Systematic Monitoring of Extreme X-ray Variability from Weak-line Quasars

Abstract

We present the results of a multi-cycle Chandra program to systematically monitor the X-ray variability of 10 weak-line quasars (WLQs) that previously had limited multi-epoch X-ray observations. Three new Chandra 2.8 to 8.2 ks observations were obtained for each WLQ with C$\,$IV rest-frame equivalent widths (REWs) $\lesssim 10$ Å, substantially improving the monitoring data quality of WLQs and our ability to characterize their long-term X-ray variability behavior. We observe recurrent extreme X-ray variability in the historically variable WLQ SDSS J1539+3954, with an X-ray flux rise of a factor of $\gtrsim 6$ between 2023 and 2024 ($\gtrsim 21$ relative to 2013). Another previously X-ray weak WLQ in the sample, SDSS J0825+1155, underwent a significant X-ray flux variation by a factor of $\gtrsim 14$ between 2019 and 2023. We find the fraction of WLQs exhibiting evidence of extreme X-ray variability to be $0.20^{+0.17}_{-0.07}$. In the context of the thick disk and outflow (TDO) model, the substantial fraction of WLQs displaying extreme X-ray variability may suggest that the variability is driven by the intrinsic motion of the TDO wind rather than changes in the height of the TDO disk. We performed a statistical comparison between the distribution of variability amplitudes of WLQs and general radio-quiet quasars. We find that these underlying distributions are statistically different, with WLQs having $\approx 6.8$ times higher odds of exhibiting an extreme X-ray variability event than the general radio-quiet quasar population.

Figures (14)

• Figure 1: Long-term 0.5--2 keV X-ray light curves for the sample of WLQs. Solid points indicate significant source detections with 1$\sigma$ error bars. Points with downward arrows represent 90% confidence level upper limits on the flux from instances of non-detections.
• Figure 2: Maximum 0.5--2.0 keV X-ray flux-factor change. We are unable to constrain the maximum flux-factor change for two of ten WLQs in our sample (SDSS J110409.96+434507.0 and SDSS J163810.07+115103.9)
• Figure 3: Logarithm of the full-band count flux ratio between two observations as a function of the rest-frame time difference between the start of the observations. The down-sampled population of radio-quiet quasars from 2020MNRAS.498.4033T is shown in blue, with circles as detected values. Our sample of 10 WLQs is shown in pink, with detected values marked as squares. The upper and lower limits of both samples are represented with upward (downward) pointing triangles as upper (lower) limits. The 2020MNRAS.498.4033T sample contains 556 observation pairs across 216 quasars, and the WLQ sample contains 59 observation pairs of 8 quasars. We note that 2 of the WLQs in the sample were undetected across all epochs, and as such we are unable to constrain the amount of variability between epochs for those objects. The dotted grey lines indicate the minimum and maximum values of $\Delta t$ of the chosen time bin to compare the populations. $\Delta t_{\mathrm{min}} = 6.44 \ \mathrm{Ms}$ and $\Delta t_{\mathrm{max}} = 169.11 \ \mathrm{Ms}$.
• Figure 4: Distribution of full-band count flux ratios for observations of time separations $6.44 \ \mathrm{Ms} \leq \Delta t \leq 169.11 \ \mathrm{Ms}$. The arrows represent the X-ray limits and their direction and the $3\sigma_{\mathrm{MAD}} = 0.536$ threshold is denoted by the dashed orange lines.
• Figure 5: ZTF g/r/i-band light curves of the sample of WLQs extracted from ZTF's forced photometry service (ZFPS) 2019PASP..131a8003M. The vertical lines mark the dates of the Chandra observations of a given WLQ. Any Chandra epochs that pre-date the ZTF survey are noted in the top left corner of each set of light curves.