Was Capotauro a Pair-Instability Supernova, Not a z≈30 Galaxy?

Some objects in the early universe refuse to sit still. In December 2022, JWST's CEERS survey identified a compact red source nicknamed Capotauro. It was detected in the NIRCam filters F444W and F410M but was missing in F356W — an F356W-dropout, the kind of color jump that can signal an extremely distant galaxy. The first analysis favored a luminous galaxy at z≈30, where redshift z is a measure of cosmic distance and time (higher z means earlier), with M_UV≈−21.4 and an inferred number density that would worsen the tension with galaxy-formation models. Then, about 800 days later, JWST obtained a NIRSpec spectrum as part of CAPERS. Convolved with the F444W filter, that second epoch implied the source had brightened by about 20% at 4.4 micrometers. The paper treats that variability as motivation to consider a transient nature.

JWST has found more bright early galaxies than virtually all pre-JWST models expected. Aside from a few extreme contaminants such as CEERS-93316, spectroscopic confirmation rates for many candidates have been remarkably high (about 80%), so the tension with galaxy-formation theory is not easy to dismiss. If photometric candidates at 15<z<28 were confirmed, the inferred ultraviolet luminosity density would exceed some model predictions by a large factor, and by z≈25 matching such densities would become virtually impossible for current galaxy-formation models. One possible escape is contamination: some candidates may not be ultra-distant galaxies at all. Capotauro is an extreme test case.

A red dot that brightened by 20%

Capotauro was first identified in CEERS as an F356W-dropout. Its F444W and F410M fluxes made it a plausible luminous galaxy at z≈30 with M_UV≈−21.4 — an ultraviolet brightness that would be remarkable so early. But the second epoch, roughly 800 days later, showed the source brightening by about 20% at 4.4 micrometers. That second epoch was a NIRSpec spectrum, not a direct NIRCam image; the F444W brightness was synthesized by convolving the spectrum with the F444W filter. Variability over that baseline is a clue that the light may not come from a stable galaxy.

A pair-instability supernova at z≈15 fits

That led Andrea Ferrara and colleagues to ask whether Capotauro could instead be a single exploding star — specifically, a pair-instability supernova, or PISN. A PISN is a stellar explosion expected from very massive, metal-free stars. In the paper's models, the progenitor is about 250–260 solar masses, and the luminosity is powered by radioactive nickel. The most massive models reach peak bolometric luminosities above 10^44 erg/s, making high-redshift PISNe potentially detectable with JWST.

Ferrara and colleagues compared Capotauro's two epochs of photometry and its NIRSpec spectrum with a library of PISN light curves, spectral energy distributions, and synthetic spectra. They allowed the redshift and time since explosion to vary, then asked which combination best reproduced the observations.

In their PISN model fits, the preferred solution is not a z≈30 galaxy. The fits place the explosion at redshifts between about 11.6 and 15.3. The most successful models are the most massive hydrogen-free ones, especially He125 and He130. Only He125 and He130 satisfy the 1σ F356W upper limit; the other three fiducial models are excluded also by the 3σ upper limit. The favored He130 model is a roughly 250–260 solar-mass metal-free star exploding at z≈15.32, with the second epoch about 115 days after the explosion in the supernova's rest frame. That lower redshift helps explain the F356W dropout: the model spectra drop sharply below about 3–3.5 micrometers because ultraviolet light is absorbed by the expanding ejecta.

The NIRSpec spectrum gives the idea a second test. The He130 PISN model matches the observed spectrum reasonably well, including an overall rise toward longer wavelengths and absorption from elements freshly made in the explosion. The paper notes expected iron-group blanketing and magnesium absorption. The residuals are mostly within the large error bars, though the fit is not perfect.

A brown dwarf fits the spectrum, but not the variability

The authors also test a very different explanation: Capotauro could be a local, cool brown dwarf. The coolest brown dwarfs, Y dwarfs, have effective temperatures below about 600 K and spectra peaking near 5 micrometers. The authors compare with WISE 0359-54, a Y0 brown dwarf about 450 K and 13.57 parsecs away. Scaled to match Capotauro's F444W flux, such an object would sit about 785 parsecs away.

Surprisingly, the brown dwarf spectrum fits about as well as the PISN spectrum. The key difference is beyond 5 micrometers: the brown dwarf spectrum falls off, while the PISN model continues to rise and then flattens. Deeper MIRI photometry and/or a NIRSpec spectrum could help separate them. But the brown dwarf does not naturally explain the 800-day variability.

The paper lays out several tests. Confirming the variability with NIRCam would reduce worries about cross-instrument calibration; the current difference is at the >2σ level, and residual calibration uncertainties remain possible. Deep MIRI photometry could check whether the source stays bright beyond 5 micrometers, as PISN models predict. High-resolution JWST spectroscopy could identify the expected absorption features more cleanly. A second NIRCam F444W image could measure proper motion, the sideways motion across the sky. A brown dwarf at 785 parsecs would move about 0.053 arcseconds per year, still compatible with the current upper limit of 0.137 arcseconds per year.

If the PISN interpretation survives further observations, Capotauro would become the most distant supernova yet observed and a rare window onto Population III stars — the metal-free first generation. That would be a scientific prize in itself. It would also warn galaxy surveys that some ultra-high-redshift candidates may be transients, not galaxies.

The work is an arXiv preprint and presents possible evidence, not a confirmed identification. More epochs, deeper mid-infrared data, and higher-resolution spectra are needed to decide between a primordial explosion at z≈15 and a cold brown dwarf much closer to home.

For now, Capotauro remains a red dot that brightened. That simple fact has turned it from an ultra-high-redshift galaxy candidate into a puzzle about which kind of object can flicker in JWST's early-universe census — and whether we may already have seen the death of one of the universe's first stars.