The cold molecular gas regulates the activity of active galactic nuclei in massive galaxies

1. Problem Statement

The physical mechanisms regulating Active Galactic Nuclei (AGN) feedback in massive galaxies remain a subject of debate. While it is widely accepted that supermassive black holes (SMBHs) co-evolve with their host galaxies, the specific driver of AGN activity is unclear.

• The Debate: Does AGN activity depend primarily on the black hole mass ($M_{BH}$) or the availability of cold molecular gas ($M_{H2}$)?
• Context: Previous studies have yielded conflicting results. Some suggest a strong link between jet power and Bondi accretion rates (implying hot gas accretion), while others find correlations with molecular gas mass. Conversely, some studies report no significant dependence of AGN activity on molecular gas properties.
• Gap: Many previous analyses relied on small sample sizes, potentially leading to spurious correlations. There is a need for a large-sample study to disentangle the effects of molecular gas mass from other correlated parameters like stellar mass ($M_*$), bulge mass ($M_{bulge}$), and black hole mass.

2. Methodology

The authors conducted a comprehensive statistical analysis using a large, multi-source sample of nearby AGNs.

• Sample Selection:

  • Size: 411 AGNs with available measurements for molecular gas mass ($M_{H2}$), stellar mass ($M_*$), star formation rates (SFR), and black hole mass ($M_{BH}$).
  • Sources: Data were aggregated from multiple surveys, including Palomar–Green quasars, Swift-BAT AGNs, the CARS survey, the QFS survey, xCOLD GASS, Arizona Radio Observatory, and MaNGA.
  • Constraints: The sample includes both early-type ($T \leq 0$) and late-type ($T > 0$) galaxies, with stellar masses ranging from $10^{9.5} M_\odot$ to $10^{11.5} M_\odot$.
  • Radio Selection: To isolate nuclear activity from star formation, the authors selected "radio excess" sources where the observed radio luminosity significantly exceeds the expected luminosity from star formation ($L_{radio} \gg L_{SF}$).

• Tracers of AGN Activity:

  • Radio Luminosity: Measured at 1.4 GHz and 3 GHz (tracers of jet activity).
  • [O III] Luminosity: Tracer of ionizing radiation from the AGN.
  • 5100 Å Luminosity: Tracer of accretion disk continuum (estimated via H$\beta$ luminosity).

• Statistical Analysis:

  • Linear Regression: Used the LINMIX package to perform linear regression accounting for measurement uncertainties in both variables and intrinsic scatter.
  • Correlation Analysis: Used PYMCCORRELATION with bootstrapping to calculate Spearman's rank correlation coefficients ($r$) and p-values.
  • Partial Correlation: Used pingouin.partial_corr to determine the intrinsic relationship between variables while controlling for confounding factors (e.g., correlating $L_{radio}$ with $M_{H2}$ while holding $M_{BH}$ and $M_*$ constant).
  • Bias Correction: Addressed Malmquist bias by analyzing low-redshift subsamples separately.

3. Key Contributions

1. Large-Scale Empirical Evidence: Provided one of the largest samples to date (411 AGNs) to test the relationship between AGN power and cold gas reservoirs, moving beyond the limitations of small-sample studies.
2. Disentangling Drivers: Systematically compared the correlation strength of AGN luminosity against $M_{H2}$, $M_{BH}$, $M_*$, and $M_{bulge}$ using partial correlation analysis to identify the primary driver.
3. Morphological Differentiation: Analyzed differences between early-type and late-type galaxies, finding that the gas-AGN correlation is stronger in early-type systems.
4. Feedback Implications: Linked AGN activity not just to fueling but also to star formation, suggesting a potential positive feedback loop in massive galaxies.

4. Key Results

A. Molecular Gas is the Primary Driver

• Strong Correlations: Radio luminosity ($L_{1.4/3GHz}$) and [O III] luminosity ($L_{[O III]}$) exhibit strong, significant correlations with molecular gas mass ($M_{H2}$) and molecular gas content ($M_{H2}/M_*$).
  • Example: $L_{1.4GHz}$ vs. $M_{H2}$ yields $r = 0.55$ ($P \approx 10^{-20}$).
• Weakness of Other Parameters: Correlations between AGN luminosity and $M_{BH}$, $M_*$, or $M_{bulge}$ are significantly weaker.
  • Example: $L_{1.4GHz}$ vs. $M_{BH}$ yields $r = 0.28$ ($P \approx 10^{-6}$).
• Partial Correlation Dominance: When controlling for $M_{BH}$ and $M_*$, the correlation between AGN luminosity and molecular gas remains robust (e.g., $r \approx 0.53$ for $L_{1.4GHz}$ vs. $M_{H2}$). Conversely, when controlling for molecular gas, the correlations between AGN luminosity and $M_{BH}$ or $M_*$ diminish significantly or vanish (e.g., $r$ drops to $\sim 0.14$ for $L_{1.4GHz}$ vs. $M_*$).
• Conclusion: Molecular gas mass is the most tightly correlated property with AGN activity, suggesting it is the most important driver of nuclear activity.

B. Morphological Differences

• The correlation between AGN activity and molecular gas is stronger in early-type galaxies than in late-type galaxies. This suggests that AGN feedback mechanisms (specifically radio mode) may be more efficient or dominant in gas-poor, massive ellipticals where cold gas is a scarce but critical resource.

C. Connection to Star Formation

• Positive Correlation: There is a significant correlation between AGN luminosity (Radio, [O III], 5100 Å) and Star Formation Rates (SFR), even after controlling for $M_{BH}$ and $M_*$.
• Synchronized Growth: This supports the "co-evolution" scenario where the same cold gas reservoir fuels both the central black hole and star formation.
• Positive Feedback: The authors suggest that AGN activity might have a positive feedback influence on star formation in these massive galaxies, potentially mediated by jet interactions or magnetic fields, rather than purely negative (quenching) feedback.

D. Robustness Checks

• The results hold true even after correcting for Malmquist bias by restricting the analysis to low-redshift sources.
• The slopes of the $L_{AGN}$ vs. $M_{H2}$ relations are consistent with theoretical predictions (e.g., Kennicutt-Schmidt law and Hickox et al. models) where AGN luminosity scales linearly with gas mass and SFR.

5. Significance

• Regulation Mechanism: The study provides strong observational evidence that cold molecular gas mass is the fundamental regulator of AGN activity in massive galaxies, superseding black hole mass as the primary predictor of current nuclear power.
• Model Constraints: These findings challenge models that rely solely on Bondi accretion of hot gas or $M_{BH}$-dependent prescriptions. Instead, they support models where cold gas accretion is the dominant fueling mechanism, even in early-type galaxies.
• Feedback Paradigm: The results suggest a more complex feedback loop where AGN activity and star formation are synchronized and potentially mutually reinforcing (positive feedback) in the presence of abundant cold gas, rather than a simple "quenching" scenario.
• Future Simulations: The observed correlations provide a direct empirical basis for refining gas accretion sub-grid physics in cosmological simulations (e.g., IllustrisTNG, EAGLE), moving away from purely empirical or hot-gas accretion models toward cold-gas driven prescriptions.