7000 words
First published in Analog, Sept 1994, reprinted with minor modifications.
Molvray home page
 | Molvray science fiction home page

Gamma Ray Bursters: Unexplained Lights in the Sky

  Mia Molvray
© 1994
Distribution of GRBs, 1998, BATSE

Figure 1. Distribution of gamma ray bursts observed to date (1998). (From the BATSE web site. See also this non-specialist's BATSE site.)

      Gamma ray bursters were first noticed in the mid-1960s by military scientists looking at data coming in from the new Vela satellites that had been lofted to help verify whether the Russians really were abiding by the Test Ban Treaty of 1963. Since nuclear explosions are highly energetic events, they release gamma rays. The new space technology could scan Earth for bursts of gamma rays, and thus discover when tests were carried out, and how big the blasts were. The Cold War was in full swing, both sides were arming for Mutually Assured Destruction, and it was essential to know just how far the other side had gotten.

       It's easy to imagine the consternation at headquarters when word came in that not only must the Russians be cheating, they already had nukes in outer space! But the idea that the Russians could have developed space-based nuclear capability without any word leaking back to the U. S. was so implausible that the data were checked and rechecked.

      By 1967 the military knew that these strange bursts came from beyond the solar system. How far beyond, or why, is still unknown. The bursters were declassified in 1973, and their distribution in the sky was and continues to be mapped (Fig. 1), and their energies recorded, but that is all we know about them. Speculation abounds, of course, and includes violent events associated with neutron stars, black holes, or even antimatter.

      However, all these ideas require tailored assumptions, and one idea that may fit hasn't even been considered. The first notion, in the 1960s, that these were someone's spaceships, may turn out to be right -- except that they don't belong to the Russians. Given the characteristics of bursters, the hypothesis of alien spaceships is as plausible, perhaps more so, than some of the other ideas put forward. Of course, that requires the acceptance of aliens as plausible, and therein, no doubt, lies the problem.

       The assumption that aliens are improbable doesn't really appear to square with current knowledge. Astronomical data indicate increasingly that planets are common around sun-like stars. Water is a reasonably common molecule, and the likelihood of liquid water at the appropriate distance from a star is good. Paleontological data show that life arose almost the (geological) minute the last heavy meteor bombardment stopped. Indications of life appear so quickly that some think it may evolve very easily in liquid water, and may even have done so several times on Earth, dying out during successive heavy bombardments until they finally stopped. Once life arises, evolution to fill the various available niches appears inevitable. If none of these events are unlikely, the only major uncertainty is whether tool-use is likely to advance to the level of interstellar technology. In my view the probabilities imply that those who will not consider the possibility of alien technology are the irrational ones, not vice versa.

       I'll go into the pros and cons of the "acceptable" ideas, and then turn to the "unacceptable" idea: that, just maybe, some of the bursts come from the interstellar engines of a starfaring people.

Gamma Background

      Gamma rays are the most energetic form of light, with shorter wavelengths than ultraviolet and x-rays, and with energies between 100 kiloelectron Volts (keV), or 105 eV, up to 1015 eV. A large dose of light with 1015 electron volts of energy would not merely tan, but would kill a gamma-sunbather in moments. These highly energetic photons are produced by interactions between fields (magnetic, gravitational, electric) and high energy charged particles (electrons, protons, or charged subatomic particles). The fields energize and constrain the particles, forcing them to interact, resulting in high energy collisions that release specific subatomic particles and photons, including gamma rays, with characteristic energies. The particles and photons released allow scientists to work backward and deduce precisely the reaction that produced that particular pattern. For instance, if a proton and proton collide (also referred to as positive hydrogen ions) they can start the fusion reaction sequence diagrammed in Figure 2.

fusion reactions electron+positron trails
Figure 2. Left: proton-proton chain reaction resulting in protons, helium and energy. Two protons with one neutron each fuse to form a proton with two neutrons (deuterium ion), a beta particle, a neutrino, and 0.42 MeV of energy; an added proton can then fuse with the deuterium ion to produce a helium nucleus with three neutrons, a photon, and 5.49 MeV of energy; and in the final step the two helium nuclei fuse to form an ordinary helium nucleus with four neutrons (alpha particle), two protons, and 12.8 MeV of energy. This is the process that generates light and heat in stars like the sun. Right: Tracks made in a cyclotron by electron - positron pair production from gamma ray bombardment. In this case the gamma ray excited an existing electron. The reaction also works backward to produce gamma rays when electrons and positrons annihilate each other.

      The megaelectron volts of energy produced could energize a photon to any point on the spectrum, including gamma ray, or the energy could be dissipated in other ways, as heat for instance. As another example, if a positron, which is an antimatter electron, and an electron collide, they annihilate each other completely and release nothing but energy in the form of photons at 511 keV. The reaction on the extreme right in Figure 2 is the inverse, where a gamma ray's energy generates a positron.

      Gamma rays are produced in nuclear reactions, including fission in nuclear power plants or radioactive decay, fusion in nuclear bombs or stellar interiors, reactions in cyclotrons where the particles are speeded and constrained by magnetic fields, and more exotic events such as matter- antimatter annihilation.

       Our atmosphere is not fully transparent to any wavelength shorter than visible light, whether ultraviolet, x-ray, or gamma, and hence astronomical objects emitting gamma rays could only be studied easily once orbiting satellites were available. The earliest ground-based studies were carried out using Cherenkov radiation, secondary radiation generated in the atmosphere by collisions between molecules and incoming gamma radiation, but both the number and quality of observations were necessarily limited. The state of knowledge advanced with high altitude balloon studies, with incidental data from spacecraft launched for various purposes, such as solar observation, and with sporadic data from satellites launched to studyhigh energy phenomena, but only for a few days at a time. The real explosion of information came with orbiting satellites able to perform continuous observations. There are currently two civilian satellites monitoring these frequencies: the Russian Granat, launched in 1989, and the US Compton Gamma Ray Observatory launched in 1991.

Burster Background

      The Compton GRO has several instruments on board for studying the high energy universe, but the appropriately named Burst and Transient Source Experiment (BATSE) in particular finds gamma ray bursters. Its eight modules facing out in eight directions conduct a continual all-sky survey (except where its view is blocked by Earth), recording a burst candidate whenever two or more detectors simultaneously note the gamma count rising more than 5.5 standard deviations above the background. As a practical approximation, this means there is much less than a one percent chance that BATSE is mistaking background gamma rays for a burst event, and that the background level at the time of an event must be taken into account when determining its intensity. This instrument does not make images; it records the energy, duration, and direction of burst events, and allows spectra to be calculated.

      BATSE has expanded our knowledge of gamma ray bursters enough to make us realize how little we know about them. About 800 bursts occur per year, and the typical burster blooms in a few seconds, radiating nothing detectable except gamma rays, and disappears after a few seconds or minutes. For a high energy event to happen so quickly, the source has to be very compact, theoretically no more than a few hundred kilometers across, or it would take too long for the reaction to propagate through the whole mass.

      In the two years of data so far available, BATSE has shown us that bursters have the following characteristics:

Theories about Bursters

Isotropic Distribution

       The distribution of bursters throughout the sky (Fig. 1) is really what has astrophysicists scratching their heads. They know intense nuclear reactions produce gamma rays, they know that the energies and field strengths needed are most commonly found on neutron stars, and so they were pretty sure that the bursts should be associated with neutron stars in some way.

      Some scientists favored the idea of binary systems, with a neutron star pulling in ionized hydrogen and helium from a lower-mass companion along its magnetic fields, funneling the gas to its poles where the gas builds up until explosive helium fusion finally starts, releasing a sudden burst of gamma rays. Others suggested that random chunks, such as asteroids or passing planets, attracted by the star's immense gravity, accelerate toward it so rapidly that the final collison happens at a significant fraction of the speed of light, generating the rays. Some even suggested that the chunks could be antimatter. Yet others thought it might be the sudden release of energy from a massive neutron starquake.

      The problem is that neutron stars, strange as they are, occur within galaxies, like their more ordinary siblings. Since we are near an edge of our galaxy, and the galaxy is flattened, events confined to the galaxy are concentrated in the strip formed by the Milky Way. The early data on gamma ray bursters showed them more or less evenly distributed in the sky, but it was assumed this was only because the data were too patchy to show the true, disk-like pattern. Once BATSE showed that bursters really were distributed throughout the sky, all the neutron star theories no longer fit the data.

      There are two ways to have an even all-sky distribution. Events that are so close to us they fit within the one thousand light year thickness of the galaxy in our neighborhood will appear to be omnidirectional. On the other hand, events on a cosmological scale and out at the edge of the universe will also appear to be evenly distributed. Of course, if we receive a given amount of light in a gamma ray burst from right next door, the source must be comparatively weaker than if that same amount of light was part of a burst whose light had been spread over the whole universe. So, to understand what sort of events are causing the bursters, it is essential to know how far away they are.

Distances to Bursters

       No direct evidence exists at this point about how far away gamma ray bursts really are. Distances have been estimated using distribution data, which suggest that bursters occur either within about 300 light years of Earth, or within 150,000 - 300,000 light years (out at the Magellanic Clouds and beyond to the halo of dark matter assumed to control the gravitational dynamics of our galaxy), or out at billions of light years. The first number is well within the local thickness of the galactic disk (about 1000 light years). The second range of numbers refers to a distance well beyond the visible stars of the galaxy, whose globular clusters extend out to 40,000 light years, but within the range both of the Magellanic clouds (150,000 light years) and the halo of dark matter that is assumed to control the gravitational dynamics of the galaxy. These estimates suggest where the bursters aren't, but for a better idea of where they are, we will have to wait till spectral data are available.

       Redshifts provide a direct indication of distance, but their calculation requires accurate spectra. BATSE records energy levels, which are raw data whose distribution, duration and strength can be interpreted directly and hence have been studied right from the first availability of the data. BATSE also records spectra, that is, the full complement of wavelengths making up the light, but transforming that data into something interpretable requires calculation. Theoretical background for some of the calculations is still being worked out, and the programming to calculate gamma ray spectra has only recently been developed, so that very few are available for study, and those that are available come from bright bursters. It is possible that statistical studies of many more observations gathered over the next few years may allow an overall red-shift to be estimated.

      What few spectral data have been acquired so far speak against the idea that light from the bursters has travelled very far. No evidence of red- shifts has been found up till now, which is completely implausible for light travelling billions of years. Cyclotron resonance, a somewhat delicate pattern caused by magnetic fields at the gamma ray source, is unlikely to be visible in energy that has travelled across the whole universe, and the positron annihilation peak would no longer appear at 511 keV (due to redshift). Yet both Russian and Japanese satellites, as well as balloon studies, have detected these two patterns. BATSE data have not yet confirmed them, but spectral analysis has only just started, and it is too early to say that these patterns won't be found. What has been found in the BATSE data are spectral patterns that break at relatively low energies (Fig. 3), which is said to occur in high energy reactions. A reaction with sufficient energy close to home could be something relatively small, like a antimatter explosion limited to a few tons of mass, but an event at the edge of the universe would have to be so energetic as to be physically impossible. The low energy breaks in BATSE spectra suggest that at least some of the sources are within about 3000 light years of Earth.

GRB spectrum Figure 3. BATSE spectrum of burster #910601_69736, showing attentuation of gamma rays more energetic than approximately 700 keV. (From Schaefer et al.)

      Gamma ray burster data require a compact source, and they suggest a close source, whose emissions are limited to the highest energies of light, and which has some kind of magnetic containment. But instead of working on testing the obvious hypothesis of alien engines (perhaps because the probability of getting funding is considerably lower than the likelihood that aliens exist), scientists have concentrated on natural explanations.

Neutron star theories

      Astronomers have not really contemplated the close-to-home hypothesis, because no known astronomical phenomenon in our neighborhood could generate gamma ray bursts without other associated radiation. Instead many astronomers have concentrated on modifying theories involving neutron stars so that they fit the observed distribution.

      The true diehards have modified the local neutron star hypothesis to fit an isotropic distribution. Instead of neutron stars within the galaxy, they postulate high-speed neutron stars moving fast enough to escape galactic gravitation. By the time the stars reach 150,000 or more light years from the main galactic disk, which is still considered "local" in cosmological terms, they are assumed to have aged enough to change from pulsars to gamma ray bursters.

      The local neutron star theory requires many assumptions tailored to fit the theory. Pulsars have not been observed as strong gamma ray sources, and adherents to this theory assume that exceptionally strong magnetic fields on some stars turn them into bursters. The theory requires a very unusual initial population of stars to form the unusual high velocity pulsars, a population unlike any currently seen, and whose possible historical existence is not supported by the distribution of elements in the galaxy. Also, since our galaxy is not unique, similar halos of pulsars/bursters should be seen around other galaxies, but they are not.

      A galactic halo of neutron stars has enough evidence against it that many astrophysicists lean toward a cosmological hypothesis, where the bursters come from billions of light years away. Mere infalling matter cannot generate enough energy to create a gamma ray burst detectable from that far, so the hypothesis shifts to collisions of whole neutron stars, or of neutron stars and black holes, or of "strange" stars composed of quarks. These can indeed generate the energy required to make a burst visible to us, (though not enough to explain the energy breaks discussed above), and it is possible that there would be enough such events, given the vastness of the universe, to account for the observed frequency of bursters.

      There are, however, problems with the cosmological collisions idea. The first is the same problem of location, transposed to a different key. If neutron stars, or black holes, or the like are involved, the bursters ought to occur more commonly in the neighborhood of clusters or superclusters of galaxies. They don't appear to, though additional data may indicate some clustering that is not yet evident.

Lack of associated radiation

       Another objection to the collision hypothesis is the lack of any other radiation associated with the gamma ray bursters. Colliding stars or black holes would have to release one percent, possibly more, of their nuclear energy as pure gamma rays. This is unlikely in any case. Nor have theorists arrived at a consensus concerning the observable characteristics of such a cataclysm, that is, no-one knows what it would look like if it happened. And, in addition, where is the other 99% of the energy? We should also be seeing UV bursters, and even visible light bursters, sometimes all coming from the same event. But we don't.

       One explanation advanced for the lack of other radiation is that the extremely violent collisions supposed to generate bursts release a blast wave travelling out at a large fraction of the speed of light. When these relativistic particles hit the interstellar medium, kinetic energy is transformed to a tremendous increase in radiative energy, such that the photons radiated by the excited material are strongly increased in energy, or blue-shifted. Light of all wavelengths is blue-shifted until nothing but gamma rays is present.

      The astrophysical theoretician Paczynski has pointed out that relatively little radiation, including gamma radiation, might escape from neutron star or black hole collisions, since their gravity is so strong and their substance so dense. To overcome this problem, Paczynski suggests that bursters occur from collisions between "strange" quark stars, which would theoretically be expected to release gamma rays.

      A few times bursters have lasted long enough to allow other instruments to be focused on them. On one occasion a multi-day burster led to the identification of a faint source at optical and UV wavelengths, showing what appears to be a gamma-ray nova. Perhaps, associated radiation would be found in other bursters if we could just get a long enough look at them. But the very oddity of this particular source suggests that the nature of the others would be quite different.

Weak source deficit

      Yet another difficulty is that gamma rays interact with infrared photons in the intergalactic medium, losing their energy. We ought, logically, to be seeing a large number of very weak sources, if they're coming all the way from the edge of the universe, because so many would have been attenuated on the way. But in fact, the opposite pattern occurs. There is a deficit of weak sources, not an excess (Fig. 4).

weak source deficit Fig. 4. Number versus strength of gamma ray bursts. Strong bursters toward 100 on the x-axis, weak toward 1. Dashed line shows expected numbers of bursts for any given strength in a perfectly spherical distribution. The number of weak sources is somewhat too low for such a distribution, and the number of strong sources somewhat too high (though the sample size of the latter is small, making conclusions about strong sources suspect). (From Meegan et al.)

      The weak source deficit has another implication as well. It means we must not be at the center of burster distribution. To visualize this, imagine yourself as the guest of honor at the world's most enormous fireworks display. All around you, as far as the eye can see, fireworks are being set off. Right out to the horizon, you can see progressively fainter and fainter bursts. If, on the other hand, you approached one inner edge of this massive ring of fireworks, you'd have more very bright sources near you, and you wouldn't be able to see the weakest sources furthest away across the circle. This distribution is what BATSE has seen so far: a slight excess of strong sources, and a deficit of weak sources. If we're not at the center of the bursts, then they can't be happening out at the edge of the universe, because we are at the apparent center of everything happening that far away. An alternate explanation could be that different types of sources generate gamma bursts, and that some types are not evenly distributed.

Time Profiles

      Bursts share the characteristic of brevity, as their name implies, usually lasting from one to a few hundred seconds. Everything else about their timing varies (Fig. 5), which supports the idea of different causes generating the bursts. Variables include: the initial phase when they build up intensity, known as rise time, the peak energy, the number of peaks, their duration, and the duration of the whole burst.

      The duration of rise time, typically as long as a few seconds, is yet another problem for the colliding neutron star hypothesis. The energy release from such a collision would be over in a few milliseconds and would have a much faster rise time. The energy structure of the whole burst, with various levels of energy arriving at various times, is also too broad for what is expected in such a collision. Paczynski suggests that possibly the collision results in the brief existence of an infalling accretion disk that accounts for the longer lifetime of gamma bursts, though still not for their "long" rise times.

time distribution Fig. 5. Time profiles of gamma ray bursters. (From Fishman et al.). (Four included in full size image.)

Other Possibilities

      So far we know that gamma ray bursts are found throughout the sky, must come from very compact objects, and don't appear to have any other associated radiation. We also know of no natural process that might have this set of characteristics. Prevailing theories that attempt to press the bursters into a natural mold seem to require as many favorable assumptions as a government budget.

      What else could we consider? Could we be seeing hapless chunks of matter, or antimatter, bumping into cosmic strings? Because gamma rays are generated in strong fields, the potentially superconducting ends of the strings would have to be involved, according to Paczynski. However, cosmic strings pack so much energy in such a small space that all the energy would be released in a millisecond, according to theory, and the burster time profiles do not fit well. Still, further analysis of bursters from the perspective of string theory may bring interesting things to light.

      A local phenomenon that forms a spherical shell around us is the Oort cloud of comets, but the only source both small enough and energetic enough to generate the observed bursters at that distance would be antimatter annihilation. Astrophysicists are very reluctant to imagine free-flying chunks of antimatter so close to home, especially since none has ever been seen in the inner solar system.

      However, an even stranger idea may fit the facts at least as well as the natural explanations. At least some of the gamma ray bursters are local, and are due to antimatter annihilation, but are not free-flying. This antimatter is carefully contained in magnetic fields aboard starships accelerating for interstellar journeys. The ships may be accelerating to superlight speeds, or they may be the slow-moving barges of the fleet, ferrying tools or medicines or foods at sublight speeds to young colonies.

      Just as cargo planes and barges use different fuel, not all the spaceships may burn exactly the same fuel in exactly the same way, and the rate at which they burn may depend on the cargo, accounting for different time profiles. They might use positrons and electrons, protons and antiprotons, hydrogen and anti-hydrogen, or maybe something even more exotic. Since these are not accidental chunks of antimatter, it is not surprising that similar chunks haven't appeared in our solar system. The number of bursts accounted for by this "strange" hypothesis could be anywhere from few to many. Who knows, maybe this alien civilization has just started interstellar colonization and can only afford a few flights a year. Whether it is many or few, alien spaceships are potentially distributed widely in our local space, would account for the lack of other associated radiation, and would provide a plausible source of compact, high energy events not seen within our solar system (Table 1).

      It is worth noting that theorists suggest that bursters from the edge of space may be subject to gravitational lensing, and that the arrival of identical bursts separated in time would be proof of a distant origin. However, an equally good reason for identical bursts from the same place is that a ship has departed from the same port. Thus events that actually support the alien hypothesis may be subsumed under the cosmological one. A similar problem could arise in the calculation of overall red-shift, mentioned earlier. If alien ships account for only a few bursters, they could be lost in the data. Pooling data from different events should only be done when they are known to come from the same source, though then, of course, if their source is known, the need to pool the data disappears.

      Table 1. Gamma ray burster characteristics supportive and not supportive of the alien technology hypothesis.

pro con either
no other associated radiation no repeaters isotropic distribution
controlled burns (slow rise times,varying time profiles) insufficient time profile similarity weak source deficit
compact source
511 keV spectra
cyclotron resonance?
local source? (at least in some cases?)

      The major problem with this idea of alien gamma ray ships, is that one would expect the aliens to have ports, and therefore one would expect bursters to be concentrated there. The isotropic distribution of the bursters is a problem yet again. One would also expect similarity in some of the time profiles, though if alien ships account for only a few bursters, it may take time for similarities to become evident. In this connection it would be helpful if time profiles, and spectra when we get them, were mapped on to the distribution, so that possible local areas of similarity were evident.

      Another possible problem is that repeated bursts have never been observed. The aliens may require only one massive burst for their purposes, but logically it would seem that a series of bursts would allow for more controlled acceleration and a smoother ride. But who knows, they may be speed demons with the ability to control inertial effects; or jumps to hyperspace may not work like that....

      One question, initially raised by Fermi when the presence of alien civilizations was being seriously considered, is "Why haven't they visited?" I can think of several reasons. One is that space is too vast for the probability of bumping into Earth to be good. However, if the bursters all around us are ships within 300 light years, that is not a good answer. Another possibility is that Earth is a backwater. After all, Okiwi Bay in New Zealand may be an extraordinarily beautiful place, and not all that far away, but very few people have ever visited it, or even have a map detailed enough to find it. And finally, another answer may be, "They have visited, we just don't admit it." Though almost all UFO sightings have natural explanations, there are a few that cannot be explained away.

      The wonderful thing about mysteries is the possibility they give us of seeing new wonders. Gamma ray bursters are sure to show us worlds undreamed of, -- and some of them may even be populated.


      Chupp, E. 1992. The gamma-ray cosmos. Science, 258: 1894-1896.

      Fishman, G. J. et al. 1989. Proceedings GRO Science Workshop, 2.39-2.44 (Goddard Space Flight Center, Greenbelt, Maryland).

      Li, Hui, and C. D. Dermer. 1992. Gamma-ray bursts from high-velocity neutron stars. Nature, 359: 514-516.

      Meegan, C. A. et al. 1992. Spatial distribution of gamma-ray bursts observed by BATSE. Nature, 355: 143-145.

      Paczynski, B. 1988. Gamma-ray bursts from cusps on superconducting cosmic strings at large redshifts. The Astrophysical Journal, 335: 525-531.

      Paczynski, B. 1992. Estimating redshifts for gamma-ray bursts. Nature, 355: 521-522.

      Schaefer, B. E., et al. 1992. High-energy spectral breaks in gamma-ray bursts. The Astrophysical Journal, 393: L51-L54.

      Schwarschild, B. 1992. Compton observatory data deepen the gamma ray burster mystery. Physics Today, Feb.: 21-24.

Molvray sf home page