Questing a Coherent Definition of Weak-line Quasars and its Physical Implications

TL;DR

The paper addresses the lack of a universal, physically meaningful definition for weak-line quasars (WLQs) by analyzing a large SDSS DR16 quasar sample and identifying outliers in three UV relations: the L1350Å–C IV blueshift relation, the Baldwin effect, and the log $L_{2500Å}-\alpha_{\rm ox}$ relation.A data-driven definition emerges from two CIV EW breakpoints, $EW(C\ IV) = 8.9\pm0.2$ Å and $EW(C\ IV) = 19.3\pm0.3$ Å, classifying WLQs as $EW(C\ IV)<8.9$ Å, normal quasars as $EW(C\ IV)>19.3$ Å, and bridge quasars as intermediate objects, with substantial WLQ and bridge fractions in DR16Q.The authors find that WLQs exhibit stronger attenuation of high-ionization lines than low-ionization lines, and that line attenuation correlates positively with ionization energy, consistent with a shielding gas model where a puffed inner disk (slim-disk regime) blocks high-energy photons from reaching the BLR.Evidence for a unified physical picture includes a roughly equal split between X-ray weak and X-ray normal WLQs, composite spectra showing diminished HILs relative to LILs, and the proposed link between accretion-state transitions and EW(C\ IV) through bridge quasars, highlighting the role of accretion-driven geometry in quasar emission.

Abstract

Weak-line quasars (WLQs) are a subset of type 1 quasars with remarkably weak high-ionization broad emission lines but normal optical/UV continua. Using 371,091 quasars from SDSS DR16, we define WLQs by analyzing outliers in three relations: the L1350-CIV blueshift, the Baldwin effect, and the logL2500-alpha_ox. We find two CIV EW thresholds: $8.9\pm0.2$Å and $19.3\pm0.3$Å. WLQs (EW(CIV)<$8.9\pm0.2$Å) have enhanced CIV blueshifts, deviate from the Baldwin effect, and include many X-ray weak objects (nearly half). Normal quasars (EW(CIV)>$19.3\pm0.3$Å) show typical properties, while bridge quasars (intermediate EW) are transitional. WLQs show a positive correlation between line attenuation and ionization energy: high-ionization lines (e.g., HeII, CIV) are suppressed by ~3-4σ compared to low-ionization lines (e.g., MgII, OI). This supports the shielding gas model, where a thick inner accretion disk obscures high-energy photons, suppressing high-ionization lines, while low-ionization lines are less affected. We suggest that WLQs and normal quasars correspond to slim and thin disk regimes, respectively, with bridge quasars as a transitional phase. This work provides a unified criterion for WLQs and highlights the role of accretion-driven shielding gas in their spectral features.

Figures (12)

• Figure 1: EW distributions of HILs. In each panel, the black histogram and red lines are the EW distribution and its best-fit lognormal model, respectively. The vertical dash and dash-dotted lines are the $2\sigma$ and $3\sigma$ lines of their lognormal distributions, respectively. In addition, the redshift coverage of each HIL, the number of sources, mean value and the $3\sigma$ range of the best-fit lognormal distribution are depicted in the upper left corner of each panel. As the figure shows, each HIL EW distribution exhibits a prominent tail toward low EW values, which are consistent with the results in 2009ApJ...699..782D and 2012ApJ...747...10W.
• Figure 2: EW Distributions of LILs. The legend is the same as that in Figure \ref{['HILs_distributions']}.
• Figure 3: EW distributions of narrow emission lines. The legend is the same as that in Figure \ref{['HILs_distributions']}. All narrow lines show a tail toward high EW values but none toward low EW values, which are exactly opposite to those of HILs.
• Figure 4: The relationship between C iv blueshift and monochromatic luminosity at 1350 Å in the rest frame $L_{1350\rm \textup{\AA}}$. The red contours from darker to lighter represent regions containing 68%, 95% and 99.7% of the sources respectively. The black solid line represents the linear equation fitted by MCMC method. The black dashed line marks the boundary of outliers for the $L_{1350\rm \textup{\AA}}-$ C iv blueshift relation. Region below this dashed line encompasses 95% of the objects, while sources above the line will be considered outliers.
• Figure 5: The fraction of outliers in the $L_{1350\rm \textup{\AA}}-$ C iv blueshift relation as a function of EW(C iv). We divide the sources based on their EW(C iv) using overlapping bins to account for the limited number of sources at small EWs. The standard deviation of EW(C iv) within each bin is shown as the x-axis error, while the y-axis error represents the binomial uncertainty. The three segments of the dark blue polyline represent the best fit linear relation in each region, while the vertical blue lines with lighter shaded areas represent location of the two breakpoints and their uncertainty ranges.