Observable Characteristics of Extraterrestrial Technological Civilizations

 Robert A. Freitas, Jr.

Xenology Research Institute, 8256 Scottsdale Drive, Sacramento, California 95828, USA.
Journal of the British Interplanetary Society 38:106-112 (1985).
Advanced extraterrestrial civilisations which make extensive use of the fusion fuel resources of their local star and planetary system have numerous potentially observable characteristics. A circurnstellar nuclear fuel molecular effusion cloud, the principal observable, rapidly dissociates and neutralises to the atomic ground state, permitting the detection of hydrogen and tritium hyperfine transition radio lines at 1420 MHz and 1516 MHz, respectively. The negligible natural abundance of neutral atomic ground-state tritium suggests that its hyperfine line, the "tritium waterspout" centred in the radio SETI "waterhole" band, is ideal for interstellar communication and future SETI searches. Other possible observables of advanced civilisations include redshifted neutrino point sources, an artificial radio spectrum, anomalous blackbody radiation, fission waste absorption lines, Doppler and stellar spectral anomalies, and extraordinary magnetic fields.

Note: This web version is derived from an earlier draft of the paper and may possibly differ in some substantial aspects from the final published paper.



Conditions favourable to the development of life, intelligence, and civilisation may be widespread in the Galaxy. Current programmes to detect these civilisations, usually involving searches for powerful radio beacons, assume that alien societies are purposely calling attention to themselves [1-3]. A better approach is to search for the natural and anticipated byproducts of technological civilisation. This avoids the tenuous assumption that extraterrestrial intelligences are actively seeking contact or communication. Still, observing any but the most technologically rapacious civilisations is difficult with present equipment [4].

One of the best ways to detect evidence of intelligent activity around another star is to search for the material effluents of a spacefaring industrial civilisation. The mainstay of technical civilisation is energy, and nuclear fusion energy is the only plausible source for long-lived societies. The bulk of planetary mass in a solar system is likely to be fusionable hydrogen and helium, and the sun is a natural fusion reactor, so both may plausibly be employed.

However, solar energy alone may be regarded as insufficient because it restricts the maximum rate of power consumption. Civilisations may wish to release greater energies than the 1026-1027 watts developed by their sun. Further, a G0 star typically emits 1044 joules during its main sequence lifetime, whereas 1 Msun (solar mass) of fusion fuel carefully burned in a controlled reactor releases 2 x 1041 joules - a considerable motivation for astrophagy [5]. [Ed. Note 1]

Advanced civilisations may turn to artificial fusion as a supplementary source of energy. They will begin by cannibalising fusionables from their planets to be burned in space has based reactors. Later, as their appetite for energy grows, they may begin draining light element fusionables from their star. G0 and G1 stars differ by 0.04 Msun, and an artificial 1% mass deficit would be undetectable with current instrumentation. Even a 10% deficit might go unnoticed because of our incomplete understanding of stellar nucleosynthesis.

To recover the full energy theoretically available via fusion, hydrogen atoms must be burned all the way to iron - Fe has the maximum binding energy per nucleon, and thus represents the natural endpoint of all fusion reactions. It is likely that heavy element products of fusion burning will be used for various constructive purposes - artificial structures, atmospheres, etc. - in addition to being burned for additional energy. But the closer the endproducts approach Fe the less fusion energy they can generate per atom burned, a law of diminishing marginal return which gives a very steep decrease in fusion efficiency and rising fusion ignition temperatures with increasing atomic mass.

The most likely mode of fusion energy production is hydrogen burning to 4He at temperatures < 4 x 106 K, much like the PPI, PPII, and PIII chains in stellar nucleosynthesis [6], which releases 80% of the available nuclear binding energy in the fuel. Intermediate fusion products include D (2H), T (3H), 3He, 7Li, 7Be, 8Be, and 8B. The temperature must be raised to 100 x 106 K to burn 4He to 12C to extract another 7% of the binding energy, and another 11% can be extracted only by resorting to temperatures in excess of 109 K. The remaining 2% of the binding energy escapes as neutrinos. Low temperature fuels will be burned in preference to higher temperature fuels, with the latter perhaps stored in tankage until required.

The D-3He reaction has been discussed as a potential propulsion reaction for interstellar rockets [7], and Powell [8] argues that a large-scale fusion-based civilisation will generate major surpluses of 3He. The D-T reaction is the most widely studied because it is one of the easiest to ignite, yields the second-highest energy of any fusion reaction after D-3He, and is expected to be the fuel used in the first commercial fusion reactors on Earth [9-[1]. T decays to 3He with a half-life of 12.5 yr and B--emits the second-softest radiation after 187Re (~ 103 watts/kg) of any radioactive element, thus may be regarded as a biologically benign radioisotope. All three isotopes are plausible fusion fuel candidates, in addition to H and 4He.


Emission characteristics of large-scale fusion energy facilities in circurnstellar orbit include neutrons, accidental leakage of high-energy plasma fuel/product mixtures, energetic neutrinos, contaminated reactor wall components, gamma ray photons, and fusion fuel diffusion and accidental spillage from storage bunkers. Reaction neutrons not absorbed by breeder blankets and containment walls decay with a half-life of 750 sec into high-energy protons and electrons which merge into the solar wind. Escaped plasma particles cannot be slowed appreciably by normal processes and are too energetic to de-ionise, dispersing rapidly into the interstellar medium along heliomagnetic field lines and becoming indistinguishable from the primary cosmic ray background. The neutrino (n) emission is observable in principle as a point source with a 1-10 eV gravitational redshift relative to solar emanations, but is unmeasurable using present-day detector technology [12-[3] and must await the development of high-resolution neutrino spectroscopy. Contaminated components rich in artificial radionuclides are easily localised and recycled, and hence arc not readily observable.

Gamma rays are produced in the reactions H(n, g)D, H(D, g)3He, 3He(4He, g)7Be, 7Li(p, g)24He, and 7Be(p, g)8Be, and by annihilation of positrons from p(p, e+, v)D reactions, with energies less than I McV, but these should remain confined within the reactor vessel. As few as 1035 photons/sec at 1017 Hz released isotropically from near a star 10 pc distant should be detectable by the orbiting Einstein X-Ray Observatory, but such a flux rate (~ 10-4 rad/sec, human lethality ~ 103 rads) would probably render the circurnstellar shell biologically uninhabitable.

Fusion fuel effluence is plausibly observable over interstellar distances. Fuel will be stored as diatomic ( H2, D2, T2) or monoatomic (3He, 4He) gas at To 300 K. For hydrogen gas emitted from an artificial circurnstellar shell of radius R (1 AU), the mean thermal velocity vo = (2kTo/2mp)1/2 1.57 km/sec << (2GM/R)1/2 = 42.2 km/sec = solar escape velocity, where 2mp = hydrogen molecular mass, k is Boltzmann's constant and G is the gravitation constant. Leakage will occur from the surface of the circurnstellar shell at some rate DM kg/sec, and molecules are subsequently swept from the solar system by solar wind particle collisions over the age of the civilisation t. The most abundant effluent is molecular hydrogen. We assume a thick cloud model such that DM/mp >> Jw, the solar wind proton flux, and the entire solar wind is absorbed by the cloud. If n is solar wind proton number density near the circurnstellar shell and vp is solar wind velocity at the shell, then Jw = 4pnvpR2 and the rate of ejection of H2 is limited to Jw. Also, in a perfectly elastic collisional sequence two protons in molecular form are ejected but one striking proton is halted, for a net collisional ejection rate of 1/2 (4pnvpR2), So the number of hydrogen molecules present in the artificial cloud NH2 ~ t[(DM/2mp)-(4pnvpR2)/2].

Molecular hydrogen is dissociated with an efficiency b = ( H2 residence time)/( H2 dissociation lifetime) = tr/td 1. For photodestruction of H2 near the Sun, the best value is td = 5 x 1010 sec [14-[6]. In the thick cloud model tr = collision time (tc) + ejection time (te) = (DMt/2mp)/ (4pnvpR2) + (Rb-R)/(vp/2), where Rb is the distance at which the effusion cloud density falls to the interstellar background. Rb ~ R (no/nb) 1/2, where no is emission number density near the shell (no = DM/4pR2(2mpvo)) and nb = interstellar background number density (~ 0.1 H cm-3), so Rb = 0.0026 DM 1/2 AU ~ radius of heliopause. For R 1 AU, n = 5 cm-3, vp ~ 400 km/sec, and Rb ~ 100 AU: b = 1.0 for NH2 1046, 0.001 - 1.0 otherwise. Ionisation time for H in the Solar System is 1016-1017 sec [15-[6], as compared to 5 x 1010 sec for neutralisation of H+ [17]. Hence fusion fuel leakage may produce a cloud of fully neutralised atoms surrounding the target star if NH 1046, with number of cloud atoms NH = 2 NH2 ~ DMt/mp. Other neutral fuel atoms may also be present in varying lesser amounts.

What is DM? Fusion fuel emissions are due largely to diffusion. Leakage rates are negligible if a small number of very large fuel tanks are employed exclusively with radius r ~ 109 metres, near the theoretical maximum structure size in a 1 AU heliocentric orbit for normal building materials [18]. However, small-scale users will require more convenient storage, of necessity a very large number of small tanks. If the tanks are spherical and of thickness t, the entire fleet holds a mass of gas Mg, and Do is the classical diffusion coefficient, DM = 3DoMg/rt. For hydrogen gas stored in a zirconium and tritium stored in stainless steel at 300 K, Do ~ 10-15 metres2-sec-1 [ [9-20]. Thus for example, Mg 10-4 MJ (Jovian mass) = 2 x 1023 kg, r = 10 m and t = 1 cm gives DM = 1010 kg/sec leakage. For t = 103 - 109 yr, NH = 1047 - 1053 atoms, more if losses due to fuel processing and transfer operations are taken into account. So diffusion arguments cannot rule out the existence of dense artificial clouds near an extraterrestrial civilisation.  In the most fundamental limit, if Lf is total fusion luminosity available to the civilisation, eH = fusion efficiency of hydrogen = 0.92% c2 joules/kg, and c = speed of light, then DM Lf/eH. For Lf ~ Lsun (Solar luminosity), DM 1012 kg/sec hydrogen leakage.

2.1 Natural Background

The natural background levels of neutral atomic H, D, T, 3He, and 4He in the neighbourhood of a late-type main sequence star should approximate the local interstellar medium. This is because even though all fusion isotopes are emitted as solar wind ions and in solar flares [21-25], their velocities are too high for braking or neutralisation to occur so there is no enhancement of local background- D, T, and 3He are unimportant in advanced stellar evolution [26], and the decay of tritium further ensures its natural absence in the neutral atomic state.

Current measurements of natural neutral hydrogen abundances in the interstellar medium range from 0.01-0.2 H cm-3 [27-29], with a mean value usually taken as 0.1 H cm-3 [30-31]. Relative to hydrogen, other candidate fusion fuel isotopes have the following natural abundances: D/H = 2 x 10-6 - 3.5 x 10-4 [27, 32-36]; T/H < 10-11 [25]; 3He/H = 1-4 x 105 [32-33, 37]; and 4He/H = 0.069 - 0.1 [32-33]. The mean excess of fusion fuel hydrogen atoms is NH/(4/3pRb3) ~ 7 x 10-14 DMt (H m-3), plausibly up to ~ 1013 H m-3

2.2 Optical Anomalies

Optical emission lines from an artificial light-element effusion cloud should not exist because of the low cloud temperature and its relatively great distance from possible sources of excitation. Observable absorption lines from the hypothetical artificial cloud must have column densities exceeding those of natural isotopes in the stellar atmosphere, which is unlikely. It is more difficult to resolve the Lyman and Balmer lines of tritium from those of deuterium than the deuterium lines from those of hydrogen, so although the natural background of stellar neutral atomic tritium is negligible a column density equivalent to that of naturallyoccurring stellar deuterium would probably be required for tritium detection.

The D/H ratio has been determined by measuring interstellar deuterium absorption in the UV, from orbiting observatories, in early-type stars and a few late-type stars [27, 29, 35-36, 38]. Except in the case of Alpha Centauri, measured H concentration and D/H in the line of sight appear consistent with other observations and independent estimates for the interstellar medium. D has yet to be detected in any stellar atmosphere [39-41]. 3He has been discovered spectroscopically in only a few early-type peculiar stars [42-45] but not in any later-type star except the Sun [46].

A final consideration is that to recognise a detected line as a local anomaly, off-star comparison spectra are required - an impossible requirement for optical absorption lines except in the rare case of a close visual binary.

2.3 Radio Anomalies

In radio frequencies, excited neutral atoms can be observed via recombination lines in energetic environments such as HII regions [47]. For instance, Palmer [48] measured the H 109a recombination line at 5008.923 MHz and the 4He 109a recombination line at 5010.964 MHz to determine the relative abundance of helium in various nebulae. From the earlier discussion, we expect that most leakage gas will be present in the neutral atomic state and that recombination lines should be relatively weak.
2.3.1 Hyperfine Transition Lines
The spontaneous magnetic dipole hyperfine transition is the only plausible observational characteristic of an artificial, -neutral atomic gas cloud of hydrogen or helium isotopes in the ground state. However, in ground-state neutral helium, atomic electrons occupy all available spin states and hence the spontaneous spin transition is prohibited by the Pauli exclusion principle. The 3HeI line at 6739.7013 MHz [4950] and the corresponding 4HeI line arise from a hyperfine transition in the 2s level, but excitation from 1s2, 1S0) to the metastable triplet state (1s 2s, 3S1 requires 21.25 eV, 87% of the first ionisation energy, which is not available near late main sequence stars. Among ground-state hydrogen isotopes, the deuterium hyperfine line at 327.384 352 5222(17) MHz [51] has a brightness temperature below that of the galactic synchrotron background, and hence could only be observed [34, 52-54] in absorption against very bright radio sources with state-of-the-art equipment. This rules out the radio detection of artificial deuterium enhancements near main sequence stars except in the rare instance of a normal star occulting a very active radio source. Thus the neutral hydrogen ground-state hyperfine transition line at 1420.405 751 768(3) MHz [55] and the neutral tritium ground-state hyperfine line at 1516.701 9064(16) MHz [56] are the two most promising observational candidates.
2.3.2 The Tritium Waterspout
It has most commonly been argued [1-3] that a technological civilisation wishing to attract attention to itself to initiate contact would employ a powerful radio beacon. Operation on a single narrowband frequency against a quiet background would produce an obviously artificial signal and make most efficient use of available transmitter power.

It is interesting that the tritium line lies almost in dead centre of the traditional SETI "waterhole" region between the H and OH-spectral lines 11]. Thus, in addition to its value in a search for an artificial effusion cloud of fusion tritium, the tritium hyperfine line is virtually unique in that its detection alone is unambiguously artificial - no natural process could account for its presence. There is no possibility of confusion with H 163a and H 162a recombination lines, which lie at 1504.646 MHz and 1532.520 MHz, respectively, nor with the two 4He lines 163a and 162a at 1505.259 MHz and 1533.144 MHz.

Thus the tritium line is the ideal choice as an interstellar communication frequency from the standpoint of acquisition. It is a unique signpost to intelligence, leaping up out of the waterhole to form the "tritium waterspout." The 1516 MHz tritium hyperfine line lies well outside major radio broadcasting bands allocated to aviation communications, aeronautical and maritime satellites at 1542.5-1558.5 MHz, and space operations telemetry at 1525-1535 MHz [57]. A variety of small fixed and mobile allocations exist between 1435-1525 MHz for transmissions between fixed stations and land/coastal radar tracking systems, so care must be taken to eliminate these potentially troublesome, though obvious, sources of RFI.

Searches for purposeful narrowband CETI beacon signals .might also be conducted near the hydrogen fine-structure transitions (e.g., 1058 MHz, 3250 MHz, 9910 MHz [49]) , the hyperfine lines for triplet-state 3HeI, ground-state 3HeII (8665.649 867(10) MHz [58], metastable 2S1/2 state 3HeII (1083.354 9807(88) MHz [59]), and other hyperfine lines for 3He [60], 4He, other elements, and various neutral and ionised molecules [61]. (Hyperfine transitions of excited neutral atomic hydrogen isotopes all lie <200 MHz and thus would be difficult to observe). However, many of these transitions may occur naturally [37] and none is distinguished as especially attractive for SETI work.

2.3.3 Current Observational Status
To date only a handful of full-sky hydrogen-line surveys have been performed at various ranges of galactic latitude, and there are only a few published attempts to detect 21-cm radiation emission from stars (most of them as part of SETI programmes). Of the more than two dozen SETI searches to date [62]. none is likely to have detected an artificial hydrogen cloud. Kraus [63-64] and co-workers [65] are searching the entire sky rather than individual stars, and correct their frequency of observation to the Galactic Standard of Rest rather than the usual Local Standard of Rest. SETI observations by Wielebinskii and Seiradakis in 1977 and by Israel and Tarter in 1981 employed 4-20 MHz bandwidths, so a cloud with a 10 KHz linewidth probably could not be distinguished. Drake [66], Horowitz [67] and Tarter, [68] used narrow bandwidths of 0.015-600 Hz/channel, which also would be unlikely to have detected a cloud. The U.C. Berkeley SERENDIP programme was an all-sky parasitic search which could not easily detect hydrogen-line enhancements at individual stars [69]. Verschuur [70] and Zuckerman with Palmer [71] searched numerous stars at the proper frequencies and reasonable bandwidths but, as with many of the above studies, failed to incorporate an off-star comparison measurement thus ruling out detection of artificial clouds. Published data on three additional stars surveyed by Verschurr [70] cannot rule out the possibility of a cloud up to several flux units of intensity. Thus existing SETI searches cannot yet exclude the existence of an artificial hydrogen cloud near even the closest stars. A search for 21-cm excess would also be sensitive to the existence, but not the content, of hypothetical radio messages possibly being transmitted to us now.

Individual stars have been observed in radio frequencies, including cm-band generally [72-75], near 21-cm (e.g., 1415 MHz [76-77]), and 21-cm hydrogen line observations. As even cm-band emissions of normal stars are expected to be too weak to detect, searches have concentrated on peculiar and highly energetic stars such as early-type stars, Of stars, Wolf-Rayet stars, various emission-line and shell stars, magnetic variables, flare stars, and novae [78-79]. A very few normal single stars have been observed at cm wavelengths, with negative results except in the case of X1 Orionis, a G0 V star 10.0 pc from the Sun [80]. Of the 21-cm hydrogen-line observations, the targets have been A, B, and O stars with known or anticipated interstellar optical absorption lines [81-85], sky positions near such stars [86], a few peculiar stars such as Rho Ophiuchi [87], and general sky survey positions unrelated to individual stars for galactic HI mapping. None of these would be sensitive to artificial hydrogen clouds near Sunlike stars.

There is only one recent report of a search of individual stars at the tritium hyperfine line [88]. Although it is at, observationally convenient frequency for radio sky mapping, probably not much more information would be gained over existing 1415 MHz maps. The only survey spanning the tritium line, by Kardashev and Gindilis in 1972 covering various frequencies between 1337-1863 MHz, used an all-sky dipole antenna [62] which would not be sensitive to point sources of tritium hyperfine radiation.

2.3.4 Future Hyperfine Line Observations
If R = radius of volume containing the leakage atoms, Ds is the distance to the star, a = antenna beamwidth, and N is the number of atoms in the field of view, then for an unresolved source (a >> Rb/Ds) the brightness temperature TB of the artificial cloud is given by TB = CN/A, where the projected area A = pRb2 and C = (3hc3/32pk)(A10/w2), for w = hyperfine frequency (Hz) and h = Planck's constant [89]. A10 is the computed Einstein A transition probability, which is (fol. Field [90]) 2.869 x 10-15 sec-1 for hydrogen and 3.493 x 10-15 sec-1 for tritium. No experimental value for tritium is yet available [91]. The flux of radio energy of hyperfine wavelength l reaching the Earth is F = (2k/l2)TbW = (2k/l2) (CN/Ds2), where W = p(Rb/Ds)2. For the detection limit, taking Ds in parsecs and F in flux units, NH = 2.77 x 1045 Ds atoms H and NT = 2.27 x 1045Ds atoms T.

A civilisation which has effluxed more than DMHt ~ NHmp = (2.77 x 1045) FHDs2mp = 5 x 1020 kg (2 x 10-7 MJ) of hydrogen fusion fuel during its lifetime t could be detected at F 1 Jy sensitivity and Ds < 10 pc. Tritium decays with a half-life th (~ 12.5 yr) regardless of t, so a civilisation which leaks tritium at a rate greater than DMT ~ (2.27 x 1045)FTDs2mp(th/ln(2))-1 = 7 x 1011 kg/sec (~ 1 Lsun wastage, for eT = 0.64% c2) could be detected at F 1 Jy and Ds 10 pc. Thus the tritium limit is less restrictive for t > 22 yr, but has the advantage that any detection is unambiguously artificial.

The observational frequency in each case must be corrected to the Local Standard of Rest by compensating for Earth's rotation ( 2 KHz), Earth's orbital velocity around the Sun ( 140 KHz), and the Sun's radial velocity toward the target star ( 100 KHz). The bandwidth in searches for artificial clouds should span roughly the expected cloud thermal velocities, about 7.7 KHz for thermal line broadening at 300 K.

A tritium line search is also sensitive to SETI beacons or signals. Assuming a 15 KHz bandwidth 20 K detector with a one hour integration time, such a search could detect a 6 MW, 26-metre transmitter antenna 10 pc away pointed at Earth. Within 20 light-years of the Sun there are 86 stars, 80 of stellar class F-M, about 70 of which are visible from the northern hemisphere. There are many reasons for excluding 0, B and A stars from the search [92], such as the probable lack of planets, the severe UV environment, and the brief residence time on the main sequence with the concommitant reduced time for the emergence and evolution of life. Fifty-three of the nearest stars have now been examined for narrowband tritium line emissions, using the 26-metre radiotelescope at Hat Creek Radio Observatory in California, to a sensitivity of 1-20 Jy [88]. No detections were made.


Five other charactersitics of advanced technological civilisations may be visible across interstellar distances. Internal communications and power transmission equipment may generate radio leakage radiation upon which we may "eavesdrop." Large circurnstellar structures may produce an anomalous blackbody radiation signature. The use of fission rather than fusion nuclear fuels might give rise to anomalous solar absorption lines. Large-scale movements of hot photospheric material may produce unusual spectral line broadening and clearly artificial ghost tines. Finally, very large-scale technical activities near a star may require or establish enormous magnetic fields which may be observable via Zeeman line splitting.

3.1 Artificial Radio Spectrum

Sullivan et al. [93] performed an extensive survey of all sources of artificial radio energy leakage from Earth. An Arecibo-size antenna could detect terrestrial UHF television stations from two light-years away, and the US BMEWS military radars from 20 light-years away. A Cyclops array [1] could increase these ranges to 25 light-years and 250 light-years, respectively. Sullivan et al note that geopolitical boundaries and other information about human society can be deduced from Earth's leakage radiation, but an advanced circurnstellar civilisation would present a vastly more complex picture. However, greater efficiency as well as greater energy are available to advanced societies so it cannot be assumed that extraterrestrial technological civilisations are necessarily "noisier" than Earth.

3.2 Anomalous Blackbody Radiation

Solar optical luminosity is decreased according to the fraction f of the circurnstellar sphere blocked by optically dense orbiting material structures (i.e., a "Dyson shell"), reducing the stellar visual magnitude by -2.5 log(1-f) magnitudes. In nearby stars this produces an evidently distant but otherwise normal star with an infrared [94] and radio excess. The radio excess is difficult to detect. At 5 GHz, Arecibo could only detect fully-occulting (f = 1) artificial shells closer than 0.1 parsec. A proposed 10 km space-based radiotelescope array [95] could reach at least to 10 parsecs, and a 100 km system could reach 100 parsecs although this lies considerably beyond existing technology.

The infrared excess is easier to observe. A shell of orbiting artifacts creates a bimodal blackbody spectrum with two peaks of reciprocal amplitude, one near 500 nm (spectral class G V star) and the other near 10 microns (300 K shell of rotating bodies at 1 AU). Although this spectral signature is not unambiguously artificial, it is clearly unusual and invites further close scrutiny. The best current near-IR ground-based survey [96] at 0.8 micron could only have detected an optically dense artifact shell nearer than 0.01 pc, but the Infrared Astronomical Satellite (IRAS) permits detection [97] of fully-occulting (f = 1) Dyson shells out to 1000 pc and 1% occulting shells (f = 0.01) to 100 pc. Care must be taken to develop criteria for distinguishing artificial shells from stars such as Vega (recently discovered in IRAS data) and Be star MWC 349 [77], both of which display an infrared excess caused by natural circurnstellar material.

3.3 Fission Product Absorption Lines

Whitmire and Wright [98] suggest that extraterrestrial civilisations might use the local star as a repository for radioactive fissile waste materials, and Gray et al [99] conducted a brief search for spectral line enhancements of the expected waste elements for three stars, using an optical telescope at Kitt Peak National Observatory in Arizona. However, there are several deficiencies in this approach. For instance, considerations of convective mixing and stellar lifetimes restrict the possible candidate stars to the approximate spectral range A5-F2, yet these stars are not thought to be suitable candidates either for the formation of planets or for the natural origin and evolution of life.

The major fissionable on Earth is thorium, which is about three times as abundant as uranium and probably represents more available energy in the minerals of the Earth's crust than from both uranium and fossil fuels. The abundance of thorium in the crust is about 12 ppm, about 4 ppm for titanium, 10-6 for radium, and 10-9 for polonium and several other rare naturally-occurring isotopes. The mass of Earth's crust down to 20 km assuming a mean density 2670 kg-m-3 is 3 x 1022 kg, of which 16 ppm are fissionable or 4 x 1017 kg. Collected and burned as fission fuel, this would release about 0.1% mass energy or only 4 x 1031 joules. Ultimately, all Solar_ System fissiles could conceivably be mined and burned. Fissile abundances of Th and U in the Solar System [32] are 3.2 x 10-10 and 1.5 x 10-10, respectively, so total fissiles are 9.4 x 1020 kg and the total available energy is 8 x 1034 joules.

On the other hand, the total amount of fusionable hydrogen (the most abundant fusion fuel) is about 2 x 1020 kg on Earth, 2 x 1021 kg for all of the planets but not the Sun, and 2 x 1030 for the entire Solar System. Fusion fuel can release up to 0.92% of its mass energy, so these correspond to 2 x 1035 joules from terrestrial sources, 2 x 1042 joules from planetary sources, and 2 x 1045 joules for the entire Solar System.

So while an expanding civilisation might resort to fissionables as a last effort, clearly fusion fuels are more cost effective. They are easier to mine or extract and to transport. They are generally nonradioactive or, in the case of tritium, only weakly so, and hence may be stored more safely so far as biological beings and computers are concerned. Combustion of all fissionables in the Solar System, including those in the Sun, would release only as much energy as burning the fusion fuel on Earth alone. Fusion fuels are clearly the method of choice for artificial energy generation on an energy/kilogram basis.

Even if spaceborne fission reactors are employed, using stars as waste repositories is inadvisable for several reasons. First, if the atomic masses of the isotopes comprising the discarded matter lie much below or above 56 (Fe), nuclear binding energy is still available and is lost if not recovered. Second, the infall of matter towards a star represents a conversion of gravitational energy to kinetic energy, which is absorbed by the star and effectively lost to the orbital civilisation, along with the gravitational energy originally added to the matter to raise it from a planetary surface or atmosphere in the first place. Third, addition of foreign matter to stellar photospheres may disturb the natural ionic mix, alter flare and sunspot activity, and cause other undesirable side effects. Finally, there is little mixing between the photosphere and solar interior, so material deposited in stars remains there over long periods. If the star is later mined for its fusion energy physical resources, the discarded matter must again be raised against a strong gravity field.

3.4 Doppler and Stellar Spectral Anomalies

Large-scale technical activities may involve the removal of mass from the local star of radius Rsun to the circurnstellar shell of radius R. The movement of hot photospheric plasma from Rsun to R will produce a weak optical (wavelength l) Doppler component of order Dl = lDv/c. Dv = (2GMsun(Rsun-1R-1))1/2 = 617 km/sec, so at 500 nm, Dl ~ 1 nm. This is a potentially larger effect than optical spectrum line broadening due to thermal velocities and photospheric turbulence (>0.01 nm), stellar rotation (>0.4 nm), or interstellar radial motion (>0.1 nm in the solar neighbourhood), and can easily be distinguished from double-line spectroscopic binaries.

The exact appearance of the artificial anomaly depends upon unknown stellar mass extraction trajectories. A single locus for removal and collection would give rise to single anomalous ghost lines shifted 1 nm. Several loci would make multiple ghost lines. Co-rotating loci would produce time-variable anomalies. Non-localised loci would result in large, probably asymmetrical, line broadening. These lines could be quite bright. If the natural energy available to the extraterrestrial civilisation ~ Lsun, then the total plasma mass in transit from Rsun to R at any given moment is of order Mp ~ 2LsunR/Dv3 = 5 x 1020 kg, about the mass of the solar photosphere. In addition to direct searches for ghost lines, observed line anomalies in stars classified as P-Cygni, Wolf Rayet, T Tauri, and Ap magnetic and spectrum-variable should be re-examined for cases of possible spectral misclassification. Finally, any A0-A9 star showing both strong neutral H and neutral metal lines could be a misclassified late-type star surrounded by an artificial hydrogen cloud.

3.5 Anomalous Magnetic Fields

Technical activities in an artificial circurnstellar shell of radius R could conceivably involve current loops of order R giving rise to a free space magnetic induction of order B = moi/R and magnetic flux FB = 4pBR2 over the spherical habitat shell of energy EB = FBi = (4p/mo)R3B2, where mo = permeability constant. The highest natural fields found in main sequence stars are 0.1-1 tesla, so a B = 10 tesla field would be clearly artificial and for R = 1 AU represents EB = 3 x 1042 joules. To establish this field would require burning ~ I Mj as fusion fuel, which seems excessive. Artificial fields B < 103 gauss, while observable via Zeeman line splitting (about 10-4 nm) in optical spectra, would not be unambiguously artificial when detected near late main sequence stars.


The most observable characteristic of an advanced extraterrestrial civilisation which makes extensive use of the fusion fuel resources of its local star and planets is an effusion cloud of molecular fuel elements. This cloud rapidly dissociates and neutralises to the atomic ground state. Optical anomalies would be difficult to observe, but the detection of both the hydrogen and tritium hyperfine transition radio lines is relatively straightforward. Existing searches at these lines are not sensitive to artificial fusion fuel clouds. In addition, the low natural abundance of neutral atomic ground-state tritium suggests that its hyperfine line, centred in the radio-SETI waterhole band, has minimum background noise and thus is ideal for interstellar communication and future SETI searches. With one exception [88], no SETI searches at the tritium line towards individual stars have been reported to date. Other observables of advanced civilisations, including redshifted neutrino point sources, an artificial radio spectrum, anomalous blackbody radiation, fission product absorption lines, Doppler and stellar spectral anomalies, and extraordinary magnetic fields, might also serve as the basis for future SETI research but are considerably more challenging observationally.


The author thanks Francisco Valdes for contributions and comments on an earlier version of the manuscript. This research was supported by the Xenology Research institute.


  1. Billingham, J.and Oliver B., Project Cyclops: A Design Study of a System for Detecting Extraterrestrial Intelligent Life. Revised Edition, NASA CR-1 14445 (1973).
  2. Morrison, P., Billingham, J.and Wolfe, J., eds., The Search for Extraterrestrial Intelligence, SETI, NASA SP-419 (1977).
  3. Billingham, J.and Pesek, R., eds., Communication with Extraterrestrial Intelligence, Pergamon Press, Oxford (1979).
  4. Freitas, R. A. Jr., "Extraterrestrial Intelligence in the Solar System: Resolving the Fermi Paradox," JBIS, 36, 496-500 (1983).
  5. Freitas, R. A. Jr., "Terraforming Mars and Venus Using Machine Self-Replicating Systems (SRS)," JBIS, 36, 139-152 (1983).
  6. Turan, J. W., "Theories of Nucicosynthcsis," in A. G. W. Cameron, ed., Coosmochemistry, D. Reidel, DordrechtHolland, pp. 23-49 (1973).
  7. Martin, A. R., ed., "Project Daedalus - Final Report," JBIS Supplement (1978).
  8. Powell, C., "Interstellar Flight and Intelligence in the Universe," Spaceflight, 14, 442-447 (1972).
  9. Glasstone, S. and Lovbcrg, R. H., Controlled Thermonuclear Reactions, Van Nostrand Reinhold, Scarborough, Canada (1960).
  10. Petrie, T. W. R., "A Review of Possible Fusion Fuels," in 'G. H. Miley, ed., Advanced Energy Conversion for Fusion Reactors, University of Illinois Report No. C00-2218-18, pp. A I -A 17 (1974).
  11. "Starfire - A Commercial Tokamak Fusion Power Plant Study," Argonne National Laboratory, ANL/FPP-80-1, September 1980.
  12. Saenz, A. W., Uberhall, H., Kelly, F. J., Padgett, D. W. and Seeman, N., "Telecommunication with Neutrino Beams," Science, 199, 295-297 (1977).
  13. Roberts, A., "DUMAND: The Ocean as a Neutrino Detector," in H. Faissner, H. Reithler, P. Zerwas, eds., Proceedings of the International Neutrino Conference 1976, Friedr. Vieweg & Sohn Verlagsgesellschaft MBH, Braunschweig, pp. 688-702 (1977).
  14. Hollenbach, D. J., Werner, M. W. and Salpeter, E. E., "Molecular Hydrogen in HI Regions," Ap. J., 163,165-180 (1971).
  15. O'Donnell, E. J. and Watson, W. D., "Upper Limits to the Flux of Cosmic and X-Rays in Interstellar Clouds," Ap. J., 191, 89-92 (1974).
  16. Jura, M., "Formation and Destruction Rates of Interstellar H2," Ap. J., 191, 375-379 (1974).
  17. Watson, W. D., "Formation and Destruction of Interstellar Molecules," in Roger Balian, Pierre Encrenaz, James Lequeux, eds., Atomic and Molecular Physics and the Interstellar Matter, American Elsevier Pub. Co., New York (1975).
  18. Dyson, F. J., "The Search for Extraterrestrial Technology," in R. E. Marshak, ed., Perspectives in Modern Physics, John Wiley & Sons, New York, pp. 641-655 (1966).
  19. Kearns, J. J., "Diffusion Coefficient of Hydrogen in Alpha Zirconium, Zircaloy-2 and Zircaloy-4," J. Nucl. Materials, 43, 330-338 (1972).
  20. Austin, J. H. and Elleman, T. S., "Tritium Diffusion in 304- and 316-Stainless Steels in the Temperature Range 25 to 222 oC," J. Nucl. Materials, 43, 119-125 (1972).
  21. Shlosman, I., Koziovsky, B. Z. and Shaviv, G., "Dwarf-M Stars as a Source of He3 in the Interstellar Medium," Astron. Astrophys., 73, 358-360 (1969).
  22. Lai, D.. Libby, W. F., Wetherill, G., Leventhal, J. and Alton, G. D., "Collecting a Sample of Solar Wind: An Experimental Study of Its Capture in Metal Films," J. Appl. Phys., 40, 3257-3267 (1969).
  23. Ramaty, R. and Kozlovsky, B., "Deuterium, Tritium, and Helium-3 Production in Solar Flares," Ap. J., 193, 729-740 (1974).
  24. Fireman, E. L., "Solar-Wind Tritium Limit and the Mixing Rate of the Solar Atmosphere," Ap. J., 205, 268-272 (1976).
  25. Fireman, E. L., DeFelice, J. and D'Amico, J., "The Abundances of 3H and 14C in the Solar Wind," Earth Planet Sci. Lett., 32, 185-190.
  26. Cujec, B. and Fowler, W. A.. "Neglect of D. T. and 'He in Advanced Stellar Evolution.--- Ap. J. 236. 658-660 (1980).
  27. Dupree, A. K., Baliunas, S. L. and Shipman. H. L., "Deuterium and Hydrogen in the Local Interstellar Medium," Ap. J., 219, 361-369 (1977).
  28. Vidal-Madjar, A., Laurent, C., Bruston, P. and Audouze, J., "Is the Solar System Entering a Nearby Interstellar Cloud?" Ap. J., 223, 589-600 (1978).
  29. McClintock, W., Henry, R. C., Linsky, J. L. and Moos, H. W., "Ultraviolet Observations of Cool Stars. VIL Local Interstellar Hydrogen and Deuterium Lyman-Alpha," Ap. J., 225, 465-481 (1978).
  30. Adams, T. F. and Frisch, P. C., "High Resolution Observations of the Lyman Alpha Sky Background," Ap. J., 212, 300-308 (1977).
  31. Ajello, J. M., "An Interpretation of Mariner 10 Helium (584 A) and Hydrogen (1216 A) Interplanetary Emission Observations," Ap. J., 222, 1068-1079 (1978).
  32. Cameron, A. G. W., "Abundances of the Elements in the Solar System," Space Sci. Rev., 15, 121-146 (1973).
  33. Reeves, H., "On the Origin of the Light Elements," Ann. Rev. Astron. Ap., 13, 437-469 (1974).
  34. Pasachoff, J. M. and Cesarsky, D. A., "Further Observations at the Interstellar Deuterium Frequency," Ap. J., 193, 65-67 (1974).
  35. York, D. G. and Rogerson, J. B. Jr., "The Abundance of Deuterium Relative to Hydrogen in Interstellar Space," Ap. J., 203, 378-385 (1976).
  36. Laurent, C., Vidal-Madjar, A. and York, D. G., "The Ratio of Deuterium to Hydrogen in Interstellar Space. IV. The Lines of Sight to Delta, Epsilon, and Iota Orionis," Ap. J.. 229, 923-941 (1979).
  37. Rood, R. T., Wilson, T. L. and Steigman, G., "The Probable Detection of Interstellar 3He+ and its Significance," Ap. J.. 227. L97-L101 (1979).
  38. Rogerson, J. B. Jr. and York, D. G., "Interstellar Deuterium Abundance in the Direction of Beta Centauri," Ap. J., 186, L95-L98 (1973).
  39. Peimbert, M. and Wallerstein, G., "A Search for Deuterium in Stellar Spectra 11. Normal Stars of Types B. A, and F," Ap. J., 142, 1024-1027.
  40. Beckers, J., "A Search for Deuterium on the Sun," Ap. J., 195, L43-L45 (1975).
  41. Krat, V. A. and Pravdjuk, L. M., "On the Abundance of Deuterium in the Solar Atmosphere," Solar Phys., 61, 279-281 (1979).
  42. Sargent, W. L. W. and Jugaku, J., Ap. J., 134, 777 (196 1).
  43. Sargent, A. I., Greenstein, J. L. and Sargent, W. L. W., "A Search for Neon in the Spectra of Peculiar A and B Stars." Ap. J., 157, 757-768 (1969).
  44. Dworetsky, M. M., "Helium-3 in Iota Orionis B," Ap. J., 184, L75-L78 (1973).
  45. Hilditch, R. W. and Hill, P. W., "Identification of He3 in the Binary System 68u Herculis," Mon. Not. R. Asir. Soc., 172, 29P-34P (1975).
  46. Hall, D. N. B., "Spectroscopic Detection of Solar 3He," Ap. J., 197, 509-512 (1975).
  47. Kardashev, N. S., "The Possibility of Detection of Allowed Lines of Atomic Hydrogen in the Radiofrequency Spectrum," Soviet Astr. AJ., 3, 813-820 (1960).
  48. Palmer, P., Zuckerman, B., Penfield. H. and Lillcy, A. E., "Determinations of Helium Abundance from Radiofrequency Recombination Lines," Ap. J., 156, 887-901 (1969).
  49. Barrett, A. H., "Spectral Lines in Radio Astronomy," Proc. IRE, 46, 250-259 (1958).
  50. White, J. A., Chow, L. Y., Drake, C. and Hughes, V. W., "Hyperfine Structure of the Metastable Triplet State of Helium Three," Phys. Rev. Left., 3, 428-429 (1959).
  51. Wineland, D. J. and Ramsey, N. F., "Atomic Deuterium Maser," Phys. Rev., A5, 821-837 (1972).
  52. Weinreb, S., "A New Upper Limit to the Galactic Deuteriumto-Hydrogen Ratio," Nature, 195, 367-368 (1962).
  53. Cesarsky, D. A., Moffet, A. T. and Pasachoff, J. M., "327-MHz Observations of the Galactic Center: Possible Detection of a Deuterium Absorption Line," Ap. J., 180, L1-L6 (1973).
  54. Sarma, N. V. G. and Mohanty, D. K., "Search for the Deuterium'Absorption Line at 327 MHz in the Direction of the Galactic Center," Mon. Not. R. Astr. Soc., 194,181-184 (1978).
  55. Hellwig, H., Vessot, R. F. C., Levine, M. W., Zitewitz, P. W.. Allen, D. W. and Glaze, D. J., "Measurement of the Unperturbed Hydrogen Hyperfine Transition Frequency," IEEE Trans. Instrum. Meas. IM-19, 200-209 (1970).
  56. Mathur, B. S., Crampton, S. B., Kleppner, D. and Ramsey, N. F., "Hyperfine Separation of Tritium," Phys. Rev., 158, 14-17 (1967).
  57. Reference Data for Radio Engineers, Internati. Tel. & Tel. Corp., New York (1975).
  58. Schuessler, H. A, Fortson, E. N. and Dehmelt, H. G., "Hyperfine Structure of the Ground State of 3He+ by the IonStorage Exchange-Collision Technique," Phys. Rev., 197, 5-38 (1969).
  59. Prior, M. H. and Wang, E. C., "Hyperfine Structure of the 2s State of 'He+," Phys. Rev. A, 16, 6-18 (1977).
  60. Lawler, J. E., Ferguson, A. I., Goldsmith, J. E. M., Jackson, D. J. and Schawlow, A. L., "Doppler-Free Intermodulated Optogalvanic Spectroscopy," Phys. Rev. Lett., 42,1046-1049 (1979).
  61. Blicharski, J. S., "On Radiofrequency Transitions and Nuclear and Electron Polarization for Hydrogen Molecular Ions H2+," Can. J. Phys., 46, 823-827 (1968).
  62. Tarter, J., "Searching for THEM: Interstellar Communications," Astronomy, 10, 6-22 (1982).
  63. Kraus, J., Big Ear, Cygnus-Quasar Books, Powell, Ohio, (1976).
  64. Kraus, J., "We Wait and Wonder," Cosmic Search, 1, 3, 3 1 31-34 (1979).
  65. Dixon, R. S. and Cole, D. M., "A Modest All-Sky Search for Narrowband Radio Radiation Near the 21-cm Hydrogen Line," Icarus, 30, 267-273 (1976).
  66. Drake, F. D., "Project OZMA," Phys. Today, 14, (4) 40-46 (1961).
  67. Horowitz, P., "A Search for Ultra-Narrowband Signals of Extraterrestrial Origin," Science, 201, 733-735 (1978).
  68. Tarter, J., "An American View of SETI," Spaceflight, 21, 57-58 (1979).
  69. Bowyer, S., Zeitlin, G., Tarter, J., Lampton, M. and Welch, W. J., "The Berkeley Parasitic SETI Program." Icarus, 53, 147-155 (1983).
  70. Verschuur, G. L., "A Search for Narrow Band 21-cm Wavelength Signals from Ten Nearby Stars," Icarus, 19, 329-340 (1973).
  71. Zuckerman, B., "Searches for Electrmagnetic Signals from Extraterrestrial Beings," in Michael H. Hart, Ben Zuckerman, eds., Extraterrestrials.- Where Are They? Pergamon Press, New York, pp. 8-18 (1982).
  72. Oster, L., "Radio Emission from Stellar and Circurnstellar Atmospheres." In B. Baschek, W. H. Kegel, G. Traving, eds., Problems in Stellar Atmospheres and Envelopes, SpringerVerlag, New York, pp. 301-323 (1975).
  73. Woodsworth, A. W. and Hughes, V. A., "Observations of Radio Stars at 10.6 GHz," Astron. Astrophys., 59, 105-111 (1977).
  74. Haisch, B. M., Linsky, J. L., Slee, 0. B., Hearn, D.. R. Walker, A. R., Rydgren, A. E. and Nicholson, G. D., "A Coordinated X-Ray, Optical, and Microwave Study of the Flare Star Proxima Centauri," Ap. J., 225, L35-L37 (1978).
  75. Feldman, P. A. and Kwok, S., J.R.A.S., 73, 271 (1979).
  76. Braes, L. L. E. and Miley, G. K., "Radio Detection of Cygnus X-3," Nature, 237, 506 (1972).
  77. Braes, L. L. E., Habing, H. J. and Schoenmaker, A. A., "MWC 349, a New Radio Star," Nature, 240, 230 (1972).
  78. Braes, L. L. E., "Radio Continuum Observations of Stellar Sources," in F. J. Kerr, S. C. Simonson, III, eds., Galactic Radio Astronomy, D. Reidel Publ. Co., Dordrecht-Holland, pp. 377-381 (1974).
  79. Altenhoff, W. J., Braes, L. L. E., Olnon, F. M. and Wendker, H. J., "A Search for Radio Emission from Stars," Astron. Astrophys., 46,11-17 (1976).
  80. Gary, D. E. and Linsky, J. L., "First Detection of Nonflare Microwave Emission from the Coronae of Single Late-Type Dwarf Stars," Ap. J., 250, 284-292 (1981).
  81. Lawrence, R. S., "Radio Observations of Interstellar Neutral Hydrogen Clouds," Ap. J., 123, 30-33 (1956).
  82. Howard, W. E. III, Wentzel, D. G. and McGee, R. X., "On a Correlation Between the Radial Velocities of Optical and Radio Interstellar Lines," Ap. J., 139, 988-1001 (1963).
  83. Van Woerden, H., I.A.U. Symp. No. 31, p. 3 (1967).
  84. Habing, H. J., "An Atlas of 21-cm Line Profiles in the Directions of Stars with Interstellar Absorption Lines," Bull, Astr. Inst. Netherlands, 20, 120-128 (1968).
  85. Goldstein, S. J. Jr. and MacDonald, D. D., "Observations of the 21-cm Hydrogen Line Toward High-Latitude Stars," Ap. J., 157, 1101-1118 (1969).
  86. Takakubo, K., Bull. Astr. Inst. Netherlands, 19, 125 (1967).
  87. Meszaros, P., "Dust and Atomic Hydrogen near Rho Opbiuchi," Astrophys. Space Sci., 2, 510-519 (1968).
  88. Valdes, F. and Freitas, R. A. Jr., "A Search for the Tritium Hyperfine Line from Nearby Stars," paper presented at the 13th IAA International Review Meeting on Communication with Extraterrestrial Intelligence (CETI), 35th Annual Congress of the IAF, Palais de Beaulieu Lausanne, 8-13 October 1984. [Also]
  89. Pacholczyk, A. G., Radio Astrophysics. W. H. Freeman, San Francisco, (1970).
  90. Field, G. B., "Excitation of the Hydrogen 21-cm Line," Proc. IRE, 46, 240-250 (1958).
  91. Fuhr, J. R., Personal communication, U.S. Dept. of Commerce, National Bureau of Standards, Data Center on Atomic Transition Probabilities, 18 November 1982.
  92. Dole, S. H., Habitable Planets for Man, Blaisdell Pub. Co., New York (1964).
  93. Sullivan, W. T. III, Brown, S. and Wetherill, C., "Eavesdropping: The Radio Signature of the Earth," Science, 199, 377-388 (1978).
  94. Dyson, F. J., "Search for Artificial Stellar Sources of Infrared Radiation," Science, 131, 1667-1668 (1959).
  95. Buyakas, V. I. et al., "An-Infinitely Expandable Space Radiotelescope," in J. Billingham, R. Pesek, eds., Communication with Extraterrestrial Intelligence, Pergamon Press, New York, pp. 175-201 (1979).
  96. Craine, E. R., "Optical Infrared Sky Survey," Astron. J., 83, 1598-1606 (1978).
  97. Reynolds, R. T., Tarter, J. C. and Walker, R. G., "A Proposed Search of the Solar Neighborhood for Substellar Objects," Icarus, 44, 772-779 (1980).
  98. Whitmire, D. P. and Wright, D. P., "Nuclear Waste Spectrum as Evidence of Technological Extraterrestrial Civilization," Icarus, 42,149-156 (1980).
  99. Gray, R. H., Dixon, R. S., Ehman, J. P. and Talent, D. L., "21-cm Radio Emissions with Geometric Fine Structure," preprint, 1982.

Editors Notes:

  1. Astrophagy is also mentioned in should be "The Search for Extraterrestrial Artifacts (SETA)," JBIS 36(November, 1983):501-506.

Creation date: July 19, 1998
Last Modified: January 1, 2001
HTML Editor: Robert J. Bradbury