Extended Emission-Line Regions (EELRs)

• Extended Emission-Line Regions (EELRs) are large nebulae of ionized gas, often extending over 10 kpc, that reveal AGN radiation and diverse excitation mechanisms.
• They exhibit a variety of morphologies, such as biconical structures, filaments, tidal tails, and shells, influenced by galaxy interactions and environmental effects.
• Observational techniques like narrow-band imaging and integral-field spectroscopy enable detailed analysis of their ionization physics, kinematics, and feedback energetics.

Extended Emission-Line Regions (EELRs) are spatially resolved nebulae of ionized gas extending beyond the classical narrow-line region of a galaxy and emitting strong optical forbidden and recombination lines, most commonly [O III] $\lambda5007$, H$\alpha$, H$\beta$, [N II], [S II], and, in high-excitation cases, He II or [Ne V]. In AGN-related work, a practical definition often requires projected extension $>10$ kpc together with line ratios in the AGN regime of BPT diagrams or the presence of high-ionization lines such as He II $\lambda4686$. EELRs are used to trace anisotropic ionizing radiation, the interaction of jets and winds with the ambient medium, and “light echoes” of nuclear activity on $\sim10^4$–$10^5$ yr timescales (Keel et al., 2024, Sun et al., 2018, Husemann et al., 2012).

1. Definition, scope, and relation to ENLRs

The term EELR sits within a broader vocabulary that also includes the narrow-line region (NLR) and the extended narrow-line region (ENLR). In obscured and unobscured AGN, the NLR is typically described as ionized gas on scales of $\sim0.1$–$10$ kpc, while in the most powerful systems the same phenomenon extends to tens of kiloparsecs and is then termed an EELR (Sun et al., 2018). In quasar work, the ENLR is often defined more specifically as the subset of the EELR whose dominant ionization mechanism is photoionization by the QSO UV continuum (Husemann et al., 2012). In nearby Seyferts, ENLR and EELR are sometimes used nearly interchangeably for galaxy-scale AGN-ionized gas, including hollow or filled bicones of a few kpc and, in exceptional cases such as Mrk 783, structures extending to $\sim35$ kpc (Congiu et al., 2017).

The acronym is not restricted to a single physical mechanism. At $\alpha$0, spatially extended low-ionization emission-line regions with LINER-like line ratios have been identified on $\alpha$1–$\alpha$2 kpc scales and argued to arise from post-AGB stars and/or slow shocks rather than AGN (Hviding et al., 2018). In the Coma cluster, deep narrow-band imaging revealed extended H$\alpha$3 clouds associated with ram-pressure stripping, again labeled as extended emission-line regions despite a different excitation and dynamical origin (Yagi et al., 2010). This suggests that “EELR” is best treated as an observational category whose physical interpretation depends on line diagnostics, kinematics, and environmental context rather than on extent alone.

2. Morphology, size scales, and host environments

EELRs display a wide morphological range. Reported structures include hollow or filled ionization bicones, linear filaments, tidal tails, shell- or bubble-like rims around radio lobes, disk-like or streamer-like systems, one-sided extensions, and bar-linked rings. In the host of AT2019qiz, VLT/MUSE data reveal a bi-conical structure $\alpha$4 kpc across (Xiong et al., 25 Mar 2025). In radio-loud samples, a median equivalent radius of $\alpha$5 kpc has been reported, with many systems filling wide cones with half-opening angles of $\alpha$6–$\alpha$7 (Shih et al., 2014). In the MURALES survey of low-redshift 3C radio galaxies, FR II EELRs are almost ubiquitous and typically extend between $\alpha$8 and $\alpha$9 kpc, with maxima of $\beta$0 kpc, whereas FR I systems usually confine line emission within $\beta$1 kpc except for rare cases such as 3C 348 (Balmaverde et al., 2022). In type 2 quasar work, the ionized structure can be even larger: J0123+00 shows an ionized bridge reaching $\beta$2 kpc (Villar-Martin et al., 2011).

Environment strongly conditions detectability. Distant, kinematically quiescent EELRs are closely linked to tidal debris in interacting and merging systems (Keel et al., 2021). The TELPERION merger sample, built from 92 strongly interacting or merging systems containing 198 galaxies, found four distant EELRs in that subsample and concluded that interacting and merging AGN have a much higher detection rate than noninteracting AGN (Keel et al., 2024). In radio galaxies, the large-scale emitting gas is often elongated at large angles to the radio axis and has been interpreted as part of merger-formed gaseous “superdisks” whose appearance depends on source age and jet–disk orientation (Balmaverde et al., 2022). Nearby Seyferts add further geometric diversity: IC 5063 is described as a hollow bicone with one edge nearly in the host disk, NGC 7212 as a more filled bicone, and HE 2211-3903 as a nearly face-on ring at $\beta$3 kpc linked to the nucleus by a bar-like structure (Congiu et al., 2017, Scharwaechter et al., 2011).

Large statistical imaging samples show that EELR size scales systematically with AGN power. Using reconstructed [O III] maps for 300 obscured AGN in Subaru/HSC data, the isophotal area and radius of the emission-line region were found to correlate strongly with rest-frame $\beta$4m luminosity, with an area–luminosity slope of $\beta$5 and an implied radius slope of $\beta$6; the resolved systems extend up to $\beta$7 kpc and frequently show filamentary or biconical morphologies (Sun et al., 2018).

3. Ionization physics and diagnostic frameworks

The standard classification of EELRs relies on emission-line diagnostics. In AGN-dominated systems, strong-line ratios such as $\beta$8, $\beta$9, $>10$0, and $>10$1 are placed on BPT diagrams, with regions above the Kewley or Kauffmann demarcations interpreted as AGN-photoionized (Keel et al., 2024, Husemann et al., 2012, Congiu et al., 2017, Balmaverde et al., 2022). High-ionization lines provide additional leverage. In the new UGC 5941 cloud found by TELPERION, $>10$2, $>10$3, $>10$4, and $>10$5 still place the cloud in the AGN regime (Keel et al., 2024).

A central quantity is the ionization parameter,

$>10$6

where $>10$7 is the ionizing photon rate, $>10$8 the distance to the cloud, and $>10$9 the gas density (Reynaldi et al., 2015). In pure AGN photoionization models, high-ionization lines strengthen with increasing $\lambda4686$0 and should weaken with radius as geometric dilution lowers $\lambda4686$1 (Reynaldi et al., 2015, Congiu et al., 2017). However, many EELRs require more than a single-source photoionization picture. In CSS radio galaxies 3C 268.3 and 3C 303.1, [Ne V] $\lambda4686$2 is more extended than expected relative to He II and increases with distance even where $\lambda4686$3 declines, a result argued to favor fast shocks plus a precursor field rather than pure AGN photoionization (Reynaldi et al., 2015). In Seyfert ENLRs, high-velocity gas is commonly modeled with composite photoionization-plus-shock prescriptions; in IC 5063 and NGC 7212, the highest-velocity bins enter the shock-dominated part of a diagnostic $\lambda4686$4 diagram, while low-velocity material remains AGN-photoionized (Congiu et al., 2017).

Jet–cloud interaction is a recurring theme. CSS sources show velocity gradients aligned with the radio axis and line-ratio behavior consistent with a combination of AGN radiation and shock excitation (Shih et al., 2013). NGC 3393 provides a high-resolution circumnuclear example in which [S II]/H$\lambda4686$5 peaks along radio-lobe interfaces, soft X-rays outline narrow filaments, and the ENLR is described as a highly stratified multi-phase structure with shock contributions on scales of $\lambda4686$6 pc (Maksym et al., 2016). By contrast, some galaxy-scale ENLRs appear to be overwhelmingly photoionized. In HE 2211-3903, composite line ratios are explained as mixing between AGN and H II-region photoionization, and fast shocks are argued to be negligible at the observed spectral resolution (Scharwaechter et al., 2011).

Two recurrent complications are non-AGN excitation and contamination by star formation. In Mrk 783, line ratios indicate star-formation contamination at intermediate radii but nearly pure AGN ionization beyond $\lambda4686$7 kpc (Congiu et al., 2017). At $\lambda4686$8, extended LIER-like emission has been interpreted as stellar photoionization and/or slow shocks rather than a nuclear AGN (Hviding et al., 2018). A related practical issue is that EELRs can bias star-formation estimates from [O II]: detection of [Ne V] $\lambda4686$9 at $\sim10^4$0 has been proposed as a flag that [O II] is significantly contaminated by AGN-ionized extended gas and is therefore not a reliable SFR tracer (Maddox, 2018).

4. Observational strategies and analysis methods

EELRs have been mapped by narrow-band imaging, broadband continuum subtraction, long-slit spectroscopy, and integral-field spectroscopy. Narrow-band [O III] surveys remain especially effective for distant AGN-ionized clouds. TELPERION used an F510 filter to image [O III] $\sim10^4$1 in low-redshift AGN and interacting systems, reaching a typical $\sim10^4$2 surface-brightness sensitivity of $\sim10^4$3 and requiring confirmation of candidate diffuse features before spectroscopy (Keel et al., 2024). The earlier TELPERION phase covered 111 AGN hosts and 17 mergers with fields large enough to probe projected radii of $\sim10^4$4–$\sim10^4$5 kpc, emphasizing the importance of depth and wide angular coverage for finding light-echo clouds in tidal debris (Keel et al., 2021).

A different approach reconstructs [O III] maps from deep broadband imaging. In Subaru/HSC data, the continuum-subtracted line image is written as

$\sim10^4$6

with the scaling factor $\sim10^4$7 calibrated from SDSS/BOSS spectra after stellar-continuum fitting and PSF matching between bands (Sun et al., 2018). Applied to 300 Type II AGN, this method produced the largest uniformly processed sample of narrow-line region sizes, with 231 of 300 objects detected and resolved (Sun et al., 2018). For bright quasars, the main technical obstacle is nuclear contamination; PMAS IFU data were therefore analyzed with the iterative QDeblend$\sim10^4$8 procedure, which subtracts the PSF-scaled unresolved QSO spectrum from each spaxel to recover pure host-plus-EELR datacubes (Husemann et al., 2012).

Integral-field spectroscopy has transformed the field because it couples morphology, line ratios, and kinematics. MUSE, SNIFS, WiFeS, GMOS, MPFS, and related instruments have been used to generate narrow-band line maps, spaxel-wise diagnostic diagrams, velocity fields, and multi-component Gaussian decompositions of the emission lines (Shih et al., 2014, Shih et al., 2013, Xiong et al., 25 Mar 2025, Kozlova et al., 2019). In practice, EELR studies combine spatially resolved line fitting, continuum subtraction, and targeted extraction of off-nuclear apertures to separate nuclear NLR light from genuinely extended gas.

5. Kinematics, mass estimates, and feedback energetics

Kinematic behavior varies strongly by host class. In powerful radio-loud AGN, EELRs are often interpreted as outflows. A biconical description with wide opening angles explains much of the three-dimensional morphology in FR II quasars and radio galaxies, for which the average extent is $\sim10^4$9 kpc, the average ionized-gas mass is $10^5$0, and a simple outflow estimate gives an average mass-loss rate of $10^5$1 (Shih et al., 2014). Using

$10^5$2

and characteristic velocities of a few hundred km s$10^5$3, the corresponding kinetic energy and power are of order $10^5$4–$10^5$5 erg and $10^5$6–$10^5$7 (Shih et al., 2014). Yet these same radio-loud samples also show complexity: roughly $10^5$8 of luminous EELRs exhibit large-scale rotation, and there is no systematic correlation between the radio-jet position angle and the major axis of rotation or outflow, which has been taken to favor broad blast-wave driving rather than direct jet entrainment (Shih et al., 2014).

MURALES reinforces the coexistence of ordered rotation and misalignment. In its nuclear few-kpc regions, 20 of 26 FR IIs and 5 of 10 FR Is show ordered rotation, typically with amplitudes of $10^5$9–$\sim0.1$0, while the median angle between the kinematic axis and the radio axis is $\sim0.1$1 (Balmaverde et al., 2022). Some FR IIs continue that rotational pattern to tens of kiloparsecs, whereas others retain only velocity symmetry with strong local disturbances (Balmaverde et al., 2022). CSS sources, by contrast, often show multi-component [O III] profiles with narrow and broad components, velocity gradients aligned with the radio axis, and outflow signatures consistent with the onset of radio-jet activity (Shih et al., 2013).

Not all EELRs are dynamically violent. In 31 low-redshift type 1 QSOs, EELRs were detected around 19 objects, but only 3 systems show radial velocities exceeding $\sim0.1$2, and those regions are associated with radio jets; the majority have quiescent or gravitationally driven kinematics (Husemann et al., 2012). Several post-starburst or accretion-event hosts show similarly calm gas. In Markarian 950, the ionized gas is nearly uniformly blueshifted by $\sim0.1$3 to $\sim0.1$4 with $\sim0.1$5 outside the nucleus and is kinematically decoupled from the stars, arguing against AGN-driven outflow and favoring tidal debris illuminated by a decayed ionizing source (Wevers et al., 2024). In Mrk 78, distant clouds at $\sim0.1$6–$\sim0.1$7 kpc have $\sim0.1$8–$\sim0.1$9 and only modest velocity offsets, again inconsistent with fast AGN-driven outflow (Kozlova et al., 2019). These systems illustrate that an EELR can record past nuclear luminosity even when the present-day gas motions are cold.

6. Incidence, AGN fading, and links to transient accretion

Incidence estimates depend strongly on selection and surface-brightness depth. Combining all TELPERION phases, the distant-EELR rate is $10$0 per galaxy and $10$1 per Seyfert, while the merger subsample yields $10$2 per galaxy and $10$3 per Seyfert (Keel et al., 2024). Restricting to interacting or merging Seyferts across TELPERION gives $10$4, whereas none are detected around noninteracting AGN in that analysis (Keel et al., 2024). Earlier TELPERION results similarly found that $10$5 of tidally distorted AGN hosts show distant EELRs (Keel et al., 2021). Other host classes differ sharply: luminous EELRs occur in roughly $10$6–$10$7 of matched FR II quasars and radio galaxies but in none of the low-excitation radio galaxies of that sample (Shih et al., 2014), and around low-redshift type 1 QSOs the overall EELR detection rate is $10$8 after QSO–host deblending (Husemann et al., 2012).

One of the major uses of EELRs is reconstructing nuclear luminosity histories. A commonly used energy-balance test compares the current nuclear output,

$10$9

to the minimum ionizing luminosity required by recombination balance in the cloud at projected radius $\sim35$0; if $\sim35$1, the AGN must have faded over the light-travel delay $\sim35$2 (Keel et al., 2024). In NGC 5514, this implies fading by $\sim35$3 over $\sim35$4 yr; in NGC 7252, the difference exceeds $\sim35$5 over $\sim35$6 yr; in UGC 5941, the required change is more modest, with $\sim35$7 over $\sim35$8 yr (Keel et al., 2024). These are the clearest empirical cases in which EELRs operate as delayed records of a brighter AGN state.

Post-starburst TDE hosts extend this logic into time-domain nuclear astrophysics. In the AT2019qiz host, the ionizing luminosity required for all AGN-classified spaxels is $\sim35$9, whereas the current quiescent nuclear luminosity is only $\alpha$00, implying either a recently faded AGN or an ionization echo of historical activity (Xiong et al., 25 Mar 2025). In Markarian 950, the EELR requires $\alpha$01, inconsistent with the present weak nucleus; the inferred fading factor is at least $\alpha$02–$\alpha$03 over $\alpha$04–$\alpha$05 yr (Wevers et al., 2024). Among five MUSE-observed post-starburst TDE hosts, three show EELRs, corresponding to $\alpha$06, versus $\alpha$07 in a control sample of post-starburst galaxies; the reported enhancement is therefore about a factor of $\alpha$08 (Wevers et al., 2024).

A proposed alternative to the fading-AGN interpretation is that repeated tidal disruption events themselves can sustain galaxy-scale EELRs. Coupled TDE-disk plus CLOUDY simulations argue that TDE disks can provide ionizing luminosities of $\alpha$09–$\alpha$10, power EELRs out to $\alpha$11 light years in low-density gas with $\alpha$12–$\alpha$13, and generate AGN-like line ratios for $\alpha$14 yr after each event (Mummery et al., 18 Mar 2025). This does not displace the fading-AGN picture in established AGN hosts, but it formalizes an active controversy in post-starburst systems where EELRs, weak present-day nuclei, and transient accretion phenomena coexist.

In current usage, EELRs are therefore both emission-line nebulae and temporal diagnostics. Their morphology traces gas supply and geometry; their spectra distinguish AGN photoionization, shocks, stellar ionization, and environmental stripping; and their energetics record nuclear states that may no longer be directly observable.