The most promising candidates for Dyson spheres have lost that status. The excess infrared radiation that had been attributed to hypothetical megastructures around red dwarfs actually comes from galaxies located far behind those stars. The two sources could only be separated thanks to high angular resolution. Once again, the search for technological signatures of extraterrestrial civilizations has come up empty.
Heat as a Sign of Technology
The Dyson sphere concept envisions a structure surrounding a star that intercepts its radiation for the needs of an advanced civilization. Whatever the purpose of collecting that energy, it would ultimately be converted into heat and radiated outward at long wavelengths.
This reduces the search to one simple criterion: astronomers need to find a star that is unusually faint in visible light while at the same time unusually bright in the mid-infrared.
Candidates from a Sample of Millions
The Hephaistos project team, led by Uppsala University, examined about five million stars using data from Gaia, 2MASS, and WISE. Among them, a small group of M dwarfs remained with noticeable excess emission at 12 and 22 micrometers, as Universe Today reports.
Two of them, designated D and E, were observed last summer and autumn with the James Webb Space Telescope. Images taken through three filters of the Mid-Infrared Instrument (MIRI) show a separate source next to each dwarf, about one arcsecond away.
That angular separation is equivalent to the apparent width of a 0.1 mm hair viewed from a distance of twenty meters. In WISE data, the two objects blur together into a single spot at this scale, while James Webb resolves them completely.
The Nature of the Background Sources
The spectra made it possible to determine the distances to the neighboring sources of both dwarfs. The galaxy near candidate D is at a redshift of 0.922, while the companion source near candidate E is at 0.410.
The first appears as a point-like source with an extremely red spectrum, consistent with a hot, dust-obscured galaxy known as a Hot DOG — essentially a quasar hidden behind a thick layer of dust. The second turned out to be extended, with bright clumps and strong polycyclic aromatic hydrocarbon bands characteristic of intense star formation.
Why the Older Data Were Misleading
WISE has a resolution of about six and twelve arcseconds in its two longest-wavelength bands. At that level of blurring, photons from the star and the background galaxy fall into the same spot, making it impossible to separate their contributions.
An indirect clue to the confusion was already present in the archival data. The center of brightness at 12 micrometers is shifted away from the dwarf toward the galaxy by 0.75 arcseconds for candidate D and 1.50 arcseconds for candidate E.
The contribution of the star itself to the total flux falls extremely quickly. At 5.6 micrometers, the dwarf contributes about 80% of the light, while at 15 micrometers only about 2% remains for candidate D. The results are described in a preprint on arXiv, whose conclusions have not yet undergone formal peer review.
What Comes Next
The authors acknowledge that a substantial part of the entire project candidate list is likely contaminated by the same kind of chance background neighbors. The two stars that were checked show just how misleading earlier sky-survey data can be.
Hope is now being placed on infrared surveys with better spatial resolution over large areas, which are expected to begin operating in the 2030s. That should make it possible to filter out background galaxies already at the stage of compiling the candidate list.
