“If even a few of these objects turn out to be what we think they are, our discovery could challenge current ideas about how galaxies formed in the early universe,” said Haojing Yan, a University of Missouri astronomy professor. His group has discovered 300 exceptionally bright candidates for ancient galaxies from deep infrared imaging with the James Webb Space Telescope.
The research relies on the ability of JWST’s Near-Infrared Camera (NIRCam) and Mid-Infrared Instrument (MIRI), designed to capture the universe’s earliest billion years’ faint, stretched light. As photons from early galaxies travel through expanding space, their wavelengths stretch a process called redshift shifting their once-ultraviolet shine into the infrared ranges where Webb is best. For the most remote sources, this transition can be in excess of z = 14, or a time only a few hundred million years past the Big Bang.
To separate such high-redshift galaxies from the vastly more common near ones, Yan’s team used the dropout, or Lyman-break, method. It takes advantage of an abrupt drop-off in a galaxy’s spectrum at 1216 angstroms, due to neutral hydrogen soaking up ultraviolet light. At high redshift, that threshold shifts into redder filters, so the galaxy disappears in bluer bands but is still detectable in redder bands. As the work on Lyman-break galaxies has demonstrated, this can produce redshift estimates to a few percent accuracy, but spectroscopic confirmation is still the benchmark.
Spectroscopy, performed with JWST’s Near-Infrared Spectrograph (NIRSpec) or MIRI’s low-resolution spectrometer, disperses a galaxy’s light into its constituent wavelengths, revealing emission lines from elements like oxygen, carbon, and hydrogen. These signatures not only pin down redshift but also indicate star formation rates, chemical enrichment, and the presence of active galactic nuclei. In the case of record-holders like JADES-GS-z14-0, spectroscopy has revealed prominent [O III] and Hβ lines, suggesting intense starbursts and metallicities less than 10 % of the solar value.
For Yan’s candidates, complete spectra are not yet available. Instead, the team employed spectral energy distribution (SED) fitting comparing broadband photometry to theoretical models of galaxies to estimate redshifts, ages, and masses. This method, developed in the depths of surveys, can separate out genuinely distant galaxies from impostors like dusty, lower-redshift systems or small “little red dots” whose intrinsically low luminosity is amplified by black hole accretion. Deleting such active nuclei from samples in some initial Webb analyses removed apparent tensions with cosmological models, although there still exists an excess of large early galaxies.
The luminosity of many of these 300 candidates is itself a mystery. Standard. models say that galaxies in the first few hundred million years should be faint and small, their growth restricted by the time available for gas to collapse and become stars. But Webb’s sensitivity NIRCam can go as faint as ~2 nJy at 5σ in ultradeep exposures has revealed objects whose luminosities imply rapid, efficient star formation. They might be analogs of MoM z14, the “mother of all early galaxies” discovered at z = 14.44 and roughly 50 times smaller than the Milky Way but already carbon- and nitrogen-enriched.
Instrumental design is paramount to these discoveries. NIRCam’s two-module layout spans 0.6–5 microns with wide and medium filters, allowing for accurate color measurements from the Lyman break. MIRI extends coverage to 28 microns, critical for probing rest-frame optical lines in the most distant galaxies, where key diagnostics like Hα have shifted beyond NIRSpec’s range. In ultradeep integrations up to 23.8 hours for some fields MIRI can detect fluxes as low as 28 nJy, revealing nebular emission that betrays the chemical state of early star-forming regions.
If spectroscopy verifies even a subset of Yan’s sample to be true high-redshift galaxies, the ramifications will be significant. It would suggest that, in the early dense universe, gas cooled and condensed into stars much more effectively than in later times, or that feedback mechanisms that currently constrain star formation worked less well. Or else, part of the excess light is attributable to gravitational lensing by foreground mass, quietly amplifying background galaxies.
For the time being, the 300 glowing smudges in Webb’s infrared mosaics are contenders cosmic postcards from a time when the first galaxies were coming together, their light stretched thin across time and space, but still bright enough to find us after over 13 billion years.
