What Silences a Galaxy? JWST Survey Finds Fast Quenching After Starbursts—and a Range of Paths

In the Ultra Deep Survey field, JWST has scrutinized 24 massive galaxies selected as quiescent or post-starburst at redshifts 1 to 3—an era astronomers call cosmic noon, when massive quiescent galaxies were building up. They are massive, with stellar masses above 10 billion Suns.

Quenching is the astronomical term for shutting down star formation. It drives the split between blue, star-forming galaxies and red, quiescent ones. But what flips the switch? Mergers, black-hole feedback, gas exhaustion, and environmental effects have all been proposed. The new work, led by Maya Skarbinski of Johns Hopkins University and colleagues, uses JWST/NIRSpec spectra from the EXCELS survey to reconstruct how 24 massive quiescent and post-starburst galaxies at cosmic noon turned off.

Reading a galaxy's history in its spectrum

The team drew on the JWST Early eXtragalactic Continuum and Emission Line Survey, or EXCELS. The survey uses NIRSpec, JWST's near-infrared spectrograph, to split infrared light into detailed wavelengths. Those spectra carry fingerprints of stars: young, hot stars leave different marks than old, cool ones, and absorption features such as Balmer lines help reveal when a galaxy's last major burst of star formation faded.

The researchers combined the spectra with Hubble ACS and JWST/NIRCam imaging, plus archival VANDELS spectra for four galaxies. They then fit the data with Bagpipes, a Bayesian code that reconstructs a galaxy's star formation history: how much stellar mass formed over time, when star formation peaked, and how quickly it fell. They also measured emission lines, which trace ionized gas, to look for signs of active galactic nuclei (AGN)—supermassive black holes feeding on surrounding material. Photometric super-colors, a principal-component method for classifying galaxies from their colors alone, helped separate post-starburst galaxies from quiescent ones.

A burst, then a fast fade—usually

The reconstructed histories show that most of these galaxies did not fade gently. Many had large bursts of star formation, with rates of roughly 100 to 1,000 solar masses per year, followed by rapid truncation. But the timing was not uniform. The time from peak star formation to quiescence—defined here as falling below a specific star-formation-rate threshold, the star-formation rate divided by stellar mass—ranged from 0.06 billion to 1.75 billion years. Fifteen of the 24 galaxies quenched within 500 million years; six took between 500 million years and 1 billion years; three took longer than 1 billion years.

That spread is the central result. It suggests that massive galaxies at cosmic noon did not all follow one path to quiescence. Some were rapid shutdowns after a starburst; others declined more slowly. The authors also found no evidence that the quenching mechanisms changed with redshift between 1 and 3, even though the sample spans a large slice of cosmic time.

Dusty ancestors and a 600-million-year cameo

Post-starburst galaxies are a brief phase: they have recently stopped forming stars, but their light still carries the mark of a young stellar population. Among the galaxies that enter the post-starburst region, the median visibility time is 612 million years; six galaxies never enter it. The authors report a tentative trend with mass: above 10^10.7 solar masses—roughly 50 billion Suns—the median visibility time is about 340 million years, while lower-mass galaxies in the sample have a median near 660 million years. The mass trend is only about 1.8σ, so it is not a firm detection.

Using those visibility times, the team estimates that post-starburst galaxies could account for about 73% of the growth of the most massive quiescent galaxies, above 10^10.7 solar masses. That is a model-based estimate, not a direct count, but it makes the post-starburst phase an important route to quiescence.

Where did the burst come from? The number densities and star formation rates are consistent with submillimeter galaxies—dusty, intensely star-forming galaxies bright at submillimeter wavelengths—being the progenitors of the quenched population. The authors predict that these galaxies would have spent only about 70 million years above a star formation rate of 200 solar masses per year, roughly seven times shorter than their later visible post-starburst or quiescent phase. After accounting for that short visibility, the numbers are consistent with the observed density of submillimeter galaxies. The sample galaxies are also compact, with a median effective radius near 1.1 kiloparsecs, another similarity to post-starburst systems.

Weak black-hole flickers

Emission-line diagnostics offered a second clue. Of the eight galaxies with robust emission-line detections, four fell in the AGN region of both the BPT and WHaN line-ratio diagrams. That suggests ionizing radiation from a weak active black hole. But none of the 24 sample galaxies were detected in X-rays; only one appeared in radio data; none of the seven with MIRI coverage were detected; and none of the 24 were detected at 24 microns, suggesting a lack of AGN-heated dust. The AGN, if present, may be weak, obscured, or short-lived—or the line ratios may be influenced by shocks. The authors therefore argue that AGN feedback is unlikely to be the primary driver of quenching here. Instead, black holes may help keep galaxies quiet after the main shutdown.

What the survey can and cannot settle

The sample is small: 24 galaxies. The star formation histories come from fitting models to light, so they depend on assumptions about stars, dust, and the shape of the star formation history. The team tested an alternative history model and found consistent results, but individual histories for some galaxies remain poorly constrained. Line ratios alone cannot always distinguish an AGN from shocks, and merger signatures can fade quickly. The work is also a preprint, so it has not yet completed peer review.

Still, the broad picture is coherent. In this sample, many massive galaxies at cosmic noon experienced a powerful starburst and then quickly used up or lost their star-forming gas. Some took longer. A weak AGN may have flickered afterward, maintaining quiescence rather than causing it.

A universe that switches off in different ways

Astronomers have long asked whether galaxy quenching is one process or many. This JWST sample points to many. It shows how some massive galaxies go dark quickly, sometimes in a few hundred million years, and others over more than a billion years. Each spectrum is a fossil record of that shutdown, and future larger samples should sharpen the timing and tell which physical process dominates in which galaxy.