LLAGN: Low-Luminosity Active Galactic Nuclei

• LLAGN are active galactic nuclei with bolometric luminosities below 10^42–10^43 erg/s and Eddington ratios of 10^-3 to 10^-4, dominating the local SMBH population.
• They exhibit distinct spectral energy distributions lacking a prominent UV 'big blue bump', with emissions driven by nonthermal jets and radiatively inefficient flows.
• Multiwavelength diagnostics, including JWST IFU spectroscopy, reveal truncated disks, optically thin dust structures, and jet-dominated feedback impacting host galaxy evolution.

Low-luminosity active galactic nuclei (LLAGN) are accreting supermassive black holes (SMBHs) with bolometric luminosities $L_{\rm bol}\lesssim 10^{42}$–$10^{43}$ erg s$^{-1}$ and Eddington ratios $$\lambda_{\rm Edd}\equiv L_{\rm bol}/L_{\rm Edd}\lesssim10^{-3}$$–$10^{-4}$. They constitute the dominant AGN population in the local universe and exhibit distinctive spectral energy distributions (SEDs), physical accretion structures, and feedback modes that set them apart from luminous Seyferts and quasars. Their characteristic signatures include a weak or absent thermal UV “big blue bump,” prominent nonthermal emission components associated with jets and hot, radiatively inefficient accretion flows (RIAFs/ADAFs), and IR-to-X-ray properties that challenge the canonical AGN unification paradigm.

1. Defining Properties and Demographics

LLAGN are identified on the basis of their nuclear luminosity ($L_{\rm bol}\lesssim10^{42}$–$10^{43}$ erg s$^{-1}$), their low $\lambda_{\rm Edd}$ (typically $\ll10^{-3}$), and their host-galaxy spectroscopic signatures—principally as low-ionization nuclear emission line regions (LINERs) or low-luminosity Seyferts. Local surveys (e.g., the Palomar sample) find that LLAGN constitute the majority of AGN in $z\approx0$ galaxies, with radio, optical, and X-ray detections indicating a prevalence of SMBHs accreting at highly sub-Eddington rates (Nemmen et al., 2011, Saikia et al., 2018).

Broad-band SED decomposition (radio through X-ray) demonstrates that their emission is typically dominated by nonthermal components lacking the luminous, optically thick disk and torus found in bright AGN (Fernández-Ontiveros et al., 2022, Mason et al., 2012). Key distinguishing features include:

• Suppressed optical/UV continuum: no “big blue bump.”
• Compact, flat-spectrum radio cores and high-frequency turnovers ($\nu_{t}\sim0.2$–$30$ THz), attributed to self-absorbed jet synchrotron emission, often surpassing traditional disk contributions (Fernández-Ontiveros et al., 2022).
• A preponderance of parsec- and larger-scale radio jets, with jet kinetic power matching or exceeding $L_{\rm bol}$ (Mezcua et al., 2014).
• Diverse IR morphologies, reflecting a dissipated or absent torus and the emergence of optically thin, low dust-to-gas environments (Mason et al., 2012, Mason et al., 2012, Goold et al., 23 Jan 2026).
• Hard X-ray spectra ($\Gamma\sim1.5$–1.8) with weak or undetectable Compton reflection, truncated-disk features, and sparse Fe K$\alpha$ lines (Younes et al., 2018).

2. Accretion Physics and SED Modeling

The consensus physical paradigm for LLAGN accretion is the dominance of radiatively inefficient flows (RIAFs/ADAFs) in the inner regions, sometimes coupled to a truncated outer thin disk and relativistic jets. ADAFs/RIAFs differ radically from the standard thin-disk models: the flow is hot, optically thin, and two-temperature, allowing most of the accretion energy to escape via advection and outflows, with only minor fractions radiated (Nemmen et al., 2013, Nemmen et al., 2011).

Model Architecture

• Inner zone ($r\lesssim R_{\rm tr}$): RIAF characterized by low density, high temperature ($T_e\sim10^{9-12}\,$K), and strong mass loss via winds (Shi et al., 12 Feb 2025).
• Outer disk ($r > R_{\rm tr}$): Truncated standard disk, peaking thermally in NIR or MIR if $\dot m$ is sufficient (Nemmen et al., 2013).
• Jet: Relativistic outflow with internal-shock particle acceleration, supplying flat- or steep-spectrum synchrotron and self-Compton emission (Nemmen et al., 2011, Fernández-Ontiveros et al., 2022).

The ratio of jet power to bolometric luminosity, $P_{\rm jet}/L_{\rm bol}\gtrsim 1$–10, signals kinetic dominance even at extremely low Eddington ratios. Jet mass loss rates are $\dot m_{\rm jet}/\dot m\sim10^{-4}$–$10^{-2}$, and observed radio luminosities are robustly predicted only when the jet is included.

SED Signatures

• Radio–IR: Jet self-absorbed synchrotron, smoothly breaking to optically thin slopes in the IR-optical band with $-3.7<\alpha_0<-1.3$; turnover frequencies increase for hotter, more compact jet bases (Fernández-Ontiveros et al., 2022).
• NIR–MIR: Emission may arise from truncated disks, optically thin dust, and/or the high-frequency tail of jet synchrotron, with lack of a strong, Seyfert-like dust bump unless $\lambda_{\rm Edd}$ approaches $10^{-3}$ (Mason et al., 2012, Dumont et al., 2019, Mason et al., 2012).
• Optical–UV: Depletion or extreme steepness; dominated by nonthermal components without a “blue bump.”
• X-ray: Comptonized (RIAF) or synchrotron/SSC (jet), with hard spectra and low reflection fractions, consistent with a truncated disk (Younes et al., 2018).
• $\alpha_{\rm ox}$ vs. $L_{\rm UV}$:
  • LLAGN show a positive $\alpha_{\rm ox}$–$L_{\rm UV}$ correlation, but an anti-correlation with Eddington ratio below $\lambda_{\rm Edd}\sim 10^{-3}$, reflecting the transition from radiatively efficient disks to ADAFs (Xu, 2011, López et al., 2024).

Breakdowns of SED modeling for large samples (e.g., LINERs) consistently recover $\dot m\sim10^{-4}-10^{-2}$, jet fractions $P_{\rm jet}/L_{\rm bol}\sim 5$, and radiative efficiencies $\eta_{\rm rad}\sim10^{-4}$–$10^{-3}$ (Nemmen et al., 2013, Nemmen et al., 2011).

3. Infrared–Optical–UV Diagnostics and Torus Evolution

High-resolution IR imaging and spectroscopy reveal no evidence for a classical, optically thick, Seyfert-like torus in most LLAGN (Mason et al., 2012, Mason et al., 2012, Goold et al., 23 Jan 2026). Instead:

• Mid-IR: Compact nuclear point sources are detected in higher $\lambda_{\rm Edd}$ LLAGN, but many objects show flat MIR–optical slopes, strong silicate emission, and a reduction in dust-to-gas ratios, consistent with optically thin, dissipating dust structures (Mason et al., 2012, Goold et al., 23 Jan 2026).
• Silicate Features: Strong emission is frequent and is found even in type 2 nuclei, an indicator of optically thin or vertically extended dust shells; the strength-to-gas ratio $S_{10}/N_H$ exceeds Seyfert values (Mason et al., 2012, Mason et al., 2012).
• Star Formation: Host-dominated LLAGN at low $\lambda_{\rm Edd}$ exhibit extended PAH bands in the IR, signaling circumnuclear star formation, but this is not a universal feature (Mason et al., 2012).
• NIR: Nuclear hot-dust emission ($T_{\rm dust}\sim600-1500\,$K) is ubiquitous and may exceed the X-ray luminosity at $\lambda_{\rm Edd}<10^{-4}$, and can be produced via either truncated disk thermalization or jet processes (Dumont et al., 2019).

JWST nuclear IFU spectroscopy (Goold et al., 23 Jan 2026) reveals that at $\log\lambda_{\rm Edd}\lesssim-3.5$ the SED of LLAGN becomes increasingly UV-deficient. Concomitantly, high-ionization emission lines ([Ne V], [O IV]) systematically weaken, and IR line ratios demand a power-law photoionization SED lacking the UV bump. Warm molecular H$_2$ emission and elevated excitation temperatures are ubiquitous, pointing to mechanical heating via shocks and X-ray dominated regions.

4. Observational Correlations, Fundamental Planes, and Feedback

Multiwavelength Correlations

• Fundamental Planes: LLAGN jointly trace the “fundamental plane of black hole activity”—a relation among radio luminosity, X-ray luminosity, and black hole mass—alongside hard-state X-ray binaries, supporting scale-invariant jet-accretion coupling (Saikia et al., 2018).
• Radio Properties: Parsec-scale, unresolved radio cores are nearly universal, and identified jets are aligned over parsec-to-kpc scales in objects with $Q_{\rm jet}>10^{42}$ erg s$^{-1}$ (Mezcua et al., 2014).
• Radio Luminosity Function (RLF): The nuclear RLF turns over at a critical accretion rate ($\dot m_{\rm Edd}\sim 5.1\times10^{-5}$), possibly marking the transition to “jet suppressed” or “inefficient” states where jets cannot be sustained (Saikia et al., 2018).

Feedback Modes

Kinetic energy output in jets and (in some cases) winds dominates the feedback from LLAGN, delivering maintenance-mode heating of the host galaxy's halo. The “kinetic dominance” ($Q_{\rm jet}\gtrsim L_{\rm bol}$) at $\lambda_{\rm Edd}\lesssim10^{-4}$ is ubiquitous and has important implications for galaxy evolution at low redshift (Mezcua et al., 2014, Fernández-Ontiveros et al., 2022).

5. High-Energy Emission, Winds, and Multi-Messenger Constraints

X-rays and Gamma Rays

LLAGN X-ray properties (hard power laws, weak/broad iron lines, undetectable Compton humps) are consistently reproduced by RIAF/ADAF + jet models, with inner disks truncated at $R_{\rm tr}\gtrsim 100\ r_g$ and no evidence for classical reflection signatures (Younes et al., 2018). X-ray polarization studies, such as with IXPE, further support slab-like or wedge-like coronal geometries and exclude extreme “pancake” coronae (Chakraborty et al., 3 Mar 2025).

Fermi-LAT studies reveal that only radio-bright, jet-dominated LLAGN are significant gamma-ray emitters; gamma-loudness strictly correlates with compact jet power (Menezes et al., 2020). X-ray-bright, radio-quiet LLAGN are not detected in gamma rays, consistent with their emission arising mainly from the RIAF rather than SSC processes (Menezes et al., 2020, Fernández-Ontiveros et al., 2022).

Winds and Outflows

LLAGN winds are consistent with predictions from hot-flow theory: the inferred wind velocity at the truncation radius tracks a fixed fraction of the local $v_{\rm K}$, and the mass flux in the wind grows linearly with radius, leading to $\dot M_{\rm wind}/\dot M_{\rm BH}\gg 1$. For M81*, this scaling is quantitatively confirmed, with profiled exponents matching GRMHD expectations (Shi et al., 12 Feb 2025). The feedback energy from the wind is subdominant to the jet (often $P_{\rm jet}/P_{\rm wind}\sim4$–25), but still critical for the circumnuclear medium.

UHECRs and Neutrinos

LLAGN jets with $P_j\lesssim10^{46}$ erg s$^{-1}$ are efficient accelerators of UHECRs up to $\gtrsim10^{20}$ eV via shocks at $\gtrsim10^3\,r_g$, primarily in radio-bright objects. Monte Carlo modeling links the UHECR flux distribution to observed radio core flux and BH mass, reproducing the Pierre Auger energy spectrum and arrival direction distribution (Dutan et al., 2014). Stochastic acceleration of nonthermal protons in RIAFs can account for a diffuse PeV neutrino background consistent with IceCube, provided only a small fraction ($\sim1\%$) of the accretion power is channeled into nonthermal ions (Kimura et al., 2014).

6. Accretion–Emission Scaling Relations and Evolutionary Connections

$\alpha_{\rm ox}$ and Bolometric Corrections

In deviation from luminous AGN trends, LLAGN display a flat or anti-correlated $\alpha_{\rm ox}$ vs. $L_{\rm bol}/L_{\rm Edd}$ for $\lambda_{\rm Edd}\lesssim10^{-3}$, a physiognomy captured by ADAF models (Xu, 2011, López et al., 2024). CIGALE SED fitting confirms a nearly constant, lower-than-classical $k_X$ ($k_X\sim9$–10) for LLAGN and robust recovery of $L_{\rm bol}$ even in galaxy-dominated photometry (López et al., 2024).

Evolutionary State

The torus and broad-line region recede or disappear below $\lambda_{\rm Edd} \lesssim 10^{-3.5}$. Thin disk emission weakens, mechanical and RIAF-jet feedback dominate, and the nuclear SEDs shift from disk- to jet-dominated regimes (Mason et al., 2012, Fernández-Ontiveros et al., 2022, Goold et al., 23 Jan 2026). This is marked in emission-line diagnostics by a dramatic weakening of high-ionization lines and a shift towards UV-deficient, X-ray- and shock-heated ISM (Goold et al., 23 Jan 2026).

7. Future Prospects and Methodological Advances

Recent advances leverage JWST IFU spectroscopy for nuclear emission-line diagnostics (Goold et al., 23 Jan 2026), high-cadence X-ray polarimetry for testing coronal geometries (Chakraborty et al., 3 Mar 2025), deep-learning surrogate models for fast SED fitting (Almeida et al., 2021), and global multiwavelength census with physically motivated SED templates (López et al., 2024). Open questions include the parameterization and universality of transition points ($\lambda_{\rm Edd}\sim10^{-3.5}$), the microphysical origin of jet/kinectic dominance, and the full multi-messenger contribution of LLAGN to cosmic backgrounds and feedback.

Ongoing and future work (e.g., ensemble IXPE/eXTP, JWST, ALMA, Fermi, IceCube) will further map the demography, energetics, and accretion physics of LLAGN, refining models for both the physics of quiescent SMBHs and their key role in cosmological galaxy evolution (Chakraborty et al., 3 Mar 2025, Goold et al., 23 Jan 2026, Almeida et al., 2021).