Transient Signatures of Star-Envelope Collisions in Little Red Dots

Tomoya Suzuguchi, Kohei Inayoshi

First listed 2026-05-29 | Last updated 2026-05-28

Abstract

Little red dots (LRDs) are compact high-redshift objects, newly discovered by the James Webb Space Telescope. Although LRDs exhibit broad Balmer emission lines suggestive of the presence of active galactic nuclei (AGN), their spectral features differ significantly from those of ordinary AGN. Recent studies suggest that their characteristics can be explained if accreting supermassive black holes (SMBHs) are embedded within dense gaseous envelopes and surrounded by compact stellar clusters. In this scenario, stars in the cluster can scatter onto plunging orbits that intersect the envelope and collide with its surface. Here we investigate the observational properties of such star-envelope collisions as luminous transient events. We find that collisions involving red supergiants with radii of $\sim 10^{3}~R_\odot$, together with gaseous envelopes whose masses are comparable to those of the central SMBHs, are the most promising targets due to their high luminosities and long durations. For compact clusters with sizes of $\lesssim 10~{\rm pc}$, such massive stars can participate in star-envelope collisions within their lifetimes at event rates reaching $\sim 0.3~{\rm yr}^{-1}$ per LRD. We show that these transients are detectable with future wide-field surveys such as the Nancy Grace Roman Space Telescope if they occur at relatively low redshifts ($z \lesssim 1$). Detection of such transients would provide strong evidence for the envelope+stellar-cluster scenario of LRDs and offer a unique probe of the envelope mass, which is otherwise difficult to constrain from LRD spectra alone.

Short digest

This paper works out the transient flares expected when stars from the compact stellar clusters invoked around little red dots plunge into the surface of an optically thick black-hole envelope. The most promising events are collisions by red supergiants with radii of about 10^3 R_sun into envelopes whose masses are comparable to the central SMBH, because those give the brightest and longest-lived signals. For cluster sizes of ≲10 pc, such massive stars can still reach the envelope within their lifetimes, yielding rates as high as about 0.3 yr^-1 per LRD. If analogous systems exist at lower redshift, Roman-like wide-field surveys could detect these flares out to z ≲ 1, making them a clean test of the envelope+stellar-cluster LRD picture and a rare handle on the envelope mass itself.

Key figures to inspect

• Figure 1. Use this schematic to orient the reader to the paper’s core physical picture: an SMBH embedded in a luminous, optically thick envelope that makes the red optical continuum, plus a compact stellar cluster that supplies the blue UV light and occasionally launches stars onto plunging trajectories. It is the cleanest one-panel summary of the envelope+cluster scenario that the transient calculation is meant to test.
• Figure 2. This is the central parameter-space figure for the transient itself, showing how the duration, peak luminosity, and peak temperature change with envelope mass for different stellar radii and Eddington ratios. It directly visualizes the paper’s headline result that red-supergiant collisions and envelopes with masses around the SMBH scale give the most detectable events because they are both brighter and longer lasting.
• Figure 4 links the transient model to stellar-population realism by showing which stellar masses can actually be injected into the envelope before the stars die, as a function of cluster size, UV luminosity, and Eddington ratio. This is the figure that justifies why very compact clusters, at roughly the ≲10 pc level, are needed for red-supergiant collisions to occur within stellar lifetimes.
• Figure 5. This figure turns the feasibility argument into an observable prediction by giving the event rate of star-envelope collisions across the same cluster and luminosity parameter space. It is the quantitative support for the abstract’s quoted rates, including the cases that reach about 0.3 events per year per LRD.
• Figure 6. Use this as the observability synthesis figure: it converts the modeled transients into observer-frame SEDs at multiple redshifts and overlays Roman and LSST sensitivity limits for different visit stacks. It is the best single figure for the paper’s practical bottom line that these events are mainly a low-redshift discovery opportunity, with Roman favored for detections out to about z ≲ 1.