X-ray transients in the Chandra archive

Introducing the cumulative distribution discriminator (CuDiDi)

¹ KTH Royal Institute of Technology, Department of Physics, 10691 Stockholm, Sweden
² The Oskar Klein Centre for Cosmoparticle Physics, AlbaNova University Centre, 10691 Stockholm, Sweden

^★ Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Received: 16 September 2025
Accepted: 13 March 2026

Abstract

X-ray transients on sub-observation timescales represent a diverse and underexplored class of astrophysical phenomena, from stellar flares and magnetar bursts to extragalactic fast transients and supernova shock breakouts. We present a systematic search for such events across 20 212 Chandra ACIS observations using a new detection pipeline that combines source identification, light-curve analysis, catalogue cross-matching, and a novel statistical classifier, the cumulative distribution discriminator (CuDiDi). From 1420 initial candidates, we identified a high-confidence golden sample of 765 transients spanning a broad range of timescales, fluxes, and spectral shapes. The candidates are distributed across the whole sky and show a wide range of durations with a median of 10 ks. A subset of fast events lasting ≲30 s displays very soft spectra and is likely due to flaring dwarf stars, although extragalactic phenomena cannot be ruled out for all of them. The comparison with previously published samples showed that CuDiDi identifies most known transients while imposing somewhat stricter variability criteria, and it also extends the total sample of Chandra transients to include shorter events. We deliver a comprehensive catalogue of sub-observation Chandra X-ray transients and establish a general method for exploiting archival datasets to uncover rare short-lived high-energy phenomena.