Core-collapse supernovae—the final explosions of massive stars—remain poorly understood in their very first moments. It is then that the shock wave breaks out at the surface (a supernova), and its radiation carries the imprint of the star’s structure and surrounding material. As Fritz Zwicky predicted, supernovae are transient events, but their first instants remained elusive. Until recently, the only solid example of an X-ray breakout was supernova SN 2008D. The new flash EP260321a, discovered by the numerical simulation Einstein Probe mission in 2026, offered a chance to test breakout models in detail within a dense stellar evolution wind. Its X-ray signal is essentially the scream of a star, breaking through the gaseous curtain it had itself shed.
The researchers used a semi-analytical model based on the thin-shell approximation, where the compressed plasma between shock waves is treated as a single shell. The calculations included relativistic corrections for ejecta traveling at up to 0.7 times the speed of light. The free parameters were the maximum shell velocity (Γₘₐₓ=1.5), the kinetic energy of the high-velocity ejecta component, the density of the interstellar medium (ρ∝r⁻² profile), and its outer radius. Synthetic bolometric light curves, converted to the X-ray band, were compared with photometry data from the WXT and FXT instruments aboard Einstein Probe.
The model successfully reproduced both the peak luminosity (~8×10⁴⁴ erg/s) and the duration (~1750 s) and decay shape of EP260321a’s light curve. This required a shell with energy Eᵣₑₗ≈3.5×10⁴⁹ erg (excluding the main body of the supernova) and a dense stellar evolution wind with a mass-loss rate of about 1.2×10⁻³ M☉/yr. The dense medium must extend to a radius of ~350 R☉, corresponding to a mass of only ~10⁻⁵ M☉. The transition to a steep luminosity drop after ~1200 s marks the moment when the shell exits the dense gas. The spectrum is close to a Planck's law with a temperature of ~140 eV, but the equilibrium temperature predicted under thermodynamic equilibrium Stefan–Boltzmann law turned out lower than observed—this requires accounting for non-equilibrium effects.
The results confirm that shock breakout in a dense wind is a natural mechanism for long-duration X-ray transients like EP260321a. Small amounts of circumstellar gas (10⁻⁵ M☉), ejected just days before the explosion, can generate powerful emission, making such events a sensitive probe of the final stellar evolution of massive stars. The constraints on the energy of the high-velocity ejecta (a few 10⁴⁹ erg) and the absence of a relativistic jet align EP260321a with ordinary supernovae, yet set it apart from gamma-ray burst low-luminosity events. Each new breakout brings us closer to deciphering these death cries that tell of the last days of massive luminaries.
A growing sample of X-ray breakouts from the Einstein Probe and SVOM missions, plus future ultraviolet surveys (ULTRASAT, UVEX), will cover a wider range of ambient densities. Accounting for 2D effects and detailed numerical simulation of radiative transfer under non-equilibrium conditions will help explain the spectra. Direct pre-explosion observations in the optical using spectroscopy could capture mass-loss episodes.
The results will impact the theory of massive star evolution, the physics of shock waves in relativistic plasma, and the classification of fast X-ray transients.
Detailed spectral modeling of SN 2026gzf in the optical for an independent assessment of the ejecta structure, and a search for similar events with emphasis on pre-explosion photometry.
The paper addresses the problem of the mass-loss mechanism in the final days of massive stars, as well as the origin of low-luminosity gamma-ray bursts and their connection to ordinary supernovae.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
The authors applied a semi-analytic thin-shell model with relativistic corrections to the X-ray transient EP260321a, discovered by Einstein Probe. The model reproduced the light curve, duration, and spectrum of the outburst, showing that the shock wave broke out through a dense wind with a mass-loss rate of ~1.2×10⁻³ M☉/yr, ejected days before the explosion.
Where its strength is
The work demonstrates that even a small mass of circumstellar gas (~10⁻⁵ M☉) can produce powerful X-ray emission when the shock breaks out, and provides quantitative constraints on the energy of high-velocity ejecta and the density profile of the wind.
Where it could go next
Worth trying
Why this may matter
Such events turn short X-ray flashes into a quantitative tool for 'weighing' pre-explosion mass loss, which is critical for understanding the final stages of massive star evolution and the connection between supernovae and gamma-ray bursts.
Nearby works in our archive
The archive is crowded around this work: 6 reviews nearby, closest at 0.76. This ground is well walked — and there is plenty to compare against.
If the same mass-loss rate had persisted throughout the progenitor star’s life, it would have shed several solar masses — meaning the enhanced wind blew only briefly, likely due to wave heating during the silicon-burning stage.
The X-ray 'scream' of a dying star recalls the final signal transmission in Stanisław Lem’s novel Solaris, where the ocean sends mysterious bursts carrying information about its structure.