What If Most Selected Post-Starburst Galaxies Aren't Really Dead?

A post-starburst galaxy is the astronomical equivalent of a fire that has just gone out. Its spectrum shows signs of a recent burst of star formation, in the form of intermediate-age stars, but almost none of the short-lived massive stars that would signal ongoing star birth. Astronomers have treated these objects as rare, fast transitions between the blue, star-forming galaxies and the red, quiescent ones. A new simulation suggests that many galaxies given that label are not finished at all. They are impostors.

What a post-starburst galaxy is supposed to be

Galaxies in the nearby universe come in two broad types: blue, star-forming galaxies and red, quiescent galaxies. The transition between them is not instant. A small population, called post-starburst galaxies, appears to be caught just after a burst of star formation has been rapidly shut down. Their light is dominated by stars formed in the recent past, and their star formation rates have plunged.

That makes them important witnesses to cosmic quenching, the process that turns active galaxies into dormant ones. But the details are murky. Some observed post-starburst galaxies contain surprising amounts of cold molecular gas, the raw material for stars. Others may be hiding ongoing star formation behind dust. And it is still debated whether mergers, stellar winds, or black holes are the main culprits behind the shutdown.

A simulation built to catch mistaken identities

Elia Cenci and colleagues used the FIREbox cosmological simulation to test whether the way astronomers select post-starburst galaxies actually captures the objects they intend to study. FIREbox models a cosmological volume 22.1 comoving megaparsecs on a side using the FIRE-2 galaxy-formation model, which includes gas cooling, star formation, and stellar feedback but not black hole feedback. The team then post-processed 266 simulated galaxies at redshifts z=0.7 and z=1, when the universe was younger than today, with radiative transfer, emission-line, and spectral synthesis codes to produce mock colors and spectra.

Each simulated galaxy was viewed from 16 different angles, mimicking how dust and orientation can change a galaxy's appearance. The researchers then applied the same rest-frame optical color criteria used in observations to pick out post-starburst candidates. Finally, they checked the simulation's true star-formation histories. A true post-starburst was defined as a galaxy that had a recent starburst, then a drop of at least a factor of four in its star-formation rate, and now has a very low specific star-formation rate, meaning it forms few stars per year for its stellar mass. Anything selected as a post-starburst but not meeting that full definition was labeled an impostor.

The paper is available as a preprint on arXiv, and its conclusions rest on a simulation, not direct observation.

Only a small fraction are truly quenched

The simulated post-starburst fraction was broadly plausible: about 9 percent of the galaxies in the sample, which have stellar masses above about 5 billion Suns, were selected as post-starbursts. But the physical reality behind that label was very different. In the simulation, only about 8 percent of the selected post-starbursts in the fiducial sample were truly quenched after-starburst galaxies. Roughly one in ten, in broad terms, were genuine. The rest were impostors.

Most impostors had star-formation rates and gas masses comparable to normal star-forming galaxies of the same stellar mass. They looked post-starburst because dust, viewing angle, or the aperture used to collect their light altered their colors. The result does not mean that every observed post-starburst galaxy is a fake. It means that a photometric selection can be fooled more easily than astronomers might like.

The gas-rich dead galaxies are the tell

The simulation offers a natural explanation for one observational puzzle: post-starburst galaxies that still contain substantial molecular gas. In FIREbox, gas-rich post-starbursts were almost exclusively impostors. True post-starbursts had low molecular gas masses, generally below the detection limits of current ALMA observations that trace carbon monoxide, a standard proxy for cold molecular gas. If a supposedly dead galaxy still has a large reservoir of star-forming fuel, it is more likely to be a star-forming galaxy in disguise.

Infrared colors could separate real from fake

True post-starbursts and impostors did differ in their infrared light. The real ones had lower mid-infrared emission at a rest-frame wavelength of 24 micrometers relative to near-infrared emission at 2 micrometers. Mid-infrared light at that wavelength is a proxy for recent star formation; near-infrared light is a proxy for stellar mass. So true post-starbursts appear faint in the star-formation-sensitive band for their stellar mass. Impostors look more like ordinary star-forming galaxies.

But the separation is not clean. There is significant scatter, and the authors present it as a possible diagnostic rather than a perfect test. Deeper multi-wavelength observations, including infrared and radio data, would be needed to confirm which galaxies are genuinely quiet.

Mergers explain some, but not all

The simulation also examined galaxy interactions. Post-starburst candidates were more likely than non-post-starburst galaxies to have experienced an interaction, such as a merger or close passage, in the past billion years. Among true, temporarily quenched post-starbursts, about 20 percent had interacted in that period. Among the broader after-starburst galaxies that were also selected as post-starbursts, about half were in interacting systems. Mergers can help trigger starbursts and rapid changes, but they are not the whole story.

The black-hole implication, and its limits

Using a Bayesian framework that combines the simulation with observational estimates, the team went further. Assuming that about 27 percent of observed post-starbursts are impostors and that about half of galaxies are quenched, they estimate that roughly 73 percent of true post-starburst galaxies must be shut down by a mechanism other than stellar feedback. The most likely candidate is feedback from a supermassive black hole, known as active galactic nucleus, or AGN, feedback.

That inference comes with important caveats. FIREbox does not include black holes at all, so it cannot directly simulate AGN feedback. It also underpredicts the number of long-term quiescent galaxies, and the simulated sample is small. The estimate depends on observed fractions of post-starbursts, impostors, and quenched galaxies, which carry their own uncertainties. The result is a prediction about what the full population likely requires, not a direct detection of black holes switching galaxies off.

The quiet galaxies that are only resting

The new work does not dethrone post-starburst galaxies as important objects. It sharpens the question of what the label really means. Many galaxies that look post-starburst may be dusty, obscured, or simply seen from an awkward angle. Some may be catching their breath after a burst and will form stars again later. The true post-starbursts may be rarer and quieter than their impostors. Finding them will take more than optical colors alone. It will take infrared and radio eyes, and simulations that include the black holes that may help determine which galaxies stay dark.