X-ray transients in the Chandra archive: Introducing the cumulative distribution discriminator (CuDiDi)

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.

Figures (15)

• Figure 1: Flow chart of the transient candidate search method. The numbers shown at each step indicate how many sources were retained and excluded at that stage of the analysis.
• Figure 2: Source classification method using CuDiDi. Left: Sketches of representative light-curve behaviour, which illustrate typical variability patterns for persistent, repeating, and transient sources. Centre: Corresponding CDFs. The y-axis shows the cumulative number of counts $N$, normalised by the total so that each distribution reaches unity. Each CDF is split at the median (0.5) into lower (0–0.5, shaded light grey) and upper (0.5–1.0) halves. The mean values of each half, $\mu_{\text{lower}}$ and $\mu_{\text{upper}}$, are indicated. Right: CuDiDi plot showing $\mu_{\text{lower}}$ vs. $\mu_{\text{upper}}$ for each source. The diagonal band (±0.2 around $\mu_{\text{upper}}$ = $\mu_{\text{lower}}$ + 0.5) defines the persistent region. Sources above the band are classified as transients, while sources below the band are considered repeating sources. This visualisation shows how CuDiDi uses statistical asymmetries in cumulative flux distributions to separate source classes.
• Figure 3: Example X-ray light curves and their classification with CuDiDi. Left column: Total counts binned at 1 ks for visual clarity and normalised in time for three representative sources per row. Transient sources are shown in the top row, and non-transient sources in the bottom row. Middle column: Corresponding CDFs, obtained from the light curves with 1 s bins, normalised to unity. The shaded grey area indicates the lower half (0–0.5) of the cumulative fraction. Right column: CuDiDi diagram showing $\mu_{\text{lower}}$ vs. $\mu_{\text{upper}}$, where $\mu_{\text{lower}}$ and $\mu_{\text{upper}}$ denote the mean values of the lower and upper halves of the cumulative distributions, respectively. Transient sources are marked as squares, and non-transient sources are shown as circles. The shaded green band indicates the transient region, defined as $+0.2$ above the diagonal $\mu_{\text{upper}} = \mu_{\text{lower}} + 0.5$. Sources in this region are classified as transients, while those below fall into the non-transient category. The colours and line styles in all panels indicate observation ID: solid green line = 7048, dashed blue line = 16174, dotted orange line = 20120. This demonstrates how CuDiDi captures the distinct variability patterns of transient and non-transient sources.
• Figure 4: Histogram (green; left y-axis) and cumulative distribution (orange lines; right y-axis) of the exposure times for the 20,212 Chandra observations. The vertical dashed purple and solid dark orange lines indicate the mean and median exposure times of the full sample, respectively. The distribution for 765 transient candidates is scaled by a factor of 10 for visual clarity (hatched bars and dotted orange line).
• Figure 5: CuDiDi distribution of the photon-arrival-time asymmetry, showing $\mu_{\text{lower}}$ vs. $\mu_{\text{upper}}$ for all sources we analysed (background-density map). The diagonal black line indicates the locus of temporally symmetric light curves, $\mu_{\text{upper}}=\mu_{\text{lower}}+0.5$. The green lines mark the empirical persistent band, defined as $|\,\mu_{\text{upper}}-(\mu_{\text{lower}}+0.5)\,|\leq0.2$. Sources within and below this band are considered non-transient, while sources above are classified as transients. The 765 transient candidates in our golden sample are overplotted as light blue points. By construction they lie above the band, but their spread within this region reflects a wide range of asymmetry strengths. The bulk of the overall population clusters around $(0.25,0.75)$, as expected for approximately constant emission.