JWST has picked up a strong, narrow infrared emission line at a spot where deep space images show no object. The peer-reviewed study describes a previously unknown source in the Hubble Ultra Deep Field, one of the most intensively observed patches of sky.
The observation came from a new use of JWST's Near-Infrared Spectrograph, or NIRSpec. Instead of aiming the instrument at known targets, the team opened nearly all of its microshutters while blocking bright foreground objects. That let the researchers search the field for emission lines without first knowing where a source should be.
Why the signal looks real
The line appeared at 1.6004 micrometers in four stacked datasets. The telescope observed the field from two different orientations on Aug. 23-24 and Nov. 19, 2023, a span of 88 days. Its measured strength and wavelength remained consistent across the observations, making a simple detector glitch unlikely.
Yet the inferred position has no counterpart in existing Hubble or JWST images. That absence is especially puzzling because the line alone should have been bright enough to appear in JWST's JADES imaging from October 2022 if it had been present then.
This does not mean the light came from literal empty space. It means the spectroscopy detected emission where imaging has not revealed a persistent star or galaxy.
Every explanation has a catch
A single emission line cannot identify an object's distance or composition by itself. Cosmic expansion shifts light toward redder wavelengths, so several different atomic transitions at different distances could land at the observed wavelength.
The researchers tested four common possibilities: oxygen lines from galaxies at redshifts 2.20 or 3.29, hydrogen-alpha at redshift 1.44, and Lyman-alpha from a galaxy at redshift 12.16. None fits cleanly. Some would require additional lines that were not detected, while others would demand unusually extreme conditions.
