CIV wind properties of the SDSS-V X-ray selected quasars: strong optical-to-UV emission is key regardless of X-ray strength

1. Problem Statement

The driving mechanisms of wide-angle accretion disc winds in Active Galactic Nuclei (AGN) remain a subject of debate. While radiation line-driving is an established model, the specific spectral energy distribution (SED) conditions required to launch these winds are unclear. Previous studies, largely based on optically selected quasar samples (e.g., SDSS-IV), have suggested that winds are driven by a balance between ionizing photons and the gas's ability to absorb them. Specifically, an over-abundance of ionizing X-ray photons (hard SEDs) may over-ionize the gas, suppressing line-driving.

However, most large quasar samples are optically selected, which may introduce biases (e.g., favoring bright accretion discs over strong X-ray coronae). There is a lack of large-scale, X-ray selected samples to test if the relationship between X-ray properties and wind strength holds when the selection is biased toward X-ray bright sources. Furthermore, it is unclear if the X-ray luminosity ($L_{2\text{keV}}$) or the spectral slope ($\Gamma$) directly correlates with wind strength, or if the correlation is driven by other SED components (optical/UV).

2. Methodology

Sample Selection:

• Data Source: The study utilizes a sample of 3,027 quasars detected in the deepest available eROSITA data (eFEDS and eRASS4:5) and observed by the SDSS-V (Sloan Digital Sky Survey V) Black Hole Mapper program.
• Redshift: $1.5 \leq z \leq 3.5$, ensuring the C IV $\lambda1549$ emission line falls within the observed wavelength range.
• Selection Criteria: Objects were selected based on X-ray detections (eROSITA, Chandra, XMM-Newton, Swift) and subsequent optical spectroscopy. The sample includes Broad Absorption Line (BAL) quasars (7% of the sample), which were retained as they do not qualitatively alter the conclusions.
• Redshift Correction: The authors implemented a cross-correlation algorithm (using 33 templates) to correct SDSS pipeline redshifts, which are often biased by asymmetric, blueshifted C IV lines. This improved the accuracy of C IV blueshift measurements.

Parameter Derivation:

• Wind Tracers: The strength of accretion disc winds is parametrized using the C IV emission space: C IV blueshift (velocity) versus Equivalent Width (EW). A derived metric, C IV distance, was used to reduce this 2D space to a single dimension representing wind strength (higher distance = stronger wind/larger blueshift).
• SED Characterization:
  • Optical: Monochromatic luminosity at 2500 Å ($L_{2500}$).
  • X-ray: Monochromatic luminosity at 2 keV ($L_{2\text{keV}}$) and spectral slope ($\Gamma$) derived via Bayesian X-ray spectral fitting (using BXA and XSpec).
  • Far-UV (Proxy): Since direct far-UV observation is impossible from the ground, the authors used the He II $\lambda1640$ line luminosity ($L_{\text{He II}}$) as a proxy for the ionizing far-UV continuum ($L_{54\text{eV}}$).
• Spectral Slopes:
  • $\alpha_{\text{ox}}$: Slope between $L_{2500}$ and $L_{2\text{keV}}$.
  • $\alpha_{\text{ouv}}$: Slope between $L_{2500}$ and $L_{54\text{eV}}$ (Optical-to-UV).
  • $\alpha_{\text{uvx}}$: Slope between $L_{54\text{eV}}$ and $L_{2\text{keV}}$ (UV-to-X-ray).
• Statistical Analysis: The authors employed Hierarchical Bayesian Modelling (HBM) to infer parent distributions of X-ray parameters ($\Gamma$) across bins of C IV properties, mitigating issues with low-count X-ray spectra. They also performed 2-D Kolmogorov-Smirnov (KS) tests to compare distributions between X-ray and optically selected samples.

3. Key Contributions

1. First Large X-ray Selected C IV Analysis: This is one of the largest studies (N=3027) of C IV wind properties specifically in an X-ray selected sample, allowing for a direct comparison with previous optically selected samples (e.g., SDSS-IV).
2. Decoupling X-ray and Wind Properties: The study provides robust evidence that X-ray properties ($L_{2\text{keV}}$ and $\Gamma$) are not the primary drivers of wind strength in quasars.
3. Identification of the True Driver: The authors demonstrate that the optical-to-UV luminosity ratio ($\alpha_{\text{ouv}}$) is the strongest correlate with wind strength, superseding the traditional $\alpha_{\text{ox}}$ parameter.
4. Resolution of Selection Bias: The paper explains why X-ray selected samples appear "wind-poor" (lower blueshifts) compared to optical samples: it is a secondary effect of the correlation between optical, UV, and X-ray luminosities, rather than a fundamental difference in the physics of the winds.

4. Key Results

• X-ray vs. Wind Independence: There is no strong correlation between X-ray spectral properties ($L_{2\text{keV}}$, $\Gamma$) and C IV wind strength (blueshift or distance). The Spearman correlation coefficients are negligible.
• Shift in C IV Space: The X-ray selected sample is shifted toward lower C IV blueshifts and higher Equivalent Widths compared to optically selected samples.
  • This shift persists even after matching samples for redshift, optical luminosity, and Eddington ratio.
  • It is not solely due to eROSITA sensitivity limits; intrinsic differences in the populations exist.
• The Role of $\alpha_{\text{ouv}}$:
  • The parameter $\alpha_{\text{ouv}}$ (Optical-to-UV slope) shows the strongest correlation with C IV distance (Spearman $\rho \approx -0.67$).
  • Stronger winds are associated with more negative $\alpha_{\text{ouv}}$, meaning the optical luminosity is high relative to the far-UV luminosity.
  • The correlation between $\alpha_{\text{ox}}$ and wind strength is driven by the optical luminosity component, not the X-ray component.
• SED Evolution with Wind Strength:
  • As wind strength increases, Optical Luminosity ($L_{2500}$) increases significantly.
  • Far-UV Luminosity ($L_{54\text{eV}}$) initially increases but then decreases for the strongest winds (highest blueshifts).
  • X-ray Luminosity ($L_{2\text{keV}}$) remains roughly constant or increases only slightly; it does not need to change to drive stronger winds.
• Black Hole Mass ($M_{\text{BH}}$) and Eddington Ratio:
  • Stronger winds are found in higher $M_{\text{BH}}$ sources (above $\lambda_{\text{Edd}} \sim 0.2$).
  • This is consistent with a standard Shakura-Sunyaev disc model where higher mass black holes have cooler inner discs (less UV emission), leading to a favorable optical-to-UV ratio for line-driving.

5. Significance and Physical Interpretation

The results support a radiation line-driven wind model with the following physical implications:

• Ionization Balance: To launch a wind, the gas must not be over-ionized. The study suggests that far-UV photons (responsible for ionizing C IV) must not be too abundant relative to the optical continuum.
• X-ray Role: Hard X-ray photons are too few in number to significantly affect the ionization state of the wind-launching gas. Therefore, X-ray weakness is not a prerequisite for strong winds; rather, the suppression of far-UV emission (relative to optical) is the key factor.
• Three-Component SED: The findings favor a three-component SED model (Optical Disc + Warm Comptonizing Component + Hard X-ray Corona) over a simple two-component model. The "warm" component (or the lack of a strong far-UV continuum) plays a critical role in allowing line-driving to occur.
• Selection Bias Explanation: The "missing" high-blueshift quasars in X-ray selected samples are not absent because they lack X-rays, but because the specific SED configuration required for the strongest winds (High Optical / Low UV) is statistically less likely to be detected in X-ray surveys due to the coupling of these luminosities.

Conclusion: The strength of accretion disc winds in quasars is governed by the optical-to-UV luminosity ratio, not the X-ray strength. Strong winds require a "soft" far-UV continuum relative to the optical emission, ensuring the gas remains partially ionized and susceptible to radiation line-driving. X-ray properties are largely decoupled from this process.