JWST spectra caught an unmatched 1.6-micrometer line twice in the Hubble Ultra Deep Field

Standard NIRSpec work starts from a catalog. Astronomers pick known targets from existing observations and position the instrument's shutters over them, so the technique relies on knowing where the target is first ^[5]. Mason Huberty's team wanted to see whether the spectrograph could find objects that had not been found yet ^[4], using a technique called Multiobject Coronagraphy for Slitless Spectroscopy, or MOCSS ^[8].

The Ultra Deep Field suited that test. Hubble spent more than 270 hours on it, and it holds nearly 10,000 galaxies ^[1], so the team could check its detections against known galaxies ^[9]. Six of the seven detections described matched, about 86 percent ^[14]. The seventh is the one that did not.

The control is built into the schedule. The two windows used different observation angles, which ScienceAlert says would help eliminate artifacts and glitches ^[7]. In our view, a line that appears at both angles is harder to pin on the detector.

The flare idea needs care. ScienceAlert's account says there may be something otherwise too dim to detect that briefly unleashed a powerful flare ^[15]. The two windows are 2.9 months apart ^[13]. If "brief" means shorter than that gap, the source would have flared twice, once per window, and stayed dark in every other observation of the patch.

The thing this doesn't tell you is how many other observations that is, or how deep they go. ScienceAlert's account does not say. In our view, that coverage figure decides whether "brief" is a measurement or a guess.