Probing Periodic and Aperiodic Variability of X-ray Sources in M31, M81 and Centaurus A with Chandra

TL;DR

Using two decades of Chandra archival data, the paper conducts a uniform timing analysis of hundreds of bulge X-ray binaries in M31, M81, and Cen A to characterize intra-observation aperiodic variability, long-term inter-observation variability, and periodic signals. The authors apply Bayesian Blocks for short-term variability, Kaplan–Meier estimators for censored long-term variability, and Gregory–Loredo for periodicity, while accounting for foreground stars and background AGNs with Gaia DR3 and CDFS data. They find a population of brief, high-luminosity flares (L_X ~ 10^{37}–10^{40} erg s^{-1}) with very low duty cycles and a mixture of recurrent and isolated dips; long-term variability follows a galaxy-dependent linear rms–luminosity relation, sigma[X] = k <E[X]> with k ≈ 0.49 (M31), 0.30 (M81), and 0.67 (Cen A), consistent with propagating-fluctuation models. No significant coherent periodicities are detected in M81 or Cen A within sensitivity limits, while some M31 periods remain; selection effects and intrinsic period distributions likely explain the scarcity of detections in the external galaxies. Overall, the results support a picture where stochastic accretion-rate fluctuations drive X-ray luminosity on timescales ~10–10^4 s, with dips arising from geometric obscuration in high-inclination systems, and demonstrate the value of timing analyses in extragalactic XRB populations using the Chandra archive.

Abstract

Based on archival Chandra observations, we present a systematic timing survey of several hundred X-ray sources in M31, M81, and Centaurus A, mostly low-mass X-ray binaries (LXMBs), focusing on searching and characterizing aperiodic and periodic variability within single observation. We identify flares in 24 sources in M31, 5 in M81, and 26 in Cen~A; several display recurrent events. Flare durations span from tens of seconds to a few $10^{4}$ s, with peak luminosities of $10^{37}$-$10^{40}\ \mathrm{erg\ s^{-1}}$ and low flare duty cycles of $4.9\times10^{-6}$-$3.5\times10^{-2}$. Dipping events are found in 8 sources in M31, 1 in M81, and 5 in Cen A, including two repeaters. On multi-epoch baselines, the standard deviation of the source luminosity correlates linearly with the mean luminosity, with a coefficient of 0.49 (M31), 0.30 (M81), and 0.67 (Cen A), indicating galaxy-to-galaxy diversity. No statistically significant periodic signals are detected in M81 or Cen A, which, along with several periodic signals previously found among the M31 sources, can be understood considering a joint effect of our detection sensitivity and intrinsic distributions of the orbital period and X-ray luminosity of LMXBs. The ensemble of short-duty-cycle flares, a mix of recurrent and isolated dips, and galaxy-dependent rms--flux factor, supports a picture in which stochastic accretion-rate fluctuations modulate luminosity on $\sim$10-$10^{4}$ s. Conducted at known distances and across distinct host environments, this extragalactic survey provides uniform flare/dip samples and rms-flux scalings for bulge-dominated fields, offering empirical constraints for accretion physics and illustrating the promise of timing analyses in external galaxies using the Chandra archive.

Figures (7)

• Figure 1: Stacked Chandra/ACIS 0.5--8 keV counts images of M31 (upper left), M81 (upper right), and Centaurus A (lower panel). X-ray sources showing flares (F) or dips (D) are marked with green circles; labels correspond to the IDs in Tables \ref{['tab:flare']} and \ref{['tab:dip']}. The bulge region used in our analysis is outlined in cyan (see Section \ref{['sec:data']}).
• Figure 2: Examples of flares. The original light curve is shown as black data points with error bars. The Bayesian Blocks are shown as blue solid lines. The source information is provided at the upper left corner of each panel.
• Figure 3: Examples of dips. The original light curve is shown as black data points with error bars. The Bayesian Blocks are shown as blue solid lines. The source information is provided at the upper left corner of each panel.
• Figure 4: Left: Histograms of flare durations. Right: Histograms of flare peak luminosities. Colors denote host galaxies—blue: M31, red: M81, green: Cen A. Durations are defined as the temporal extent of the Bayesian-Blocks–identified flare segment 2013ApJ...764..167S. For each flare, the corresponding block is subdivided into five equal time bins; the maximum bin count rate is adopted as the peak, and converted to a 0.5–8 keV luminosity(see Section \ref{['sec:data']} for spectral assumptions). Distributions are shown with logarithmic binning.
• Figure 5: Histogram of the flare rates measured in this work (orange dashed line) and the Type I burst rates from the MINBAR sample (blue solid line). Only sources with a well-defined rate (i.e., those for which the mean burst or flare rate is constrained rather than a upper limit, and with at least two detected events) are included in both distributions.