A Dyson Sphere makes an extraordinary setting for science fiction. In fact, my first knowledge of the concept came from reading Larry Niven’s 1970 novel Ringworld, a book that left such an impression that I still recall reading half of it at a sitting in the drafty little parlor of a house I was renting in Grinnell, Iowa. Ringworld had just come out as a Ballantine paperback with the lovely cover you see below. I was hooked after about three pages and read deep into a night filled with wind and snow.

It could be argued, of course, that a ring made out of planetary material, a habitat so vast that it completely encircles its star, is actually one of the smaller Dyson concepts. It was in 1960 that Freeman Dyson suggested how a civilization advanced to the point of such astro-engineering might use everything it found in its solar system to create a cloud of objects, a swarm that would make the most efficient use of its primary’s light. And as you keep adding objects, you point to the ultimate outcome, a Dyson Sphere that completely envelopes the star from which it draws its energy.

A Dyson Sphere Search with IRAS

Last April I looked at Dyson spheres in the context of an article by Bruce Dorminey that considered new SETI strategies. Now I see that Richard Carrigan, a retired physicist from Fermilab, has added a new paper to the arXiv site, one that discusses the work reviewed in that earlier story. Carrigan has been examining sources identified by the Infrared Astronomical Satellite (IRAS), the idea being to look for objects that seem to be radiating waste heat in such a way that they might be Dyson Spheres of one kind or another. A fully enveloped star won’t be visible to the eye, but Carrigan’s infrared search covers the blackbody temperature region from 100 to 600 degrees Kelvin for full or partial Spheres.

The data come from an IRAS database that covers 96 percent of the sky and includes some 250,000 sources. Exciting stuff on the face of it, because unlike a conventional SETI search, a hunt for Dyson Spheres involves no necessary intent to communicate on the part of the civilization in question. And when you’re dealing with SETI, the fewer preconceptions you bring to the dance, the better. Here’s the thinking behind Carrigan’s attempt:

For a Dyson Sphere the stellar energy from the star would be reradiated at a lower temperature. If the visible light was totally absorbed by a thin “shell” a pure Dyson Sphere signature would be an infrared object with luminosity equivalent to the invisible star and a Planck or blackbody distribution with a temperature corresponding to the radius of the spherical shell formed by the cloud of objects. For a sun-like star with the shell at the radius of the Earth the temperature would be approximately 300º K.

Sorting the Evidence

A distinct signature? You would hope so, and if that is the case, we can dig through our data practicing what Carrigan delightfully calls ‘cosmic archaeology,’ using data that cover the 8 to 100 micron infrared range needed to study a Dyson Sphere’s emissions under these assumptions. Yet an identification runs into immediate problems, not the least of which is the need to differentiate any candidate from natural sources that show much the same signature. A cocoon of gas and dust around a young star, for example, might mimic an artificial source.

Carrigan goes through the possibilities — protostars, planetary nebulae, dying stars — and weighs their telltale infrared identity against a true Dyson Sphere, with notes on how to tell the natural from the potentially artificial. Here he considers the methods (italics mine):

A Dyson Sphere candidate with a blackbody distribution can have several characteristics such as a blackbody temperature, the distance from our Sun, magnitude in the infrared, and variability. It may also have a stellar signature in the visible or infrared. Slysh (1985) notes, “The confusion between red giants with thick circumstellar envelopes and possible Dyson Spheres in the IRAS survey is a serious problem, and to differentiate the two we need additional data.” …[S]ome of the source types discussed above populate the same region of an infrared color-color plot as a Dyson sphere candidate would. Non-Dyson Sphere objects can be eliminated using discriminants like spectral lines in the infrared or radio regime, implausible blackbody temperatures, established classifications, and statistical departures from a blackbody distribution.

A Dwindling List

So we still have a chance to find a true Dyson Sphere, assuming one or more are out there. If I had more money to burn, I would ring up Tibor Pacher with an offer to make another bet, this one saying that no Dyson Sphere will be found in this century. Tibor is bound to take that one, but I’ve lost several other bets recently and had better put down my cards (our other bet, on the date of the first true interstellar mission, is viewable at the Long Bets site; feel free to comment on either side of that one).

Image: A Dyson Sphere as envisioned by the producers of Star Trek: The Next Generation, from the episode “Relics.” Credit: Paramount Pictures.

Bet or no, the process of working through the database is fascinating, but the list of candidates quickly dwindles in Carrigan’s discussion. We wind up with a scant seventeen possibilities, none of them particularly promising, though worthy of further study. Carrigan comments:

This search has shown that at best there are only a few quasi-plausible Dyson Sphere signatures out of the IRAS LRS sample in the 100 < T < 600 ºK temperature region. This limit includes both pure and partial Dyson Spheres. With several possible exceptions all the "good" sources identified in this search have some more conventional explanation other than as a Dyson Sphere candidate. In spite of the fact that there are many mimics such as stars in a late dusty phase of their evolution good Dyson Sphere candidates are quite rare!

Next Steps

Where do we go from here? Compiling more on the list of seventeen Dyson candidates would be the logical next step (Carrigan discusses how). And we can search further using the more powerful Spitzer Space Telescope, an instrument with greater angular resolution than IRAS and three orders of magnitude better sensitivity in the infrared ranges needed for this work. This would extend the survey out past the center of the galaxy, but we’ll lose some of the IRAS sources, which are too bright for Spitzer’s camera to avoid saturation. And only one of the seventeen candidates Carrigan finds would be covered by such a Spitzer study.

This intriguing work reminds us how early we are in the study of Dyson Spheres, and the broader attempt to identify astro-engineering on this vast scale. The Low Resolution Spectrometer aboard the IRAS satellite was only sensitive enough to track solar-sized Dyson Spheres out to a range of some 300 parsecs, which includes a million solar-type stars. Extending that reach, and finding ways to either rule out or strengthen the case for some of Carrigan’s seventeen candidates, is work that extends our existing radio and optical SETI methods. Beefing up our infrared tools will help us determine whether a concept once considered outrageous might conceivably flag an extraterrestrial presence.

The paper is Carrigan, “IRAS-based Whole-Sky Upper Limit on Dyson Spheres,” available online.