🔭 astrophysics

Project Hephaistos -- IV. James Webb Space Telescope Observations of Two Dyson Sphere Candidates

JWST observations of two M-dwarf stars previously flagged as Dyson sphere candidates reveal that their mid-infrared excesses are actually caused by background galaxies at redshifts of approximately 0.9 and 0.4, rather than by artificial megastructures.

Original authors: Erik Zackrisson, Arjan Bik, Anita Ali Asgar, Olivia Curtis, Jason T. Wright, Tongtian Ren, Roberto J. Assef, Andrew Blain, Alexis Brandeker, Michael A. Garrett, Uma Gorti, Andreas J. Korn, Priyatam K.

Published 2026-07-13

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Erik Zackrisson, Arjan Bik, Anita Ali Asgar, Olivia Curtis, Jason T. Wright, Tongtian Ren, Roberto J. Assef, Andrew Blain, Alexis Brandeker, Michael A. Garrett, Uma Gorti, Andreas J. Korn, Priyatam K. Mahto, Armin Nabizadeh

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe is a giant, cosmic game of "Where's Waldo?" but instead of a red-and-white striped shirt, we are looking for a specific kind of cosmic magic: a Dyson Sphere.

A Dyson Sphere is a hypothetical megastructure built by an advanced alien civilization to wrap around a star and harvest all its energy. If such a thing existed, it would act like a giant, warm blanket around the star, glowing brightly in infrared light (heat) while blocking the star's visible light. For years, astronomers have been scanning the sky, looking for stars that seem to be wearing these invisible, heat-glowing blankets.

In a previous study, a team called "Project Hephaistos" found two M-dwarf stars (small, cool stars) that seemed to be wearing these blankets. They were glowing way too much in the mid-infrared wavelengths (specifically at 12 and 22 micrometers) to be normal stars. The team called them Candidate D and Candidate E. It looked like a potential first contact!

But then, the James Webb Space Telescope (JWST) arrived with its super-sharp eyes, and the plot twist happened.

The Great Cosmic Mix-Up

The paper you are reading is the "reveal" episode. The authors used JWST's Mid-Infrared Instrument (MIRI) to take a closer look at Candidates D and E. The result? There are no alien megastructures.

Instead of finding a star wrapped in a sci-fi energy net, the telescope found something much more mundane, yet equally fascinating: cosmic camouflage.

The "excess heat" wasn't coming from the stars at all. It was coming from background galaxies that were accidentally lined up right behind the stars from our point of view. It's like looking at a streetlamp at night and seeing a bright billboard directly behind it. If your eyes aren't sharp enough, the billboard's light blends with the streetlamp, making the lamp look impossibly bright.

The WISE telescope, which found these candidates originally, has a "blurry" vision (its resolution is about 6 to 12 arcseconds wide). It couldn't tell the difference between the star and the galaxy behind it. But JWST is like a high-definition camera that can zoom in and separate the two.

What JWST Actually Saw

When the team looked at Candidate D, they saw the small M-dwarf star sitting about 1 arcsecond away from a distant galaxy.

The Star: It's a normal, boring M-dwarf. It's not glowing with extra heat.
The Galaxy: It's a "Hot Dust Obscured Galaxy" (Hot DOG). This is a rare type of galaxy that is shrouded in hot dust and glows fiercely in the infrared. It's located at a redshift of 𝑧≈0.9, which means it's very far away.
The Verdict: The "alien blanket" was just a distant, dusty galaxy peeking out from behind the star.

When they looked at Candidate E, the situation was even more blended.

The Star: Again, a normal M-dwarf.
The Galaxy: This one is at a redshift of 𝑧≈0.4 and looks like a messy, clumpy "dusty starburst" galaxy. It's so bright and extended that in some infrared images, the galaxy's light completely swallows the star, making it look like the star itself is the source of the glow.
The Verdict: Just like Candidate D, the "alien heat" is actually a distant, chaotic galaxy.

Ruling Out the "Alien" and "Debris" Theories

The paper is very clear about what these stars are not.

No Dyson Spheres: If an alien structure were wrapping these stars, it would be right next to them, within a distance of less than 1 AU (the distance from Earth to the Sun). At the distance of these stars (over 200 parsecs away), such a structure would be too tiny to be seen as anything other than a single point of light, indistinguishable from the star itself. But JWST clearly saw a second object (the galaxy) sitting about 1 arcsecond away. The "blanket" isn't there; the "neighbor" is.
No Extreme Debris Disks: Could it be a disk of rocky debris from a giant collision? The authors say no. While some stars have these disks, the specific amount of heat seen here, combined with the fact that the stars look perfectly normal in visible light (no signs of being young or violent), doesn't fit the profile of a debris disk. Plus, the heat is coming from a different location entirely.

The "Hot DOG" Mystery

The paper suggests that the background galaxies are likely a specific type of object called a Hot Dust Obscured Galaxy (Hot DOG).

Candidate D's galaxy looks like a point source (a dot) and has a spectrum that fits the Hot DOG profile perfectly. It's likely powered by a supermassive black hole (an AGN) eating dust.
Candidate E's galaxy is a bit different. It looks like a dusty, clumpy starburst galaxy. While it has some features of a Hot DOG, it might just be a regular galaxy with a lot of star formation and dust.

The authors are careful to say that while the data suggests these are Hot DOGs, they are fainter than the typical Hot DOGs we usually find. They are like the "quiet cousins" of the famous Hot DOGs, hiding in plain sight because they are so close to a star in our sky.

Why This Matters (And What It Means for Aliens)

This discovery is a bit of a bummer for alien hunters, but a huge win for astronomers.

The "False Alarm" Problem: The paper explains that the WISE telescope's blurry vision makes it easy to mistake a distant, bright galaxy for a weird star. The authors calculate that if you look at millions of stars, you might expect to find a few of these "cosmic mix-ups" just by chance. In fact, they suspect that many of the original candidates from the Project Hephaistos list might be these same kind of mix-ups.
The Future: To find real Dyson spheres (if they exist), we need to be much smarter about checking for "background noise." We can't just look at the heat; we have to check if the heat is coming from the star or a galaxy hiding behind it. The paper suggests that looking for "centroid shifts" (where the center of the light moves as you change the color of the filter) is a good way to spot these imposters before we waste time on them.

The Bottom Line

The paper concludes that Candidates D and E are not Dyson spheres. They are normal stars that got a bad reputation because a distant, glowing galaxy was standing right behind them.

The universe is full of strange, hot, dusty galaxies that look like alien technology when viewed through a blurry lens. But with JWST's sharp eyes, we can finally see the truth: it's just a cosmic game of hide-and-seek, and the "aliens" were just a very far-away galaxy all along. The search for Dyson spheres continues, but these two suspects have been cleared.

Technical Summary: Project Hephaistos – IV. James Webb Space Telescope Observations of Two Dyson Sphere Candidates

Problem Statement
Project Hephaistos previously identified a sample of M-dwarf stars exhibiting significant mid-infrared (IR) excess in WISE W3 (12 $\mu$ m) and W4 (22 $\mu$ m) bands, initially flagged as potential Dyson sphere candidates due to the absence of stellar youth indicators (Suazo et al. 2024). While Dyson spheres would emit "waste heat" as IR excess, other astrophysical phenomena, such as debris disks or background interlopers, can mimic this signature. Previous analyses suggested that source confusion from background galaxies might explain some candidates, but the low angular resolution of WISE (FWHM $\approx$ 6–12 arcsec) prevented definitive confirmation. This paper addresses the origin of the IR excess for two specific candidates (D and E) using high-resolution data to distinguish between intrinsic stellar phenomena (Dyson spheres or debris disks) and extrinsic contamination.

Methodology
The authors utilized the James Webb Space Telescope (JWST) Mid-Infrared Instrument (MIRI) to observe candidates D and E as part of Cycle 4 programme GO 7199. The observational strategy included:

Imaging: MIRI imaging in filters F560W, F1000W, and F1500W to resolve the spatial distribution of flux at $\approx$ 5.6, 10, and 15 $\mu$ m.
Spectroscopy: Medium-Resolution Spectroscopy (MRS) covering 4.9–24 $\mu$ m to identify spectral features and determine redshifts.
Data Processing: Standard JWST calibration pipelines (v1.20.2 for imaging, v1.18.0 for spectroscopy) were applied. The authors employed Bayesian nested sampling (dynesty) to model the Point Spread Function (PSF), explicitly correcting for the "brighter-fatter" effect inherent to Si:As detectors. This allowed for the deblending of the foreground M-dwarf flux from the background sources.
Comparative Analysis: The derived Spectral Energy Distributions (SEDs) and spectral features were compared against templates of Hot Dust Obscured Galaxies (Hot DOGs), active galactic nuclei (AGN), and starburst galaxies.

Key Results

Spatial Resolution of Excess: MIRI imaging revealed that the mid-IR excess is not intrinsic to the M dwarfs. In both candidates, a distinct background source is located $\approx$ $\approx$ 1 arcsec from the M dwarf.
- Candidate D: The background object is a point-source dominated galaxy. The M dwarf is visible in F560W and F1000W but is largely invisible against the background galaxy in F1500W.
- Candidate E: The background object is an extended galaxy with bright knots. The M dwarf overlaps with a bright region of the galaxy in F1000W and F1500W, making direct flux measurement difficult, though the M dwarf is visible in F560W.
Spectral Characterization and Redshifts:
- Candidate D: Spectral analysis of fine-structure lines (e.g., [Ne II], [Ar III]) and a weak [Mg IV] line yields a redshift of $z \approx 0.922$ . The spectrum shows a steeply rising continuum, weak PAH bands, and a silicate absorption feature, consistent with an obscured AGN.
- Candidate E: Spectral lines yield a redshift of $z \approx 0.410$ . The spectrum is dominated by strong PAH emission features and a weak continuum, characteristic of a dusty starburst galaxy.
Flux Decomposition: Photometric modeling indicates that the M dwarfs contribute $\approx$ 80% of the flux in F560W but are subdominant in longer wavelengths (F1000W and F1500W), where the background galaxies dominate. The stellar fluxes align with Rayleigh-Jeans tails of standard M-dwarf atmospheric models (Phoenix models), showing no intrinsic mid-IR excess.
Nature of Background Objects:
- Candidate D's background galaxy exhibits properties consistent with a Hot DOG (Hot Dust Obscured Galaxy) at $z \approx 0.9$ , though with a lower star-formation rate than typical high-redshift Hot DOGs.
- Candidate E's background galaxy is consistent with a dusty starburst, potentially a "bump" DOG, though the 1.6 $\mu$ m bump is obscured by the foreground star.

Conclusions and Significance
The paper concludes that the mid-IR excess observed in candidates D and E is entirely attributable to background galaxies projected within $\sim$ 1 arcsec of the M dwarfs, which confused the lower-resolution WISE photometry. There is no evidence for Dyson spheres or extreme debris disks around these stars.

The significance of these findings lies in:

Refuting Technosignature Hypotheses: The study demonstrates that source confusion with red, IR-bright background galaxies is a viable and likely explanation for the mid-IR excess in the Project Hephaistos sample, effectively ruling out Dyson spheres for these specific candidates.
Contamination Rates: The authors discuss that the contamination probability in previous surveys may have been underestimated. They suggest that Hot DOG-like objects, which are fainter than those typically targeted in Hot DOG studies, could explain the IR excess of the broader candidate sample.
Methodological Implications: The paper highlights the necessity of high-resolution mid-IR imaging or the use of centroid shifts between optical/near-IR and mid-IR data to filter out interlopers in future Dyson sphere searches.
Astrophysical Discovery: Beyond the technosignature search, the work identifies low-redshift ( $z \lesssim 1$ ), IR-bright galaxies that are fainter than well-studied Hot DOGs. The unique configuration of these foreground M dwarfs acting as natural guide stars for the background galaxies offers a potential avenue for future adaptive optics studies of these rare objects.