times might suppress it, but helioseismology is needed to see such an effect below the solar surface. Helioseismologists have already noticed that solar oscillations behave differently in the presence of strong magnetic fields. The stronger the field, the stronger the effect, which can be either an increase or decrease in the frequency of a mode. Moreover, investigators have seen oscillations around single sunspots exhibit an intriguing phenomenon: up to 100 percent more vibrational power appears to move into a spot than moves out of it, as if the sunspot is absorbing p modes.
Such effects should provide observers with valuable tools in studying the sun's mysterious magnetic interior. For instance, how far down into the solar interior do flux tubes—threads of hot, highly magnetized gases first seen just 20 years ago—extend into the solar interior? Is the sun tunneled with these tiny sunspot like features? And what is the vertical structure of a sunspot? Does its magnetic field branch like the roots of a bush, or does the field remain bound as one massive trunk that extends more deeply?
Since the propagation of acoustic modes is dependent on temperature and density, helioseismologists are also using the modes to follow the sun's temperature changes. After dissecting modes for nearly a decade, observers have now seen temperature gradients shift over the length of a solar cycle. Having analyzed this effect with helioseismic observations extending back to 1980, Jeffrey Kuhn of Michigan State University suspects it to be a reflection of changes in large-scale magnetic fields as the sun's activity waxes and wanes every 11 years.
Modal frequencies, too, appear to change synchronously with the
solar cycle, a modulation that may also be slowly driven by varying magnetic fields. In analyzing his four large data sets, taken over the last half of the 1980s, Libbrecht saw the modal frequencies shift by one part in 10,000, due to changing magnetic activity on the sun's surface as the solar cycle progressed. Moreover, the shape of the sun's acoustic cavity itself appears to alter subtly over a full cycle. All of these observations suggest that the solar cycle may have deeper roots than previously suspected. Some even speculate, although it's very controversial, that the very core of the sun, where thermonuclear reactions take place, may somehow participate in the solar cycle.
The potential uses of helioseismology are legion. Some observers hope to study solar surface features, such as supergranules, and follow their movements over days and weeks. Others are hunting for large-scale convective flows. And helioseismologists are not stopping at the sun. Already a few adventurous observers have looked for seismic quivers in other stars, which offers astronomy the chance to plumb other stellar interiors. Analogs to the 5-minute oscillations of our sun have been reported in such stars as Alpha Centauri A, Procyon, and Epsilon Eridani. For the moment, however, the sun remains the top priority for specialists in this pioneering discipline.
Solar astronomers have long assumed that the sun's convection zone is lined with giant cells that act as monstrous conveyor belts, transporting solar material to and from the sun's fiery interior. But no hint of these massive structures has yet been found on the surface. To discern such a detailed solar topography, helioseismologists need long uninterrupted views of the sun, especially with instruments that resolve the high-degree modes. (The worldwide networks set up near Birmingham and Nice do not image the sun and so distinguish only low-degree modes.) With that need for high-degree data in mind, helioseismologists established the Global Oscillation Network Group, or GONG for short, a reminder of the acoustic qualities of the solar tremors. (As an additional reminder, a small gong is struck at the start of each annual GONG meeting.) Initiated by the U.S. National Solar Observatory in 1984, this international project now involves more than 100 observers and theorists from 61 institutions in 16 countries.
Principally funded by the National Science Foundation, the project is setting up six helioseismological field stations—identical and highly sensitive Doppler imaging instruments—around the globe at roughly
equal spacing. ''We will be able to increase our resolution 10 times over by observing, not at one site for 6 months, but at several sites around the world," points out Libbrecht. "Going beyond a basic understanding of the structure of the sun, these new helioseismological measurements are expected to turn the sun into a precision laboratory for learning about the physics of high-temperature plasmas and magnetohydrodynamics, neutrino oscillations, radiative transfer, and the dynamics of large-scale stratified convection and rotation."
The GONG instruments are being placed in California, Chile, the Canary Islands, India, Australia, and Hawaii. Contrary to many astronomical sites, a lack of air turbulence, or good "seeing," is not a prime consideration, considering the resolutions at which they will be working. Each station will be housed in a refurbished commercial cargo container and automated to the fullest extent possible, as if it were a spacecraft on the ground. "Mission Control," in this case, will be the National Solar Observatory in Tucson, where the first station was erected for testing. Much like the system already established for global radio-telescope arrays, data taken at each station will be recorded on videocassette tapes, which will be mailed periodically to the central data analysis center in Tucson. With each station recording some 2 gigabytes a week, 1.5 trillion bytes of data will be acquired after 3 years of continuous sun watching, the project's planned lifetime (although there are hopes to extend that to 5 years). Such prolific data gathering will make it one of the largest data sets in astronomy, after the Hubble Space Telescope. With special computer software, users will be able to browse, view, and select the data sets they wish to interpret within the archives.
Helioseismologists utilize a number of techniques to take high-resolution Doppler measurements. GONG scientists have chosen Fourier tachometry, because of its ability to measure Doppler velocities quickly and accurately. A given mode, one of many that make up a solar oscillation, moves some 10 centimeters per second or less. To obtain a good signal above the noise, a GONG instrument should then detect movements over the sun with a precision as small as a centimeter per second.
Timothy Brown of NCAR pioneered Fourier tachometry, and the GONG instrument is a product of the evolution of that technology. Sunlight first enters through an 8-centimeter telescope, or light-feed, that automatically tracks the sun over the course of a day. This light passes through a filter that isolates a specific spectral line, in this case the unionized nickel absorption line at 6768 angstroms. These red rays then enter the heart of the Fourier tachometer, a compact Michelson interferometer
in the form of a cube 2.5 centimeters on a side. As with many interferometers, this cube splits the incoming light into two parts, routes each beam along a different pathway, and in the end recombines the two beams. If the light waves are in phase, or in step (peak matching peak), the two beams will add up to a bright signal; if out of phase (peak matching trough), the two will combine to produce a dark image.
The Fourier tachometer's keen ability to detect small Doppler shifts on the solar surface results from the makeup of the cube: one pathway or arm is solid glass; the other is air. With such different pathways (the glass arm is 30,000 wavelengths longer than the air arm), a tiny change in the wavelength of light entering the interferometer results in a measurable change of phase when the two beams are recombined. These phase shifts thus indicate the way in which motions on the sun are increasing or decreasing the original 6768-angstrom wavelength.
The output image of the interferometer, which encompasses the entire solar disk, will ultimately be focused on an electronic detector with an array of more than 65,000 pixels. Each pixel will record the signal intensity and phase at that point on the sun. At the field stations, a complete image will be recorded once a minute. In this way GONG scientists should detect any solar oscillations with durations of 3 minutes or more.
Ultimately, atmospheric fluctuations prevent observers from studying the highest-degree modes, which are tinier and require very good resolution to resolve. Therefore, the field of helioseismology will soon take to space, free from turbulent air, as well as disruptive day/night gaps. Helioseismic instruments will be included on SOHO, the Solar and Heliospheric Observatory, a joint project of the European Space Agency and the National Aeronautics and Space Administration. Scheduled for launch later this decade, SOHO will be placed at a Lagrangian point 1.5 million kilometers sunward from Earth. There a variety of detectors will be trained on the sun for at least 2 years, although the observatory could operate for 6 years, a major slice of the solar cycle. SOHO scientists are developing an instrument called the Michelson Doppler Imager, or MDI, which will perform both long-term Doppler scans, up to 2 months of continuous coverage at one image per minute, and daily readings. In this way SOHO investigators hope to measure solar vibrational modes with degrees from 1 to 3000. With such long-term high resolution, MDI's data could enable helioseismologists to focus on the topography of active regions, zoom in on granulation, and track the movement of sunspots.
Other detectors aboard SOHO will concentrate on low-degree modes
and possibly g modes, whose extremely tiny amplitudes on the sun's surface make them the most elusive of the sun's oscillations. Years of data collection, either in space or by GONG, may be needed to discern the faint g-mode signal above the noise. But the payoff will be gig if g modes are firmly discovered, for the detection will enable solar astronomers to peer directly into the sun's core.
"The very core of the sun where nuclear reactions take place, that's the biggest prize in helioseismology," declares Gough. At the sun's core, helioseismologists will find a laboratory that is irreproducible on this planet, where matter is pulled apart, ionized, and fused at unearthly temperatures. Indeed, one of astronomy's most nagging mysteries lies at the heart of the sun, where a flood of ghostly particles called neutrinos are continually and copiously generated and released into space. Sensitive detectors on Earth, located deep underground, are capturing only one-third to one-half of the neutrinos predicted by physicists' standard model of the sun. New findings in particle physics may eventually explain this discrepancy. Perhaps the neutrino, now assumed to be massless, has a bit of mass after all, affecting its detection on Earth. Or maybe other undiscovered particles huddling in the sun's core somehow temper the solar nuclear fire. If not, solar physicists will be forced to amend their ideas on stellar structure. If answers are not forthcoming from helioseismology, they will surely arrive as investigators advance the art of neutrino astronomy, another very powerful means of probing the sun's interior.
THE ELUSIVE NEUTRINO
The neutrino was one of the first elementary particles to be postulated before it was observed. For Wolfgang Pauli, who originated the idea, it was a desperate remedy. During the 1920s, particle physicists had begun to notice a troubling anomaly. Whenever a radioactive nucleus decayed by ejecting an electron, a process called beta decay, something went awry. In carefully controlled experiments conducted at Great Britain's famous Cavendish Laboratory, investigators saw that the energy of the nucleus before it radioactively decayed was more than the energy of the system afterward (i.e., the combined energy of the depleted nucleus and the fleeing electron). It looked as if energy were actually disappearing during beta decay, and this violated one of the bulwarks of physics—the law of conservation of energy, which states that energy is neither created nor destroyed. The laws of physics governing the conservation of linear and angular momentum were being violated as
well. Atoms, it seemed, weren't playing by those rules. Energy and momentum were somehow getting lost along the way. So distressing was this finding that the distinguished physicist Niels Bohr openly discussed the possibility that atoms might at times violate some of the standard laws of physics.
Pauli wasn't as pessimistic as his Danish colleague. The Viennese physicist had an abiding faith that atoms were obeying the physical laws of the land. But to maintain that allegiance, Pauli took the rather radical step in 1930 of suggesting that an entirely new particle, invisible to ordinary instruments, had to exist to explain the discrepancies seen in the Cavendish experiments. Every time a nucleus undergoes beta decay, he proposed to some colleagues by letter, this neutral phantomlike particle is emitted and vanishes off into the night carrying off that extra bit of energy and momentum. Usually full of chutzpah, the young Pauli was actually intimidated by the outrageousness of this idea. "Dear radioactive ladies and gentlemen …," he teasingly wrote his friends, who were then attending a meeting in Tübingen, Germany. "For the time being I dare not publish anything about this idea and address myself confidentially first to you, dear radioactive ones, with the question of how it would be with the experimental proof of such a particle." It was only after the chargeless neutron was discovered in 1932 that Pauli finally got the courage to publish his unusual suggestion. The noted physicist Enrico Fermi soon dubbed Pauli's hypothetical mote the neutrino, Italian for "little neutral one." The name was apt. The neutrino seemingly had no mass, and it had no charge. Indeed, it was nothing more than a spot of energy that flew from a radioactive atom at the speed of light.
Fermi perceptively recognized that Pauli's idea would also require a whole new force, what came to be known as the weak nuclear force. It is this force that enables a neutron to convert into a proton, releasing the electron and the neutrino (actually an antineutrino) seen in beta decay. (Interestingly enough, the prestigious journal Nature rejected Fermi's idea on the weak force as too speculative and too remote to be of any interest to practicing scientists; a small Italian review, however, did publish the hypothesis. Thus, the weak force entered the world of physics with little fanfare.)
It took so long to prove that the neutrino was more than a figment of Pauli's imagination that some physicists began to call his neutrino "the little one who was not there." In fact, Pauli began to wonder whether he had committed the theorist's ultimate sin: postulating a particle that could not possibly be detected. He had good reason to be fearful.
Capturing a neutrino is an extremely difficult task. The neutrino, which only interacts via the weak force, is terribly oblivious to matter. "A neutrino could actually go through 1020 centimeters of water, on average, before it scatters," explains John Wilkerson of the Los Alamos National Laboratory, a nuclear physicist who came to specialize in neutrino physics. "It's pretty hard to put such a number in terms we can comprehend. It says that the neutrino could travel through a pool of water the diameter of 100,000 of our solar systems put end to end, about a hundred light-years, before it interacted with another particle." A neutrino could bolt right through the entire earth and feel nary a thing, as if the earth were no more substantial than a cloudy mist.
The odds of catching a neutrino are considerably increased, however, if there is a flood of neutrinos coming at you. In that way, a few out of the hordes have a chance of bumping into an atom closeby and triggering a weak interaction. As a result, the atomic nuclei are altered, sending out flashes of radiation that can be spotted with photomultiplier tubes. The construction of nuclear reactors in the 1950s at last provided the necessary neutrino spigot for carrying out this tricky endeavor.1 By 1956 Clyde Cowan and Frederick Reines, in a difficult but elegant experiment conducted at a South Carolina nuclear power plant, were finally able to corner the ghostly neutrino, or at least see its telltale footprints of light. The experiment was appropriately named Project Poltergeist.
Over each second that the giant Savannah River reactor operated, hundreds of trillions of neutrinos were generated in the reactor's core, whereupon they immediately escaped and raced through Cowan and Reines's 10-ton detector set nearby. Only a few of the slippery particles got stopped each hour. With such a sparse harvest, it took more than 3 years for the two young researchers to gather enough evidence to declare the neutrino a certified member of the particle zoo—to Pauli's great relief. Receiving news of the verification while attending a conference in Geneva, Pauli, by then in his 50s, held the telegram high and gleefully announced to his fellow physicists, "The neutrino exists!" He had lived to see his "sin" forgiven.
The particle that Cowan and Reines detected is specifically known as the electron neutrino because of its appearance in nuclear reactions involving ordinary electrons. Since then, physicists have found a second type of neutrino, the muon neutrino, which is associated with interactions involving the muon, a more massive relative of the electron. The muon is about 207 times heavier than the electron. A third neutrino, the tau neutrino, is confidently assumed to exist as well. It would be linked
to a still heavier cousin of the electron called the tau that was first seen in particle accelerator experiments in the mid-1970s. The tau is nearly 3500 times more massive than the electron. To use the jargon of physics, neutrinos come in three "flavors": electron, muon, and tau.
In the standard model of particle physics, the rest mass of each neutrino type is arbitrarily set to zero. But theories that go beyond that standard model suggest that might not be true; neutrinos may have mass. And that could be a reason why solar physicists see fewer neutrinos flying out of the sun than they expect. Why this might be so requires a short review of physicists' understanding of how the sun shines.
THE SOLAR FURNACE
In the 1910s and 1920s, before a full theory of quantum mechanics was developed, physicists could only speculate on the sun's power source. Sir Arthur Eddington, among others, suggested that atoms must somehow be fusing and releasing energy. When Eddington's critics responded that the sun's interior was simply not hot enough to bring about these transmutations, Eddington readily responded that they should "go and find a hotter place."
By 1938 Hans Bethe and Charles Critchfield in the United States and Carl von Weizsäcker in Germany at last independently showed, in a series of steps that elegantly moved from one nuclear reaction to another, exactly how the sun and stars can be powered by the welding of 4 nuclei of hydrogen into helium. Normally, hydrogen nuclei, or protons, strongly repel each other, since they are positively charged. But deep in the core of a star, where temperatures reach up to 15 million degrees and densities are 12 times that of lead, protons can gain enough energy to collide with a force that sometimes overcomes that electromagnetic repulsion. This enables some of the protons, following an involved chain of reactions as outlined by Bethe and von Weizsäcker, to stick together with a powerful nuclear glue called the strong force. In the sun about one in every 10,000 billion billion proton collisions gives rise to a nuclear reaction. Long odds, but the firestorm under way in the sun's heart still enables it to convert approximately half a billion tons of hydrogen into helium with each tick of the clock. In the process, about 4 million tons of mass is transformed each second into pure energy, which eventually bathes our entire solar system in heat and light. The sun has been doing this for nearly 5 billion years and has enough hydrogen fuel in its core to continue for about 5 billion more.
The proton-proton reaction is the first step in a series of nuclear
reactions that lead to helium production in the sun. First, two protons collide; one releases a positron and electron neutrino to become a neutron, which immediately combines with the remaining proton to form deuteron, a nucleus of heavy hydrogen.
This proton-proton reaction is the primary fusion reaction thought to occur in the sun. The deuteron proceeds to latch on to another proton, forming a light isotope of helium, helium-3, and releasing a gamma ray.
Two of these helium nuclei then collide, ejecting two protons and leaving a nucleus of helium-4.
According to calculations by John Bahcall at the Institute for Advanced Study in Princeton, New Jersey, and Roger Ulrich, over 90 percent of the neutrinos emanating from the sun are produced by the proton-proton reaction, the first step in the chain above. As much as 2 percent of the sun's energy is emitted as these electron neutrinos, whose energies span some 400,000 electron volts or less. "Given its weak interaction with matter, a neutrino can escape directly from the sun's core, whereas a photon created at the center can take 10,000 years to escape," notes Wilkerson. Here was the perfect tool for looking into the sun, as physicists estimate that every second about 66 billion of these solar neutrinos rain down on each square centimeter of the earth's surface.
There are additional nuclear reactions going on inside the sun, besides the proton-to-helium chain, and each of these side reactions gives off neutrinos as well—neutrinos with characteristic energies. For instance, particularly energetic neutrinos (some of them with energies of more than 10 million electron volts), are produced in a minor reaction involving nuclei of boron-8 (a rare isotope composed of five protons and three neutrons). Specifically, it is a reaction in which a nucleus of boron-8 decays into beryllium-7, releasing a positron and a neutrino in the process.
Though far fewer in number than the neutrinos generated by the sun's proton-proton reaction (they represent only a small fraction, a mere 0.01 percent, of the total flux of neutrinos coming out of the sun), these boron-reaction neutrinos are easier to detect. In 1946 Bruno Pontecorvo, a student of Fermi's, first suggested how such high-energy neutrinos might be captured. "His idea," says Wilkerson, "was to take an atom of chlorine-37, which is a stable nucleus with 17 protons and 20 neutrons, and add a neutrino to it. That turns the chlorine into argon-37, which is radioactive and can be watched for its decay." More than 20 years went by, though, before physicists could actually embark on such an ambitious endeavor.
The world's first neutrino observatory was finally established in America's heartland in 1967, and it has been gathering vital clues on the quirky nature of the neutrino ever since. The observatory's "telescope" is a huge tank of chlorine-rich cleaning fluid, 100,000 gallons of perchloroethylene, set in the Homestake gold mine situated nearly a mile beneath the Black Hills of South Dakota. Such a depth is required to keep the measurements free from disruptive cosmic rays. Additional shielding was added to eliminate interference from natural sources of radioactivity within the deep chamber. It took 20 railroad tanker cars to fill the tank, roughly the amount of stain remover American consumers use in a day (see Figure 1.6).
With this gargantuan apparatus, University of Pennsylvania radio chemist Raymond Davis, the founding father of neutrino astronomy, has been catching a few electron neutrinos out of the legions that are continually spewed into the solar system as the sun burns its nuclear fuel. For more than two decades now, the chlorine atoms in his cleaning fluid have been occasionally stopping some of the cagy particles. Following Pontecorvo's original scheme, the electron neutrino gives itself away by turning an atom of chlorine into traceable radioactive argon. The argon atoms, suspended within the perchloroethylene, are extracted after 2 months of exposure by bubbling helium gas through the cleaning fluid. Once the gas is collected, it is sent through a cold trap, where the argon freezes out. The extracted argon is then placed in low-background proportional counters, which record any radioactive decays of the argon-37 atoms.
But the results over the years suggest that the chlorine is capturing neutrinos (and subsequently transforming into argon) at an unexpectedly
low rate—only about one neutrino is captured every couple of days.2 This is just about one-third the number of solar electron neutrinos that theorists, such as Bahcall, estimate should be snared, given the rate of nuclear reactions thought to be going on within the sun. Was it a real effect or a bug in the experimental setup?
Not until 1986 did another experiment at last come on line to provide additional data on the solar neutrino flux. It served as a valuable check on whether the Homestake measurements were truly recording a neutrino signal three times lower than the signal predicted by solar models. Japan's Kamiokande detector, a huge vat of water originally built to look for the proton decays anticipated by physicists' grand unified theories, was reconfigured to search for solar neutrinos. The Kamiokande detector is fundamentally different than the radiochemical type used by Davis. Unlike the Homestake detector, the Kamiokande detector is surrounded by more than 900 photomultiplier tubes, which record the distinctive light given off when a neutrino interacts with an electron. The electron, upon being scattered by the solar neutrino, emits a characteristic cone of light, known as Cerenkov radiation, as it moves through the water. This light is then detected by the photomultiplier tubes, yielding data on an event-by-event basis. Within a few years, this neutrino-detection scheme recorded a deficit of solar electron neutrinos, similar to the shortfall seen for two decades in South Dakota. Since it's been running, Kamiokande observes, on average, about one neutrino scattering event every 8 days, or some 45 signals over a year. "The Kamiokande detector has the nice advantage in that it can actually point to the sun," says Wilkerson. "It became the first experiment to convincingly demonstrate that neutrinos were being emitted from the sun."
But both the Homestake and Kamiokande detectors have limitations: they can only register fairly high energy neutrinos, most of them produced by that minor boron reaction in the sun, a nuclear reaction that is very sensitive to the sun's internal temperature. It's quite possible that a slight change in the solar model, turning down the sun's core temperature a bit, for example, can explain the undersupply of high-energy neutrinos. There are several ways such a temperature change could come about; perhaps, the proportions of hydrogen and helium in the sun, the two elements that make up 98 percent of the sun's mass, are different than solar physicists think. "One of the many suggestions that have been made to explain the neutrino deficit is that the sun's helium abundance is very low, leading to a comparatively low central temperature," says Gough. Solar models show that if the abundance of helium in
the sun were as low as 19 percent by mass the sun's internal temperature should register about 13 million degrees Kelvin, causing the production of high-energy neutrinos to be curtailed—enough to explain the current deficit. According to Gough, though, a look at how the speed of sound varies through the sun, a helioseismological calculation based on an examination of the sun's acoustic modes, convincingly rules out such a helium deficit. A helium abundance of around 23 to 25 percent is needed to explain the oscillations detected in the sun's outer layers.
But there are other possibilities for explaining the neutrino deficit, conditions that are open to examination by helioseismological techniques. Gough notes that there is some evidence that the sun's rotation, just beneath the convection zone, varies on the time scale of the solar cycle. Whether the energy-generating core also varies is not yet known. There are theoretical reasons why it might, but observations have not yet been able to detect it. If future measurements do observe such an effect, it could greatly alter theorists' calculations of neutrino production rates. Such are the future questions for heliosesmology.
Other ideas for reducing the sun's temperature are being discussed in the astrophysics community. Some theorists have wondered if there is a new class of weakly interacting massive particles (WIMPs) sitting in the sun's core. Such particles, as yet hypothetical, would have the effect of lowering the core temperature and thus cutting back the number of high-energy neutrinos reaching earth's detectors. Experiments conducted so far seem to rule out a WIMP solution, but physicists won't know for sure until they can measure the lower-energy neutrinos that are generated by the sun's predominant nuclear process, the proton-proton reaction, a process not so sensitive to the sun's inner temperature. If a shortfall is seen in those neutrinos as well, though, physicists suspect that the solution to the neutrino-deficit mystery will lie in particle physics, the nature of the slippery neutrino, rather than in WIMPs or solar physics.
To help decide, more sensitive neutrino observatories are being erected around the world. ''We have now gone from a field that went for 20 years with one lone experiment to many experiments," says Wilkerson. "The decade of the 1990s should prove to be a landmark period for the study of solar neutrino physics. New experiments are poised to discover whether this deficit is a problem in our understanding of how the sun works, is a hint of new neutrino properties beyond those predicted by the standard model of particle physics, or perhaps a combination of both."
Collaborators from the United States and the former Soviet Union have initiated a search, called the Soviet-American Gallium Experiment (SAGE), deep underground in the heart of the Caucasus Mountains, near the Russian town of Baksan. There several dozen tons of liquid gallium (gallium melts at 30°C) await the shower of solar neutrinos passing through the earth to turn some of the gallium atoms into radioactive germanium. "You take gallium metal, add a neutrino to it, and you then produce germanium-71 atoms, which are radioactive," explains Wilkerson, who is a participating scientist in the SAGE experiment. More important, half of the neutrinos expected to interact with the gallium will be of lower-energy, particles produced by the proton-proton reaction, the sun's primary energy source. "It's very difficult to make these lower-energy neutrinos go away based on models of the sun," notes Wilkerson. Simply changing the sun's core temperature a bit, for example, would have little effect on their production rate. In solar physics the strength of the proton-proton reaction is fundamentally linked to the sun's observed luminosity; uncertainties in the flux of neutrinos expected from the reaction are minimal, around 2 percent. So any deficit in those lower-energy neutrinos would be better explained through particle physics (see Figure 1.7).
SAGE started out with 30 tons of gallium and has since increased its size to nearly 60 tons. With 30 tons of gallium, SAGE investigators expected to produce one atom of germanium-71 per day. Since this germanium has a short half-life, less than 12 days, it starts decaying. "After a month of running, you have about 16 atoms remaining," states Wilkerson. Through an intricate chemical extraction process, this very sparse harvest of germanium atoms is eventually isolated from the gallium and counted. After running half a year in 1990, SAGE scientists, like their counterparts in South Dakota and Japan, found considerably fewer neutrinos than theory predicted. By 1992, using data from both the 30- and the 57-ton setups, they calculated that they were finding only 44 percent of the amount expected (with the statistical uncertainty ranging from −21 to +17 percent). "We aren't claiming any new physics," cautions Wilkerson, "but the experiment seems to be hinting that something is happening. It is nearly impossible for solar physics models to account for this result and, when taken in conjunction with the Davis and Kamiokande data, offers the hint that the low solar neutrino flux may be due to new neutrino properties."
A similar gallium experiment, called GALLEX, is under way in a laboratory set within Italy's Gran Sasso tunnel in the Apennines, a mountain range northeast of Rome. Operating since 1991, the GALLEX
detector consists of a 100-ton aqueous solution of gallium chloride—a setup that GALLEX designers believed would simplify the germanium extraction process (gallium atoms make up 30 tons of that solution). Its initial findings, although very tentative (both SAGE and GALLEX must still undergo extensive calibrations), are intriguing. After running nearly a year, GALLEX captured 63 ± 16 percent of the neutrinos expected to be seen based on the standard solar model—more than the other detectors and just within shooting distance (given the statistical uncertainties) of the standard solar model, yet still a shortfall. The only way to reconcile the various GALLEX, SAGE, Kamiokande, and Homestake results with conventional solar physics is to lower the sun's core temperature by some 5 to 15 percent. But that's a conclusion that astrophysicists believe would play havoc with other well-measured solar parameters. What may be easier to tweak are the properties of the neutrinos themselves (see Figure 1.8).
The neutrino deficiency, so long seen in South Dakota and later in Japan, Russia, and Italy, might be explained if electron neutrinos have mass and so are able to "oscillate" on their way out of the sun; with mass, a fraction of them might be able to transform themselves into the other two neutrino types, the muon neutrino and the tau neutrino. Since the perchloroethylene and gallium detectors can only "see" electron neutrinos and not the other two types, this might easily explain the shortfall, why only a fraction of the expected neutrino signal is detected. By the time the solar neutrinos reach Earth, a large portion of them would have been transformed into either muon or tau neutrinos. Physics
Today editor Barbara Goss Levi, intrigued by this possible metamorphosis, once poetically described it this way:
In caverns deep under the ground
They hunt SNUs like hungry bloodhounds.
But maybe the prey
Can change 'long the way
And sneak by without being found.3
The idea that neutrinos might oscillate was first discussed by Pontecorvo in 1957. More recently, the idea has been extended by physicists Stanislav Mikheyev and Alexi Smirnov of the former Soviet Union, building on earlier work by Carnegie-Mellon physicist Lincoln Wolfenstein. According to this theory, known by its originators' last initials as the MSW model, neutrino oscillations would rarely occur in the vacuum of space, but the effect would be enhanced as neutrinos speed through more dense collections of celestial matter, such as the sun. As they propagate through this matter, the various neutrino types could "mix," making it easier for electron neutrinos to switch their identity and turn into either muon or tau neutrinos. "One of the reasons that the MSW solution is so popular," adds Wilkerson, "is that in addition to being able to accommodate the data from all four neutrino experiments, Homestake, Kamiokande, SAGE, and GALLEX, it also yields values of neutrino masses that are reasonable in most grand unified theories that go beyond today's particle physics model." The masses would not be very sizable; the electron neutrino, for instance, would still be hundreds of thousands of times less weighty than an electron. Some theorists are suggesting that electron neutrinos must have energies greater than 500,000 electron volts to experience the conversion, which may explain why the Homestake and Kamiokande detectors are noticing the biggest shortfall; each is particularly sensitive to neutrinos in that energy range.
"The MSW model is definitely the leading contender, and it's a beautiful theory," notes Wilkerson, "but it only takes a few ugly facts to kill a beautiful theory." Neutrino astronomers will have a better chance at testing the MSW model later this decade with the opening of the Sudbury Neutrino Observatory, a collaborative venture sponsored by Canada, Great Britain, and the United States. This ambitious detector is now under construction in a deep nickel mine more than 200 miles north of Toronto and will consist of 1000 tons of deuterium oxide, or "heavy" water (where the proton in each hydrogen nucleus is accompanied by a neutron), encased in a bubble of clear plastic. This casing, in turn, will be surrounded by 7000 tons of ordinary water and a vast array of sensitive
photon detectors. Numbering 9500, these phototubes will spot the bursts of Cerenkov light released whenever a neutrino interacts in the detector. More important, this heavy water detector will be able to distinguish all three flavors of neutrinos—the electron, muon, and tau. Unlike previous neutrino "telescopes," the Sudbury Observatory will be able to detect the distinctive signals released whenever an electron, muon, or tau neutrino hits a deuterium nucleus. No longer would the electron neutrino alone be recognized. Thus, the Sudbury detector should be able to prove whether the MSW model is the likely solution to the solar neutrino problem. "The Sudbury Neutrino Observatory can essentially find the smoking gun," says Wilkerson.
In the meantime, other neutrino detectors are scheduled to come on line in the near future. The Japanese are starting to build a facility similar to their current Kamiokande detector, except with 10 times more volume. This Super Kamiokande detector may be able to register up to 18,000 events per year, an efficiency hundreds of times better than the current instrument. Similar counts may be achievable in another planned experiment known as BOREXINO, a collaborative effort of Italian, U.S., and Russian scientists. Its design calls for a 100-ton liquid scintillator doped with boron, which would also be very sensitive to neutrino oscillations.
With the pace of neutrino astronomy increasing yearly, Wilkerson is optimistic that all these new detectors may at last provide the long-anticipated breakthrough. "It seems likely that the longstanding mystery of missing solar neutrino flux will be solved during the decade of the 90s," he predicts. "It will be interesting to see if the answers will come from solar physics or particle physics. If the current hint of new neutrino physics endures, it will be a curious occurrence that observing neutrinos from the sun has shed new light on physics beyond the current model of elementary particle physics."
Neutrino astronomers often report their results in terms of solar neutrino units, or SNUs. One SNU equals one neutrino capture per second for every 1036 atoms of the target material.
Bahcall, J.N. 1989. Neutrino Astrophysics. Cambridge University Press, Cambridge.
Bahcall, J.N., and R. Ulrich. 1988. Solar models, neutrino experiments, and helioseismology. Reviews of Modern Physics 60:297.
Bartusiak, Marcia. 1993. Through a Universe Darkly. Harper Collins, New York.
Bethe, H.A. 1939. Energy production in stars. Physical Review 55:434.
Bethe, H.A., and C.L. Critichfield. 1938. The formation of deuterons by proton combination. Physical Review 54:248.
Brown, T.M., J. Christensen-Dalsgaard, W.A. Dziembowski, P.R. Goode, D.O. Gough, and C.A. Morrow. 1989. Inferring the sun's internal angular velocity from observed p-mode frequency splittings. The Astrophysical Journal 343:526–546.
Christensen-Dalsgaard, J., T.L. Duvall, Jr., D.O. Gough, J.W. Harvey, and E.J. Rhodes, Jr. 1985. The speed of sound in the solar interior. Nature 315:378–382.
Christensen-Dalsgaard, J., D. Gough, and J. Toomre. 1985. Seismology of the sun. Science 229(Sept. 6):923–931.
Deubner, F.L. 1975. Observations of low wavenumber nonradial eigenmodes of the sun. Astronomy and Astrophysics 44:371.
Gough, D.O. (ed.). 1986. Seismology of the Sun and Distant Stars. D. Reidel Publishing Co., Dordrecht, Holland.
Gough, D.O., and J. Toomre. 1991. Seismic observations of the solar interior. Annual Review of Astronomy and Astrophysics 29:627–684.
Harvey, J., and the GONG Instrument Development Team. 1988. The GONG instrument. Pp. 203–208 in Proceedings of the Symposium on the Seismology of the Sun and Sun-like Stars, Tenerife, Spain, 26–30 September, 1988. ESA SP–286 December.
Harvey, J. W., J. R. Kennedy, and J. W. Leibacher. 1987. GONG: to see inside our sun. Sky & Telescope (Nov.):470–476.
Leighton, R. B., R. W. Noyes, and G. W. Simon. 1962. Velocity fields in the solar atmosphere: I. Preliminary report. Astrophysical Journal 135:474–499.
Libbrecht, K. G. 1988. Solar and stellar seismology. Space Sciences Review 47:275–301.
Libbrecht, K.G., and M.F. Woodard. 1991. Advances in helioseismology. Science 253(July 12):152–157.
Schwarzschild, B. 1992. GALLEX data can't quite lay the solar neutrino problem to rest. Physics Today XX(Aug.):17–20.
Waldrop, M. M. 1989. Gazing into the interior of the sun. Science 244(April 7):31.
Bahcall, J.N. 1989. Neutrino Astrophysics. Cambridge University Press, Cambridge.
Gough, D.O., and J. Toomre. 1991. Seismic observations of the solar interior. Annual Review of Astronomy and Astrophysics 29:627–684.
Harvey, J. W., J. R. Kennedy, and J. W. Leibacher. 1987. GONG: to see inside our sun. Sky & Telescope (Nov.):470–476.
Leibacher, J., R. Noyes, J. Toomre, and R. Ulrich. 1985. Helioseismology. Scientific American (Sept.):48–57.
Libbrecht, K.G., and M.F. Woodard. 1991. Advances in helioseismology. Science 253(July 12):152–157.
Freedman, D. H. 1991. The ghost particle mystery. Discover (May):66–72.
Freedman, H. 1986. Sun and Earth. Scientific American Books, New York.