Ruling Out Compact Jets as the Dominant Source of Radio Emission in Radio-quiet, High Eddington-ratio Active Galactic Nuclei

TL;DR

This study investigates whether compact jets dominate radio emission in radio-quiet, high/ super-Eddington AGNs by examining the radio/X-ray luminosity ratio \\mathcal{R}_{\\rm X} across a broad black-hole mass range \\bigl( M_{\\rm BH} \\sim 10^{5}\\-10^{9} M_\\odot \\bigr). The authors adapt jet and corona scaling relations to the slim-disk regime and account for X-ray weakness and saturated bolometric luminosities at high \\ell_{\\rm Edd}. Their analysis shows the observed \\mathcal{R}_{\\rm X} distribution is inconsistent with a jet-dominated origin, signaling that radio emission is more likely corona-dominated with possible contributions from strong outflows at higher accretion rates. The results challenge the applicability of the standard fundamental plane to this regime and emphasize the role of coronal processes and winds in shaping radio output, motivating deeper, high-resolution, multiwavelength follow-up to disentangle the contributing components.

Abstract

The origin of core radio emission in radio-quiet active galactic nuclei (AGNs) is still actively debated. General relativistic magnetohydrodynamics simulations often predict the launching of moderately large-scale jets from super-Eddington accretion flows, but this prediction seems at odds with observations indicating most high/super-Eddington AGNs appear radio quiet. Here, we use the ratio of radio to X-ray luminosities as a multiwavelength diagnostic to probe the origin of radio emission in a sample of 69 radio-quiet, high/super-Eddington AGNs with black-hole masses $M_{\rm BH} \sim 10^{5}-10^{9}~M_\odot$. With this wide dynamic range in $M_{\rm BH}$, we adapt existing formalisms for how jetted radio emission and accretion-powered X-ray emission scale with black hole mass into the super-Eddington regime. We find that the radio/X-ray luminosity ratios observed across this $M_{\rm BH}$ range are inconsistent with a jet-dominated model for radio emission. We discuss how our results may instead be consistent with a corona-dominated radio emission origin with a contribution from outflows at higher accretion rates.

Figures (5)

• Figure 1: Histograms showing the distributions of key sample properties. The full sample is shown by the black outline, and within each bin the subsample distributions are shown by colored bars (note that placement within each bin is based simply on sample order and does not necessarily reflect a fine x-axis value). In the lower-left panel we show the observed values of $\mathcal{R}_{\rm X}$, while in the lower-right panel we show $\mathcal{R}_{\rm X}$ corrected for X-ray weakness (where applicable; see Section \ref{['subsec:data_lum']}).
• Figure 2: Comparison of $\ell_{\rm Edd}$ to $\dot m$ as found using Eq. (\ref{['eq:mdot_saturated']}). Our sample is shown by symbols given in the legend (categorized by subsample; Section \ref{['subsec:samp3']}). The grey bars illustrate error on $\dot m$ from a $\pm$0.5 dex variance in $\ell_{\rm Edd}$. The black curve shows the relationship given by Eq. (8) of Mineshige00, modified to use $\ell_{\rm Edd} = \mathcal{F}_{\rm inc} L/L_{\rm Edd}$, where $L/L_{\rm Edd}$ is the intrinsic Eddington ratio and $\mathcal{F}_{\rm inc}$ is our adopted inclination factor (see Eq. \ref{['eq:inc_fact']}).
• Figure 3: $f_{\rm weak}$-corrected $\log \mathcal{R}_{\rm X}$ of objects in our sample vs. the model $\log \mathcal{R}_{\rm X}$ calculated as a function of their $M_{\rm BH}$ and $\dot m$ (Eq. \ref{['eq:FP_LRLX']}). The model has been normalized to our sample mean values of $\langle \log \mathcal{R}_{\rm X} \rangle \approx -5$, $\langle M_{\rm BH} \rangle \approx 10^{7} M_\odot$, and $\langle \log \dot m \rangle\approx 0$. The sample is shown by symbols given in the legend and is colorized by $\log \ell_{\rm Edd}$ (color bar). 1$\sigma$ errors were derived using the Monte Carlo scheme described in Section \ref{['subsec:data_err']} (arrows denote limits from nondetections). The green dot and bars to the left of the legend illustrate the average uncertainty for modeled $\mathcal{R}_{\rm X}$. The blue dashed line and shaded area show the 1:1 relationship between the axes with a 1 dex scatter. Were radio emission dominated by a jet, we would expect the sample to follow the 1:1 slope. Instead, the distribution of $\mathcal{R}_{\rm X}$ in our sample is effectively flat (with large scatter; see Section \ref{['sec:discussion3']}), as illustrated by the black dotted line.
• Figure 4: Comparison of $\log \mathcal{R}_{\rm X}$ with $\log \ell_{\rm Edd}$. The sample is shown by symbols given in the legend. 1$\sigma$ errors were derived using the Monte Carlo scheme described in Section \ref{['subsec:data_err']} (arrows denote limits from nondetections). The orange line shows an ordinary-least-squares fit to our sample, with a correlation confidence level of $p_{\rm null} = 0.04 \pm 0.03$. We do not expect this fit to extend to lower values of $\ell_{\rm Edd}$; it is more likely that $\mathcal{R}_{\rm X}$ is generally flat for $\log \ell_{\rm Edd} \lesssim -0.5$ but increases (albeit with greater scatter) at higher $\ell_{\rm Edd}$.
• Figure 5: Our sample examined in the context of the fundamental plane of black hole accretion. (a): Result of an ordinary-least-squares fit (dashed line) to our sample, using our Monte Carlo scheme to account for error and nondetections (Section \ref{['subsec:data_err']}). Our sample is shown by symbols given in the legend (categorized by subsample; Section \ref{['subsec:samp3']}) and is colorized by $\ell_{\rm Edd}$ (color bar). The light blue diamond shows POX 52, a massive black hole ($M_{\rm BH} \sim 1.6 \times 10^{5} M_\odot$) which was recently detected in radio for the first time by Yuan25. Arrows denote limits from radio or X-ray nondetections. (b): Fit parameters $\xi_{\rm RX}$ and $\xi_{\rm RM}$ for our sample (green star) compared to a number of other works in the literature (purple symbols), including Merloni03; radio-quiet objects at $z < 0.13$ from Li08; radio-quiet objects at $z < 1.5$ from Bariuan22; and radio-quiet objects from Wang24, limited by $0.1 < z < 1$ (denoted "Wang+ 2024 (z)", including both low- and high-$\ell_{\rm Edd}$ objects) or by $\ell_{\rm Edd} > 0.03$ (denoted "Wang+ 2024 (L)", including their full range of $0.1 < z < 4$). Model-predicted ranges of the fit parameters are shown by circles with connecting dashed lines that trace variation in the radio spectral index (open circles assume a flat $\alpha_{\rm r} = 0$ and filled circles assume a steep $\alpha_{\rm r} = -0.5$) for three origins of X-ray emission: advection-dominated accretion flow (ADAF), jet, and standard disk/corona (e.g., Sections 5.2 and 5.3 of Merloni03). Note that all three scenarios interpret radio emission as originating from a synchrotron jet, and all of the radio-quiet samples are inconsistent with these scenarios.