The redshift distribution of Einstein Probe transients supports their relation to gamma-ray bursts

TL;DR

This study analyzes the redshift distribution of Einstein Probe fast X-ray transients to test their connection to long-duration GRBs. By compiling EP redshifts and comparing their cumulative distribution to that of long GRBs using non-parametric tests, it finds no significant difference, suggesting a shared underlying population. The EP events, including gamma-ray-detected and gamma-ray-dark transients, also align with the Amati $E_p$-$E_{iso}$ relation, supporting collapsar progenitors. The results imply that many EP transients arise from massive-star deaths and reveal a potentially substantial population of soft X-ray dominated, failed-jet or dirty-fireball events that are missed by traditional gamma-ray monitors. As EP expands, the sample will enable tighter constraints on progenitor diversity and event rates, refining our understanding of the deaths of massive stars and their jet physics.

Abstract

The launch of the \textit{Einstein Probe} unleashed a new era of high-energy transient discovery in the largely unexplored soft X-ray band. The \textit{Einstein Probe} has detected a significant number of fast X-ray transients that display no gamma-ray emission, complicating their robust association to more common gamma-ray bursts. To explore their possible connection, we analyzed the redshift distribution of both \textit{Einstein Probe} fast X-ray transients and long duration gamma-ray bursts. A comparative analysis of their cumulative redshift distributions using non-parametric two-sample tests, namely the Kolmogorov-Smirnov and Anderson-Darling tests, finds no statistically significant difference. These tests favor that their redshifts are drawn from the same underlying distribution. This empirical connection between \textit{Einstein Probe} transients and long gamma-ray bursts is further supported by their agreement with the so-called ``Amati relation'' between the spectral peak energy and the isotropic-equivalent energy. Together, these results indicate that most extragalactic \textit{Einstein Probe} fast X-ray transients are closely related to long gamma-ray bursts and originate from a massive star (collapsar) progenitor channel. Our findings highlight the role of the \textit{Einstein Probe} in uncovering the missing population of failed jets and dirty fireballs that emit primarily at soft X-ray wavelengths.

Figures (7)

• Figure 1: Left: Histogram of the EP reporting latency measured as the time between GCN notice and EP trigger time. All EP transients (from July 2024 onwards) are shown in blue, and those with measured redshifts marked in red. The dashed lines mark $3$ and $12$ hours from the EP trigger. Right: Reporting latency versus discovery date for EP transients. Those with no redshift are marked in gray, and those with redshifts in red. Sourced with redshift at $z$$<$$1$ are shown by squares.
• Figure 2: The $E_\textrm{p}$-$E_\textrm{iso}$ plane of GRBs with short GRBs shown in gray and long GRBs in black. EP transients with gamma-ray detections are shown in blue, and those without gamma-rays in purple with upper limits represented by downward triangles. We explicitly highlight HETE-2 bursts in orange Pelangeon2008. The solid black line and $3\sigma$ scatter to the correlation for long GRBs are reproduced from Amati2019. The figure is reproduced from Dichiara2021EE.
• Figure 3: Left: Rest frame $0.5$$-$$4$ keV X-ray luminosity measured by EP/WXT versus redshift (Table \ref{['tab:EPtab']}). The time-averaged luminosity is represented by solid symbols and empty symbols refer to the peak luminosity. We have designated between EP-GRBs (blue) and those without gamma-rays (purple). Sources represented by squares have been found to be associated to Type Ic-BL supernovae at $z$$\lesssim$$0.4$vanDalen2024Rastinejad2025EPSrinivasaragavan2025EP0108aEP250304a-SN-GCN. Right: Rest frame luminosity versus X-ray photon index measured by EP/WXT. The gray shaded region represents the synchrotron "line of death" Preece1998.
• Figure 4: Left: Cumulative distribution of the redshift of EP transients (red) compared to short (gray; OConnor2022) and long (black; Greiner Catalog) GRBs. Right: The EP transients are further divided into those without gamma-ray detections (purple) and those with joint GRB detections (EP-GRBs; blue).
• Figure 5: Spectroscopy of EP250302a (VLT X-shooter), EP250821a (Gemini-South GMOS), and EP250827a (VLT X-shooter) used to measure their redshifts. For EP250827a, due to the large number of absorption lines, we show both the UVB and VIS arms of the X-shooter spectrum. The spectra have not been smoothed or re-binned.