Somewhere between a star and a planet, brown dwarfs drift through the galaxy too dim to see with the naked eye and too cool to sustain nuclear fusion. They are cosmic also-rans, objects that almost became stars but fell short. And according to a study posted in May 2026 on the preprint server arXiv, thousands of them have been lurking in NASA’s own data for years, waiting for human eyes to find them.
Volunteers working with the Backyard Worlds: Planet 9 citizen-science project have identified 3,006 new brown-dwarf candidates in the sun’s neighborhood by sifting through archival infrared images that automated software had already scanned and moved past. If the candidates survive spectroscopic confirmation, the discovery would more than double the known census of L- and T-type brown dwarfs near our solar system, a result that would force astronomers to reconsider how many low-mass objects populate the Milky Way’s disk and how they formed.
What volunteers found, and how they found it
The raw material behind the discovery is the CatWISE2020 catalog, a database of 1,890,715,640 infrared sources assembled from observations by NASA’s Wide-field Infrared Survey Explorer (WISE) and its NEOWISE extension. The CatWISE2020 catalog specifically draws on observations collected between 2010 and 2018 in two mid-infrared bands called W1 and W2, even though the broader NEOWISE mission continued operating until the spacecraft was deorbited in early 2025. Those infrared bands are especially well suited to spotting brown dwarfs because these objects glow brightest at infrared wavelengths. But many are so faint that they vanished into the noise during earlier automated searches.
That is where the volunteers came in. Backyard Worlds: Planet 9 operates on the Zooniverse platform and is formally supported by NASA. Participants use a flipbook-style tool that rapidly blinks between images taken months or years apart. A distant galaxy stays put; a nearby brown dwarf shifts position against the background. The human visual system turns out to be remarkably good at catching that subtle motion, often better than the algorithms that processed the same frames first.
To push the data even deeper, the project team stacked multiple WISE and NEOWISE exposures into single “coadd” frames, a technique documented in a separate technical paper. Layering years of repeat observations effectively multiplies exposure time without building a new telescope, pulling objects above the noise floor that no single-pass survey could have detected.
The pipeline worked in stages. CatWISE2020 supplied positions and brightness measurements across multiple epochs. The Backyard Worlds interface presented volunteers with time-sequenced cutouts built from the deep coadds. Participants flagged sources that appeared to move. Then the research team, led by scientists including J. Davy Kirkpatrick at Caltech’s Infrared Processing and Analysis Center, applied quantitative motion and color cuts to winnow thousands of volunteer suggestions down to 3,006 motion-confirmed candidates: 2,357 classified as L-type and 649 as T-type.
Why the numbers matter
Brown dwarfs occupy a strange gap in the cosmic hierarchy. They form the same way stars do, from collapsing clouds of gas and dust, but they never accumulate enough mass to ignite stable hydrogen fusion in their cores. The lightest ones blur into the territory of giant planets, raising questions about where star formation ends and planet formation begins.
Knowing how many brown dwarfs exist near the sun is not just a bookkeeping exercise. The local census feeds directly into estimates of the Milky Way’s total mass budget and constrains models of how stars and planets form at the lowest masses. Before this study, the nearest known brown dwarfs included the Luhman 16 pair, a binary system roughly 6.5 light-years away discovered in 2013 using WISE data. If 3,006 new neighbors hold up, it would suggest that prior surveys systematically missed a large population of cool, faint objects, with ripple effects across stellar and planetary astrophysics.
The preprint’s own language frames the result carefully: confirming the full sample would “more than double” the cataloged L and T dwarf population near the sun. That conditional phrasing matters. Different catalogs use different volume limits and completeness thresholds, so the precise multiplier depends on which prior census serves as the baseline. But even a partial confirmation rate would represent a significant expansion of the known brown-dwarf population. It is worth noting that these 3,006 candidates represent only the solar neighborhood; the total brown-dwarf population across the entire Milky Way is estimated to number in the tens of billions, so this local doubling, while dramatic for the nearby census, addresses just one small patch of the galaxy’s full inventory.
What still needs to happen
The 3,006 figure describes candidates, not confirmed brown dwarfs. Each one was identified through its infrared colors and measured motion across the sky, both strong indicators of a nearby, cool object. But spectroscopic confirmation, the gold standard for assigning a definitive spectral type, has not yet been reported for the bulk of the sample. Until telescopes collect spectra from a significant fraction of these targets, some portion could turn out to be distant red giants, unresolved galaxy pairs, or other contaminants that mimic brown-dwarf signatures in two-band infrared photometry.
Ground-based observatories with powerful infrared spectrographs can characterize the brightest candidates, measuring temperatures, surface gravities, and atmospheric compositions. For fainter targets, the James Webb Space Telescope (JWST) offers the sensitivity and spectral range needed to reach objects that ground-based instruments cannot. JWST has already proven its ability to study brown-dwarf atmospheres in detail, and a confirmed sample of this size would provide a rich target list for future observing proposals.
Volunteer accuracy adds another layer of uncertainty. NASA’s descriptions of the citizen-science workflow emphasize the value of many independent eyes scanning the same data, and previous Backyard Worlds papers have reported high hit rates for subsets that received follow-up. But extrapolating that track record to a sample this large requires caution. The project does not publish detailed false-positive statistics for this specific search, so the final confirmation rate remains an open question.
There are also selection biases to consider. The search relies on WISE’s sensitivity limits and on motion thresholds that favor objects with relatively large apparent movement. Very slow-moving or extremely faint brown dwarfs could still be undercounted even after this expansion. Crowding in dense star fields and confusion with background galaxies may also shape which regions of the sky yield the most candidates, complicating efforts to convert raw counts into robust space densities.
Why archival infrared data keep producing surprises
NEOWISE, the extended mission that kept the WISE spacecraft scanning the sky for over a decade, collected its final observations before the spacecraft was deorbited in early 2025. But the data it left behind are far from exhausted. The Backyard Worlds project demonstrates that archival datasets, when paired with human pattern recognition and deeper image-processing techniques, can still yield discoveries that eluded the original analysis pipelines.
The preprint and the CatWISE2020 catalog record are the primary sources underpinning this result. NASA’s project pages for Backyard Worlds and NEOWISE provide institutional context about the citizen-science pipeline and confirm the project’s legitimacy, but they do not independently verify the 3,006-candidate count or the doubling claim. Peer review of the preprint may still lead to refinements in the analysis or the interpretation of the candidate list.
For now, as of June 2026, the practical next step is spectroscopic follow-up. As those observations accumulate, astronomers will be able to sort genuine brown dwarfs from impostors, refine the local census, and test whether the solar neighborhood really has been hiding twice as many failed stars as anyone realized. If it has, the discovery will belong not just to professional astronomers but to the thousands of volunteers who spotted what the machines missed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
