by Isaac Asimov
The fact of the reversals is easier to ascertain, however, than the reasons for it.
In addition to long-term drifts of the magnetic field, there are small changes during the course of the day. These suggest some connection with the sun. Furthermore, there are disturbed days when the compass needle jumps about with unusual liveliness. The earth is then said to be experiencing a magnetic storm. Magnetic storms are identical with electric storms and are usually accompanied by an increase in the intensity of auroral displays, an observation reported as long ago as 1759 by the English physicist John Canton.
The aurora borealis (a term introduced in 1621 by the French philosopher Pierre Cassendi, and Latin for “northern dawn”) is a beautiful display of moving, colored streamers or folds of light, giving an effect of unearthly splendor. Its counterpart in the Antarctic is called the aurora australis (“southern dawn”). In 1741, the Swedish astronomer Anders Celsius noted its Connection with Earth’s magnetic field. The auroral streamers seem to follow the earth’s magnetic lines of force and to concentrate, and become visible, at those points where the lines crowd most closely together—that is, at the magnetic poles. During magnetic storms, the northern aurora can be seen as far south as Boston and New York.
Why the aurora should exist was not hard to understand. Once the ionosphere was discovered, it was understood that something (presumably solar radiation of one sort or another) was energizing the atoms in the upper atmosphere and converting them into electrically charged ions. At night, the ions would lose their charge and their energy, the latter making itself visible in the form of auroral light. It was a kind of specialized air glow, which followed the magnetic lines of force and concentrated near the magnetic poles because that would be expected of electrically charged ions. (The airglow itself involves uncharged atoms and therefore ignores the magnetic field.)
THE SOLAR WIND
But what about the disturbed days and the magnetic storms? Again the finger of suspicion points to the sun.
Sunspot activity seems to generate magnetic storms. How such a disturbance 93 million miles away can affect the earth is not easy to see, yet it must be so, since such storms are particularly common when sunspot activity is high.
The beginning of an answer came in 1859, when an English astronomer, Richard Christopher Carrington, observed a starlike point of light burst out of the sun’s surface, last 5 minutes, and subside. This is the first recorded observation of a solar flare. Carrington speculated that a large meteor had fallen into the sun, and assumed it to be an extremely unusual phenomenon.
In 1889, however, George E. Hale invented the spectroheliograph which allowed the sun to be photographed in the light of a particular spectral region. This picked up solar flares easily and showed that they are common and are associated with sunspot regions. Clearly, solar flares are eruptions of unusual energy that somehow involve the same phenomena that produce the sunspots (hence, the cause of flares is as yet unknown). When the solar flare is near the center of the solar disk, it faces Earth, and anything shot out of it moves in the direction of the Earth. Such central flares are sure to be followed by magnetic storms on Earth after a few days, when particles fired out by the sun reach Earth’s upper atmosphere. As long ago as 1896, such a suggestion had been made by the Norwegian physicist Olaf Kristian Birkeland.
As a matter of fact, there was plenty of evidence that, wherever the particles might come from, the earth was bathed in an aura of them extending pretty far out in space. Radio waves generated by lightning had been found to travel along the earth’s magnetic lines of force at great heights. (These waves, called whistlers because they were picked up by receivers as odd whistling noises, had been discovered accidentally by the German physicist Heinrich Barkhausen during the First World War.) The radio waves could not follow the lines of force unless charged particles were present.
Yet it did not seem that these charged particles emerged from the sun only in bursts. In 1931, when Sydney Chapman was studying the sun’s corona, he was increasingly impressed by its extent. What we can see during a total solar eclipse is only its innermost portion. The measurable concentrations of charged particles in the neighborhood of the earth were, he felt, part of the corona. Hence, in a sense, the earth is revolving about the sun within that luminary’s extremely attenuated outer atmosphere. Chapman drew the picture of the corona expanding outward into space and being continually renewed at the sun’s surface. There would be charged particles continuously streaming out of the sun in all directions, disturbing Earth’s magnetic field as it passed.
This suggestion became virtually inescapable in the 1950s, thanks to the work of the German astrophysicist Ludwig Franz Biermann. For half a century, it had been thought that the tails of comets, which always point generally away from the sun and increase in length as the comet approaches the sun, were formed by the pressure of light from the sun. Such light-pressure does exist, but Biermann showed that it is not nearly enough to produce cometary tails. Something stronger and with more of a push was required; this something could scarcely be anything but charged particles. The American physicist Eugene Norman Parker argued further in favor of a steady outflow of particles, with additional bursts at the time of solar flares and, in 1958, named the effect the solar wind. The existence of this solar wind was finally demonstrated by the Soviet satellites Lunik I and Lunik II, which streaked outward to the neighborhood of the moon in 1959 and 1960, and by the American planetary probe Mariner II, which in 1962 passed near Venus.
The solar wind is no local phenomenon. There is reason to think it remains dense enough to be detectable at least as far out as the orbit of Saturn. Near the earth the velocity of solar-wind particles varies from 220 to 500 miles per second, and it takes particles three and a half days to travel from the sun to the earth. The solar wind causes a loss to the sun of a million tons of matter per second—a loss that, however huge in human terms, is utterly insignificant on the solar scale. The density of the solar wind is about a quintillionth that of our atmosphere; and in the entire lifetime of the sun, less than 1/100 of 1 percent of its mass has been lost to the solar wind.
The solar wind may well affect our everyday life. Beyond its effect on the magnetic field, the charged particles in the upper atmosphere may ultimately have an effect on the details of Earth’s weather. If so, the ebb and flow of the solar wind is still another weapon in the armory of the weather forecast.
THE MAGNETOSPHERE
An unforeseen effect of the solar wind was unexpectedly worked out as a result of satellite launchings. One of the prime jobs given to the artificial satellites was to measure the radiation in the upper atmosphere and nearby space, especially the intensity of the cosmic rays (charged particles of particularly high energy). How intense was this radiation up beyond the atmospheric shield? The satellites carried Geiger counters (first devised by the German physicist Hans Geiger in 1907 and vastly improved in 1928), which measure particle radiation in the following way: The counter has a box containing gas under a voltage not quite strong enough to send a current through the gas. When a high-energy particle of radiation penetrates into the box, it converts an atom of the gas into an ion. This ion, hurtled forward by the energy of the blow, smashes neighboring atoms to form more ions, which in turn smash their neighbors to form still more. The resulting shower of ions can carry an electric current; and for a fraction of a second, a current pulses through the counter. The pulse is telemetered back to earth. Thus the instrument counts the particles, or flux of radiation, at the location where it happens to be.
When the first successful American satellite, Explorer I, went into orbit on 31 January 1958, its counter detected about the expected concentrations of particles at heights up to several hundred miles. But at higher altitudes (and Explorer I went as high as 1,575 miles), the count fell off; in fact, at times it dropped to zero! This might have been dismissed as due to some peculiar accident to the counter, but Explorer III, launched on 26 March 1958, and reaching an apoge
e of 2,100 miles, had the same experience. So did the Soviet Sputnik III, launched on 15 May 1958.
James A. Van Allen of the State University of Iowa, who was in charge of the radiation program, and his aides came up with a possible explanation. The count fell virtually to zero, they decided, not because there was little or no radiation, but because there was too much. The instrument could not keep up with the particles entering it, and blanked out in consequence. (This would be analogous to the blinding of our eyes by a flash of too-bright light.)
When Explorer IV went up on 26 July 1958, it carried special counters designed to handle heavy loads. One of them, for instance, was shielded with a thin layer of lead (analogous to dark sunglasses) which would keep out most of the radiation. And this time the counters did tell another story. They showed that the “too-much-radiation” theory was correct. Explorer IV, reaching a height of 1,368 miles, sent down counts that allowing for the shielding, disclosed a radiation intensity far higher than scientists had imagined.
It became apparent that the Explorer satellites had only penetrated the lower regions of this intense field of radiation. In the fall of 1958 the two satellites shot by the United States in the direction of the moon (so-called moon probes)—Pioneer I, which went out 70,000 miles, and Pioneer III, which reached 65,000 miles—showed two main bands of radiation encircling the earth. They were named the Van Allen radiation belts, but were later named the magnetosphere in line with the names given other sections of space in the neighborhood of the earth (figure 5.6).
Figure 5.6. The magnetosphere, or Van Allen radiation belts, as traced by satellites. They appear to be made up of charged particles trapped in the earth’s magnetic field.
It was at first assumed that the magnetosphere was symmetrically placed about the earth, rather like a huge doughnut, and that the magnetic lines of force were themselves symmetrically arranged. This notion was upset when satellite data brought back other news. In 1963, in particular, the satellites Explorer XIV and Imp-I were sent into highly elliptical orbits designed to carry them beyond the magnetosphere if possible.
It turned out that the magnetosphere has a sharp boundary, the magnetopause, which is driven back upon the earth on the side toward the sun by the solar wind, but which loops back around the earth and extends an enormous distance on the night side. The magnetopause is some 40,000 miles from the earth in the direction of the sun, but the teardrop tail on the other side may extend outward for a million miles or more. In 1966, the Soviet satellite Luna X, which circled the moon, detected a feeble magnetic field surrounding that world which may actually have been the tail of earth’s magnetosphere sweeping past.
The entrapment of charged particles along the magnetic lines of force had been predicted in 1957 by an American-born Greek amateur scientist, Nicholas Christofilos, who made his living as a salesman for an American elevator firm. He had sent his calculations to scientists engaged in such research, but no one had paid much attention to them. (In science, as in other fields, professionals tend to disregard amateurs.) It was only when the professionals independently came up with the same results that Christofilos achieved recognition and was welcomed into the University of California. His idea about particle entrapment is now called the Christofilos effect.
In August and September 1958 to test whether the effect really occurs in space, the United States fired three rockets carrying nuclear bombs 300 miles up and there exploded the bombs—an experiment that was named Project Argus. The flood of charged particles resulting from the nuclear explosions spread out along the lines of force and were indeed trapped there. The resulting band persisted for a considerable time; Explorer IV detected it during several hundred of its trips around the earth. The cloud of particles also gave rise to feeble auroral displays and disrupted radar for a while.
This was the prelude to other experiments that affected or even altered Earth’s near-space environment, and some of them met with opposition and vast indignation from sections of the scientific community. A nuclear bomb exploded in space on 9 July 1962 introduced marked changes in the magnetosphere, changes that showed signs of persisting for a prolonged interval, as some disapproving scientists (such as Fred Hoyle) had predicted. The Soviet Union carried out similar high-altitude tests in 1962. Such tampering with the natural state of affairs may interfere with our understanding of the magnetosphere, and it is unlikely that this experiment will be soon repeated.
Then, too, attempts were made to spread a layer of thin copper needles into orbit about the earth to test their ability to reflect radio signals, in order to establish an unfailing method for long-distance communication. (The ionosphere is disrupted by magnetic storms every once in a while and then radio communication may fail at a crucial moment.) Despite the objection of radio astronomers who feared interference with the radio signals from space, the project (Project West Ford, after Westford, Massachusetts, where the preliminary work was done) was carried through on 9 May 1963. A satellite containing 400 million copper needles, each three-quarters of an inch long and finer than a human hair—50 pounds’ worth altogether—was put into orbit. The needles were ejected and then slowly spread into a world-circling band that was found to reflect radio waves just as had been expected. This band remained in orbit for three years. A much thicker band would be required for useful purposes, however, and it is doubtful whether the objections of the radio astronomers can be overcome for that.
PLANETARY MAGNETOSPHERES
Naturally, scientists were curious to find out whether there were radiation belts about heavenly bodies other than the earth. If Elsasser’s theory is correct, a planetary body must fulfill two requirements in order to have a sizable magnetosphere: it must have a liquid, electrically conducting core, in which swirls can be set up; and it must have a fairly rapid period of rotation to set up those swirls. The moon, for instance, is of low density and is small enough not to be very hot at its center, and thus almost certainly contains no liquid metal core. Even if it does, the moon rotates far too slowly to set it swirling.
The moon, therefore, should have no magnetic field of any consequence on both scores. Nevertheless, no matter how clear-cut such deduction may be, it always helps to have a direct measurement, and rocket probes can easily be outfitted to make such measurements.
Indeed, the first lunar probes, the Soviet-launched Lunik I (2 January 1959) and Lunik II (September 1959), found no signs of radiation belts about the moon, and this finding has been confirmed in every approach to the moon since.
Venus is a more interesting case. It is almost as massive and almost as dense as Earth and must certainly have a liquid metallic core much as Earth does. However, Venus rotates very slowly, even more slowly than the moon. The Venus probe, Mariner 2, in 1962, and all the Venus probes since have agreed that Venus has virtually no magnetic field. The magnetic field it does have (possibly resulting from conducting effects in the ionosphere of its dense atmosphere) is certainly less than 1/20,000 as intense as Earth’s.
Mercury is also dense and must have a metallic core; but, like Venus, it rotates very slowly. Mariner 10, which skimmed Mercury in 1973 and 1974, detected a weak magnetic field, somewhat stronger than that of Venus, and with no atmosphere to account for it. Weak as it is, Mercury’s magnetic field is too strong to be caused by its slow rotation. Perhaps because of Mercury’s size (considerably smaller than that of either Venus or Earth), its metallic core is cool enough to be ferromagnetic and possesses some slight property as a permanent magnet. However, we cannot tell yet whether it does.
Mars rotates reasonably rapidly but is smaller and less dense than Earth. It probably does not have a liquid metallic core of any size, but even a small one may produce some effect, and Mars seems to have a small magnetic field, stronger than Venus’s though much weaker than Earth’s.
Jupiter is another thing altogether. Its giant mass and its rapid rotation would make it an obvious candidate for a magnetic field if there were certain knowledge of the conducting characteris
tics of its core. Back in 1955, however, when such knowledge did not exist and no probes had yet been constructed, two American astronomers, Bernard Burke and Kenneth Franklin, detected radio waves from Jupiter that were nonthermal: that is, they did not arise merely from temperature effects. They had to arise from some other cause, perhaps high-energy particles trapped in a magnetic field. In 1959, Frank Donald Drake did so interpret the radio waves from Jupiter.
The first Jupiter probes, Pioneer 10 and Pioneer 11, gave ample confirmation of theory. They had no trouble detecting a magnetic field (for compared with Earth’s, it was a giant) even more intense than was to be expected from the huge planet. The magnetosphere of Jupiter is some 1,200 times as large as Earth’s. If it were visible to the eye, it would fill an area of the sky (as seen from Earth) that was several times larger than the full moon appears to us. Jupiter’s magnetosphere is 19,000 times as intense as Earth’s; and if manned space vessels ever reach the planet, it would form a deadly barrier to a close approach, embracing moreover the Galilean satellites.
Saturn also has an intense magnetic field, one that is intermediate in size between that of Jupiter and Earth. We cannot yet tell by direct observation, but it seems reasonable to suppose that Uranus and Neptune also have magnetic fields that may be stronger than Earth’s. In all the gas giants, the nature of the liquid, conducting core would be either liquid metal or liquid metallic hydrogen—the latter almost certainly in the case of Jupiter and Saturn.