EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 697 (May 2025) A&A, 697 (2025) A235 Full HTML
Open Access
Issue
A&A
Volume 697, May 2025
Article Number A235
Number of page(s) 20
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/202453481
Published online 22 May 2025

© The Authors 2025

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

Although active galactic nuclei (AGNs) are mostly found in massive galaxies, in the late 1980s Filippenko & Sargent (1989) discovered an AGN in the dwarf (stellar mass M* ≤ 1010 M⊙) spiral galaxy NGC 4395. Nowadays, the use of optical spectroscopy using surveys such as the Sloan Digital Sky Survey (SDSS) or the Dark Energy Spectroscopic Instrument (DESI) has allowed us to find thousands of dwarf galaxies hosting AGNs (see Pucha et al. 2025). This has been mostly based on the use of emission line diagnostic diagrams such as the Baldwin, Phillips, and Terlevich (BPT; Baldwin et al. 1981), which enable us to distinguish between gas ionisation produced by AGNs, stars, or both. This distinction can be made by comparing the ratio of the narrow emission lines [OIII]λ5007/Hβ against [NII]λ6583/Hα, [SII]λλ6716, 6731/Hα, or [OI]λ6300/Hα (Kauffmann et al. 2003; Kewley et al. 2001, 2006). AGNs in massive galaxies often exhibit a broad Hα or Hβ component originating from dense gas clouds in the broad-line region (BLR) located a few parsecs from the supermassive black hole (SMBH; MBH > 106 M⊙). These clouds are ionised by continuum emission from the accretion disk of the SMBH. In massive galaxies, the typical full width at half maximum (FWHM) of the broad component ranges from ∼1000 to ∼5000 km s−1 (Osterbrock & Ferland 2006; Kollatschny & Zetzl 2013). In dwarf galaxies, such broad components have also been identified, but with a less pronounced width of ∼500−1000 km s−1 (e.g. Reines et al. 2013; Chilingarian et al. 2018). Assuming that the gas producing the broad emission line is moving in circular orbits in the BLR around the SMBH, a BH mass measurement can be obtained from the detection of broad Balmer lines. In the case of dwarf galaxies, this BH mass is typically MBH = 105 − 106 M⊙; therefore, it is lower than that of SMBHs (e.g. Greene & Ho 2007; Reines et al. 2013; Moran et al. 2014; Baldassare et al. 2015; Marleau et al. 2017; Chilingarian et al. 2018; Mezcua & Domínguez Sánchez 2020, 2024; Salehirad et al. 2022).

BPT diagnostic diagrams used in surveys that only obtain the spectrum at the centre of the galaxy can miss AGN in galaxies with a high star formation rate (SFR) that dilutes the AGN signatures or in cases where the AGN is displaced from the centre of the galaxy (Comerford & Greene 2014). This displacement is expected to occur in ∼50% of dwarf galaxies (e.g. Bellovary et al. 2019). This problem can be solved by observing galaxies in the X-ray and radio bands (e.g. Hasinger 2008; Mezcua et al. 2016, 2018, 2019; Hickox & Alexander 2018; Reines et al. 2020; Birchall et al. 2020; Bykov et al. 2024). Alternatively, integral field unit (IFU) spectographs allow us to identify these ‘hidden AGN’ because they provide information spaxel by spaxel (i.e we have a spectrum for each pixel of the galaxy) thanks to the different fibers arranged of each IFU (e.g. Ricci et al. 2014; da Silva et al. 2017; Wylezalek et al. 2020; Liu et al. 2020; Mezcua & Domínguez Sánchez 2020, 2024). Furthermore, by carrying out spatially-resolved emission line diagnostic diagrams (i.e. BPT spaxel by spaxel), we are able to identify which regions of the galaxy are ionised by AGNs or star formation (SF) processes (e.g. Mezcua & Domínguez Sánchez 2020, 2024; Wylezalek et al. 2020).

AGNs release energy in the form of radiation or of mechanical radio plasma coming from the gravitational energy of the material accreted by the BH. There are different launching mechanisms, such as collimated jets of charged particles or winds formed from the radiation pressure or from the accretion disk. These winds rise up to outflows when they interact with and swept up material from the interstellar medium (Harrison & Ramos Almeida 2024). Outflows can also be produced by SF processes (e.g. Shepherd 2007; Gatto et al. 2017; Romano et al. 2023; Sau et al. 2023) and can be identified by looking for broadened or shifted components in emission lines like [OIII]λ5007, [NII]λ6583 or [SII]λλ6716, 6731 (e.g. Holt et al. 2008; Liu et al. 2013; Harrison et al. 2014; Leung et al. 2019; Wylezalek et al. 2020). By combining the kinematic information of the emission lines with the BPT diagram diagnostics to access the gas excitation, we have access to a powerful way to infer the origin of the outflows and determine whether these are AGN or SF processes.

The impact of AGN outflows on the host galaxy, known as AGN feedback, plays a role in regulating the star formation due to the injection of energy and the momentum transfer in the interstellar medium of the galaxy. As a result, the growth of SMBHs is interconnected with the evolution of their host galaxies, a phenomenon referred to as BH-galaxy co-evolution (e.g. Kormendy & Ho 2013; Zhuang & Ho 2023; Capelo et al. 2024). AGN feedback and outflows are typically found in massive galaxies and they are generally suggested to prevent gas from cooling and to redistribute it (e.g. Croton et al. 2006; Bower et al. 2006; Falceta-Gonçalves et al. 2010). Nevertheless, high spatial resolution observations and recent simulations have made the understanding of AGN feedback more complex, pointing out the importance of initial conditions in different factors, such as the gas distribution or the coupling between the outflow and the interstellar medium (e.g. Tanner & Weaver 2022; Clavijo-Bohórquez et al. 2024, see Harrison & Ramos Almeida 2024 for a recent review).

In the low-mass regime, the presence of AGN outflows is a matter of debate. AGNs are typically fainter and less powerful in dwarf galaxies than in massive galaxies (but see Mezcua et al. 2023, 2024). Furthermore, star formation in dwarf galaxies is commonly assumed to be regulated by stellar winds coming from supernovae (SNe). However, recent studies show evidence for the presence of AGN outflows and feedback in dwarf galaxies. Penny et al. (2018) found evidence of AGN feedback in a subset of 69 quenched low-mass galaxies selected from the first two years of the SDSS Mapping Nearby Galaxies at Apache Point Observatory (MaNGA; Bundy et al. 2015) survey. Manzano-King et al. (2019) reported the detection of six dwarf galaxies selected from SDSS that have an outflow component in the [OIII] doublet with emission lines ratios consistent with AGN ionisation. Liu et al. (2020) used a sample of eight dwarf galaxies with known AGN taken from Manzano-King et al. (2019), and the IFU data reveals signs of outflowing gas in seven of them based on the detection of a wing in the Gaussian fit of the [OIII] line. Liu et al. (2024) studied three far-ultraviolet dwarf galaxies. They reported a fast outflow detected in multiple transitions in one source. In another source, they detected a blue-shifted [HeII]λ1640 Å emission line, likely tracing a highly ionised AGN outflow. Zheng et al. (2023) detected an escaping outflow in a dwarf galaxy with an intermediate-mass black hole (IMBH, 100 M⊙ ≤ MBH ≤ 106 M⊙) and studied the importance of the feedback by comparing the size of the outflow with the extended narrow-line region. Wang et al. (2024) reported a low-mass galaxy at redshift z = 3.1 with AGN signatures and with an absorption feature in the wings of [HeI]λ10839 Å, indicating an ionised gas outflow. Salehirad et al. (2025) reported 11 AGN dwarf galaxies with fast outflows consistent with an AGN-driven origin.

On the theoretical side, cosmological simulations considering AGN feedback in high-redshift dwarf galaxies indicate that BHs can provide a significantly amount of feedback that can quench SF (Barai & Dal Pino 2018). In contrast, Trebitsch et al. (2018) found that even in the most extreme BH growth scenario, SNe feedback quenches the AGN feedback, with this latter type shown to be negligible in comparison. In recent simulations, Koudmani et al. (2022) found a midpoint where a more moderate SNe feedback in combination with an efficient AGN could be an alternative. In this simulation, a variety of outcomes can be obtained depending on the accretion model: from no additional suppression to moderate regulation of SF to catastrophic quenching. Arjona-Galvez et al. (2024) investigated magneto-hydrodynamical simulations of the formation of dwarf galaxies: one with AGN feedback and the other with AGN feedback turned off. The AGN runs reproduce satisfactorily several scaling relations and they find that the global SF of galaxies with AGN is reduced compared to those galaxies in which the AGN has been turned off. At this point, different cosmological simulations bring on different results; thus, observational constraints are needed. Local dwarf galaxies have presumably not experienced any significant evolution and thus resemble those first galaxies formed in the early Universe (z ∼ 12). Therefore, understanding feedback processes and their contribution in dwarf galaxies could be an essential key for cosmological models of galaxy evolution.

In this paper, we present a new sample of 13 dwarf galaxies with AGN outflow candidates drawn from the final MaNGA data release (DR17)1. We compare their properties with those of massive galaxies with AGN-driven outflows. The paper is organised as follows. In Sect. 2, we study a sample of 2292 dwarf galaxies from Mezcua & Domínguez Sánchez (2024) that present AGN ionisation signatures in the spatially-resolved BPTs. We analyse the [OIII]λ5007 Å line in the stacked spectrum, followed by the individual spaxels, to detect and spatially resolve the outflows. In Sect. 3, we derive the extension of the outflows, study their kinematic and energetic properties and discuss their AGN origin. Final conclusions are drawn in Sect. 4. Throughout the paper a standard ΛCDM cosmology is adopted with H0 = 70 km s−1 Mpc−1, ΩM = 0.3, and ΩΛ = 0.7.

2. Sample and analysis

2.1. Sample of dwarf galaxies in MaNGA

The sample studied in this paper is drawn from Mezcua & Domínguez Sánchez (2024). They made a selection of 3306 dwarf galaxies (M* ≤ 1010 M⊙) from the MaNGA DR17, which is the final MaNGA data release that includes IFU observations and data products of more than 10 000 nearby galaxies. Each IFU is composed of tightly packed arrays of 19 to 127 optical fibres, with diameters from 12 to 32 arcsec. It covers a wavelength range of 3600−10 000 Å with a spectral resolution of R ∼ 2000.

Mezcua & Domínguez Sánchez (2024) made a pioneering classification of the spaxels of the 3306 dwarf galaxies based on the [NII]–BPT, [SII]–BPT and [OI]–BPT diagrams using the Kauffmann et al. (2003) classification cutoff, the Kewley et al. (2001, 2006) maximum starburst lines. They combine this classification with the WHAN diagram, used to distinguish between emission by AGN or by hot old stars (Cid Fernandes et al. 2010). This yields 2292 dwarf galaxies with strong AGN ionisation signatures (i.e. more than 20 spaxels with a signal-to-noise ratio (S/N ≥ 3) classified as AGN, SF-AGN (i.e. SF in the [NII]–BPT but AGN in the [SII]–BPT or [OI]–BPT), composite, or LINER2. For a detailed description, see Mezcua & Domínguez Sánchez 2024).

For all these 2292 dwarf galaxies, Mezcua & Domínguez Sánchez (2024) stacked the spectrum of those AGN and composites spaxels. The stacking procedure is described in Appendix A in Mezcua & Domínguez Sánchez (2020). In this paper, the stacked spectrum of these galaxies is fitted using the Galaxy/AGN Emission Line Analysis TOol (GELATO; Hviding et al. 2022), available on Zenodo3. GELATO is a Python code designed to retrieve the kinematics and line fluxes of optical spectra. It fits multiple components to lines with complex kinematics and allows the user to tie parameters between emission lines (e.g. Hα and Hβ velocity dispersion or the flux ratio between [OIII]λλ5007, 4959 Å emission lines which are known due to the transition probabilities). To run GELATO, the spectral flux, the wavelength, the inverse-variance (ivar), and the redshift (taken from the NASA-Sloan Atlas (NSA) catalogue4) must be provided in addition to the emission lines to be fitted, which are: [SII]λλ6716, 6731 Å, [NII]λλ6583, 6548 Å, Hαλ6563 Å, [OI]λλ6300, 6364 Å, HeIλ5876 Å, [OIII]λλλ5007, 4959, 4364 Å, Hβλ4861 Å, [Hγ]λ4340 Å, [NeIII]λ3869 Å, and [OII]λλ3726, 3728 Å. GELATO models the stellar component using the Extended MILES stellar library (Vazdekis et al. 2016) combining simple stellar populations (SSP). An additional power-law component is added if the continuum model passes a statistical F-Test at a 3σ level, corresponding to a likelihood of ∼99.87% (Hviding et al. 2022). Additional components are also tested with an F-Test at a 3σ level and the final model is the one with the lowest Akaike information criteria (AIC; Akaike 1974), which is an estimator of the relative quality of statistical models. We allow GELATO to fit a broad component to the Balmer lines and an outflow component to the [OIII] emission line, apart from the narrow component (all of them fitted as Gaussians). The distinction between broad and outflow component is based on the bounds of the dispersion and redshifts of these components. For the outflow component, the dispersion has a lower value of 100 km s−1 and an upper value of 750 km s−1 corresponding to the 1st and 99th percentile values of the dispersion of the second [OIII] component (Mullaney et al. 2013, for further details see Hviding et al. 2022). In this paper we use the terms ‘broad component’ and ‘outflow component’ interchangeably.

Once the stacked spectrum of each of the 2292 galaxies was fitted, the [OIII]λ5007 Å emission line was examined to determine whether a broadened or shifted component can be observed (see Fig. 1), indicating gas with different kinematics than the rest of the ionised gas in the galaxy and thus potentially tracing the outflowing gas (e.g. Liu et al. 2013; Manzano-King et al. 2019). Furthermore, a visual inspection was done to remove those galaxies in which GELATO associated a broad component due to the noise near the [OIII]λ5007 Å line. Since the final model is the one with the lowest AIC, a more complex model might be selected unnecessarily. To address this, we checked in how many stacked spectra the [OIII] lines present a better fit with a double Gaussian, rather than a single Gaussian when comparing the χ2 value of both fits and performing a second visual inspection. If the double Gaussian fit has χ2 ≤ 20 and it improves the χ2 of the single fitting by at least 20%, the galaxy is included in our sample of outflow candidates (e.g. Hao et al. 2005; Reines et al. 2013). At this point, the number of AGN dwarf galaxies that fulfil our criteria to show outflows signatures is 64.

thumbnail Fig. 1.

Top left and right: Spectra of the galaxies 8982−3703 and 12487−3704, respectively. The dark line is the stacked spectrum of the AGN spaxels. The green line is the best fit of the stellar continuum combining SSPs. The blue line is an additional power-law component, the yellow line represents the fitted emission lines, and the red line is the sum of all these components that compose the total model. The residuals after subtraction of the total model from the data are shown at the bottom. Bottom-left and right: Zoom-in on the [OIII]λ5007 Å doublet region of the stacked spectrum. The [OIII] line is fitted by a narrow (blue) and a broad (green) component.

To further restrict our sample of dwarf galaxies with possible AGN outflows, we performed a similar analysis as the one explained above, but this time for the individual AGN and composite spaxels of the 64 candidates selected. We took advantage of the MaNGA Data Reduction Pipeline (DRP)5 of DR17 data-products, which provide flux-calibrated, sky-subtracted, co-added data cubes from each of the individual exposures for a given galaxy. The wavelength, flux, and ivar of each spaxel are provided by the LOGCUBE data cube. Only spaxels with a S/N ≥ 10 were selected in order to avoid unreliable spectrum fittings (Appendix A offers a discussion of the S/N threshold selection can be found). After running GELATO for each of the individual spaxels of the 64 dwarf galaxies with AGN outflow candidates, we selected only those galaxies that present more than ten consecutive AGN spaxels with a broad component in the [OIII] line (nout > 10) in at least one region of the host galaxy. By applying this cut, a final sample of 13 dwarf galaxies with AGN outflow candidates was obtained, of which seven were classified as AGN galaxies and six as SF-AGN, according to Mezcua & Domínguez Sánchez (2024). Three of the AGN dwarf galaxies (8442−1901, 8982−3703 and 9889−1902) show additional evidence to support the presence of AGN based on the detection of radio emission and one (10223−3702) based on its mid-infrared colours, according to the criteria of Jarrett et al. (2011) or Stern et al. (2012). The analysis of the individual spaxels is also performed for the SF spaxels of these 13 dwarf galaxies in order to discuss the origin of the outflow (see Sect. 3.4).

2.2. Star formation rate measurement

We fit the spectral energy distributions (SED) using Code Investigating GALaxy EMission (CIGALE; Boquien et al. 2019) to estimate the SFR of the 13 MaNGA dwarf galaxies with AGN outflow candidates; this is because the one obtained by taking the Hα emission line from the MaNGA data-analysis pipeline (DAP; Westfall et al. 2019; Belfiore et al. 2019) does not account for the AGN contribution (see Sect. 3.4.1). We considered the single stellar population models from Bruzual & Charlot (2003), assuming a Chabrier initial mass function as per Chabrier (2003). We used the standard nebular emission model from Inoue (2011) and the dust attenuation model using the Calzetti et al. (2000) starburst attenuation curve. The reprocessed dust emission was modelled using the Draine et al. (2013) dust models. Finally, we used the AGN emission model from Fritz et al. (2006). When available, we used the following photometry: SDSS magnitudes in the optical/infrared bands (u, g, r, i, z; Fukugita et al. 1996; Doi et al. 2010) adapted to a composite model following Maraston et al. (2013). Near-infrared photometry in the J (1.25 μm), H (1.65 μm), and K (2.17 μm) bands from The Two Micron All Sky Survey (2MASS; Cutri et al. 2003, 2012; Skrutskie et al. 2006). Infrared photometry from the Wide-field Infrared Survey Explorer (WISE; Wright et al. 2010; Cutri et al. 2021) passbands W1, W2, W3, and W4 with effective wavelengths of 3.4, 4.6, 12.1, and 22.5 μm. Finally, the far- and near-ultraviolet measurements are also included from the Galaxy Evolution Explorer (GALEX; Bianchi et al. 2011, 2017; Osborne et al. 2023).

2.3. Sample of massive galaxies in MaNGA

To compare our results with a sample of massive galaxies, we used the MaNGA sample of massive galaxies (M* > 1010 M⊙) with AGN outflows of Wylezalek et al. (2020). To ensure that the analysis is carried out in the same way for both dwarf and massive galaxies, we performed the same analysis as for the dwarf galaxies described in Sect. 2.1 for the 154 massive galaxies with AGN from Wylezalek et al. (2020). We found that 46 out of these 154 massive galaxies present an extended broad component region (ten consecutive AGN spaxels with a broad component in the [OIII] line). Wylezalek et al. (2020), considering only those AGN galaxies with W80 ≥ 500 km s−1 (where W80 is the velocity width containing 80% of the flux of the emission line), reported 29 AGN outflows in their sample of massive galaxies. This cut in the W80 velocity is the main reason behind the difference between the number of outflows we found in their sample and those reported in Wylezalek et al. (2020). From the 46 massive galaxies with AGN outflow signatures we report here, 17 have W80 ≥ 500 km s−1, with 11 out of these 17 shown to be consistent with the 29 reported by Wylezalek et al. (2020). For the remaining 18 galaxies with W80 ≥ 500 km s−1 reported by Wylezalek et al. (2020), we found (according to our measurements; see Sect. 3.1) that 6 of them have W80 ≤ 500 km s−1, while the rest do not meet our criteria regarding consecutive AGN spaxels with a broad component in the [OIII] emission line.

3. Results and discussion

In this work, we report the discovery of 13 MaNGA dwarf galaxies with AGN outflow candidates identified using the MaNGA survey. Furthermore, we investigate whether the gas kinematics differs from that of the stellar kinematics. Although gas and stars can rotate perpendicularly (Sarzi et al. 2006), the misalignment may suggest past interactions (Casanueva et al. 2022; Raimundo et al. 2023; Zinchenko 2023; Winiarska et al. 2025) or the presence of potential outflows (Ristea et al. 2022). To this end, the position angle (PA) of the [OIII] total gas velocity and of the stellar velocity can be derived using the PaFit package in Python (Krajnovic et al. 2006; see Table 1). For nine out of the 13 dwarf galaxies with outflows signatures, we found that the |PAgas − PAstars|≳20°. This supports the assumption of a deviation of the gas from the stellar kinematics (see Figs. 2 and 3). The remaining four galaxies, 7992–6102, 8442–1901, 8657–6104, and 9889–1902 (which is in the process of undergoing a merger), have a PA difference of ≤12°, which does not confirm the outflowing nature of the gas.

Table 1.

Position angles.

thumbnail Fig. 2.

MaNGA analysis for one of the dwarf galaxies with AGN outflow candidates (8655−6101). Top left: Spatial distribution of the BPT-classified spaxels (red: AGN; green: composite; blue: SF; orange: LINER). The ‘N’ shows the number of AGN and composite spaxels used in the analysis. Top right: SDSS composite image. The pink hexagon shows the IFU coverage. Middle: Spatial W80 (left) and vout (right) outflow velocity distribution, where we can see the extension of the outflow. Empty squares mark the IFU coverage and grey squares those spaxels with a continuum S/N > 1. Bottom: MaNGA [OIII]λ5007 Å velocity map (left) and stellar velocity (right).

thumbnail Fig. 3.

Same as Fig. 2 but for the dwarf galaxy 10223−3702.

In the following subsections, we present the outflow kinematic and energetic properties of the 13 MaNGA dwarf galaxies. We start with the calculation of the velocity of the outflows, followed by calculations of the radius, the mass and the energy, momentum, and mass rates.

3.1. Outflow velocity and radius

The outflow velocity is defined in Manzano-King et al. (2019) as

v out = − v 0 + W 80 2 , $$ \begin{aligned} v_{\rm out}=-v_0+\frac{\mathrm{W}_{80}}{2}, \end{aligned} $$(1)

where v0 is the velocity offset between the narrow and the broad component of the [OIII]λ5007 Å line. In the case of a single Gaussian fitting, W80 could be defined as W80 = 1.09 × FWHM. However, for non-Gaussian emission line profiles, the non-parametric velocity width measurements are more sensitive to the weak broad bases (Liu et al. 2013). As our fitting is composed by the sum of two Gaussians, W80 is calculated using the following expression:

W 80 ≡ v 90 − v 10 , $$ \begin{aligned} \mathrm{W}_{80}\equiv v_{90}-v_{10}, \end{aligned} $$(2)

where v90 and v10 are the velocities at the 10th and 90th percentiles of the total flux. In order to calculate these velocities, we need to calculate the cumulative flux as a function of the velocity:

ϕ ( v ) ≡ ∫ − ∞ v F v ( v ′ ) d v ′ . $$ \begin{aligned} \phi (v)\equiv \int _{-\infty }^v F_{\rm v}(v\prime )\mathrm{d}v\prime . \end{aligned} $$(3)

The total line flux is given by ϕ(∞). Fv is the spectral flux density. In the wavelength space, the spectral flux density can be parametrised as a Gaussian distribution:

F λ = F σ λ 2 π exp ( − 1 2 ( λ − λ r ( 1 + z ) σ λ ) 2 ) , $$ \begin{aligned} F_\lambda = \frac{F}{\sigma _\lambda \sqrt{2\pi }}\mathrm{exp} \left(-\frac{1}{2}\left(\frac{\lambda -\lambda _{\rm r}(1+z)}{\sigma _\lambda }\right)^2\right), \end{aligned} $$(4)

taking into account the expansion of the Universe with the factor (1 + z). Then, F is the integrated line flux, λ is the wavelength, λr is the wavelength of the emission line in the rest-frame, and σλ is the dispersion in units of wavelength. As the dispersion of the line is dominated by the kinematics of the gas, we convert σλ into velocity units by using the Doppler effect:

σ = c σ λ λ r ( 1 + z ) · $$ \begin{aligned} \sigma = c\frac{\sigma _\lambda }{\lambda _{\rm r}(1+z)}\cdot \end{aligned} $$(5)

We notice that the integrand in Eq. (3) is in the velocity space. By definition, Fvdv = Fλdλ, meaning that ∫Fλdλ = ∫Fvdv, as long as the bounds of integration are properly adjusted. Since 1 + z = 1 + v/c = λ/λr, it implies dλ/dv = λr/c.

In the case of a double Gaussian fitting, Fv will be the sum of both Gaussian components: the narrow and the broad. GELATO outputs σ and F, so we can calculate W80 and vout. A map illustrating the spatial distribution of the outflow velocity (in terms of W80 and vout) is shown for two of the 13 dwarf galaxies with AGN outflow candidates (Figs. 2 and 3). In Table 2, the results are summarised by taking the median of all the AGN and composite spaxels with outflow signatures.

Table 2.

Properties of the outflows for the MaNGA sample of 13 dwarf galaxies.

There are two different scenarios for calculating the outflow radius: when the full extent of the outflow is not observed (Fig. 2) and when it is observed (Fig. 3). For each of these two cases, a different criteria is adopted:

  • If it is fully extended: we consider that the diameter is the largest distance of consecutive AGN and composite spaxels with outflow signatures.

  • If it is not fully extended: We consider the radius as the distance from the central spaxel to the further AGN or composite spaxel with outflow signature that is in a region with at least ten consecutive AGN or composite spaxels with outflow signatures.

As each spaxel is 0.5 arcsec size, the angular size of the outflow can be calculated. Then, through parallax, the projected radius is obtained:

R out = tan ( α ) d l , $$ \begin{aligned} R_{\rm out}=\mathrm{tan}(\alpha )\,d_{\rm l}, \end{aligned} $$(6)

where α is half of the angular size in radians and dl is the luminosity distance. We use the projected radius as a lower limit of the outflow radius, since the physical size depends on the inclination angle of the outflow with respect to the line of sight.

3.2. Energetics of the outflow

The ionised gas mass of the outflow can be calculated based on the luminosity of the Hα emission line. Considering the sum of all the AGN and composite spaxels with outflow signatures and the equation defined by Osterbrock & Ferland (2006), we have

M out = ∑ i = 1 n out m out = ∑ i = 1 n out 4.48 M ⊙ ( L H α 10 35 erg s − 1 ) ( ⟨ n e ⟩ 1000 cm − 3 ) − 1 , $$ \begin{aligned} M_{\rm out}=\sum _{i = 1}^{n_{\rm out}} m_{\rm out}=\sum _{i = 1}^{n_{\rm out}} 4.48\,M_\odot \left(\frac{L_{\mathrm{H} \alpha }}{10^{35}\,\mathrm{erg\,s^{-1}}}\right)\left(\frac{\langle n_{\rm e}\rangle }{1000\,\mathrm{cm}^{-3}}\right)^{-1}, \end{aligned} $$(7)

where Mout is the total mass of the outflow, mout is the outflow mass within individual spaxels, and LHα is the Hα luminosity of the individual outflow spaxels. Focusing on those spaxels with outflow, the electron density, ne is estimated using the [SII]λ6716 Å/[SII]λ6732 Å relation (Sanders et al. 2015):

n e ( R ) = c R − a b a − R , $$ \begin{aligned} n_{\rm e}(R) = \frac{cR-ab}{a-R}, \end{aligned} $$(8)

where R is the ratio between the [SII] emission lines flux and a, b and c are dimensionless constants equal to 0.4314, 2.107 and 627.1 respectively. The electron density ranges from 825 cm−3 < ne < 1071 cm−3.

As the detected outflows are spatially resolved, the mass (dM/dt), momentum (dP/dt), and kinetic energy (dE/dt) rates can be calculated using the following equations (e.g. Liu et al. 2020; Bohn et al. 2021):

d M / d t = ∑ i = 1 n out d m / d t = ∑ i = 1 n out m out v 0 sec θ R out , $$ \begin{aligned} \mathrm{d}M/\mathrm{d}t&=\sum _{i = 1}^{n_{\rm out}} \mathrm{d}m/\mathrm{d}t=\sum _{i = 1}^{n_{\rm out}} \frac{m_{\rm out}v_0\mathrm{sec}\theta }{R_{\rm out}},\end{aligned} $$(9)

d P / d t = ∑ i = 1 n out ( v 0 sec θ ) d m / d t , $$ \begin{aligned} \mathrm{d}P/\mathrm{d}t&=\sum _{i = 1}^{n_{\rm out}} (v_0\mathrm{sec}\theta )\mathrm{d}m/\mathrm{d}t,\end{aligned} $$(10)

d E / d t = 1 2 ∑ i = 1 n out [ ( v 0 sec θ ) 2 + 3 σ out 2 ] d m / d t , $$ \begin{aligned} \mathrm{d}E/\mathrm{d}t&=\frac{1}{2}\sum _{i = 1}^{n_{\rm out}} \left[(v_0\,\mathrm{sec}\theta )^2+3\sigma _{\rm out}^2\right]\mathrm{d}m/\mathrm{d}t, \end{aligned} $$(11)

where v0 here is the absolute value of the velocity offset between the narrow and broad component (see Eq. 1), σout is the velocity dispersion measured from the outflow component within individual spaxels, and θ = sin−1(rspaxel/Rout), where rspaxel is the angular size of the spaxel converted into physical distance. The results of the mass, momentum and kinetic energy rate range from 10−4.4 < dM/dt (M⊙ yr−1) < 10−1.2, from 104.3 < c dP/dt (L⊙) < 108.6, and from 1035.4 < dE/dt (erg s−1) < 1039.5. These results are also listed in Table 2.

3.3. Outflows escaping their host halos

To investigate whether the outflows are able to escape the gravitational potential of their host halos, the velocity of the outflows and the escape velocity, vesc, are compared. Assuming a Navarro-Frenk-White profile (NFW; Navarro et al. 1997) and making use of the abundance matching (Moster et al. 2010, 2013), we can infer the halo mass from the stellar mass of the galaxy:

M ∗ M h = 2 ( M ∗ M h ) 0 [ ( M h M 1 ) − β + ( M h M 1 ) γ ] − 1 . $$ \begin{aligned} \frac{M_*}{M_{\rm h}} = 2\left(\frac{M_*}{M_{\rm h}}\right)_{0}\left[\left(\frac{M_{\rm h}}{M_1}\right)^{-\beta }+\left(\frac{M_{\rm h}}{M_1}\right)^{\gamma }\right]^{-1}. \end{aligned} $$(12)

This expression has four parameters: the normalisation of the stellar-to-halo mass ratio (M*/Mh)0 = 0.02817, a characteristic mass of M1 = 1011.899, and β = 0.611, γ = 1.068, with these latter two slopes serving as an indication of the behaviour of the mass ratio at the low-mass and high-mass ends, respectively. Then, a gravitational potential is considered at the centre ϕ(r = 0) from a spherical NFW profile (Łokas & Mamon 2001):

ϕ ( r = 0 ) = − 4 3 c g ( c ) π G r v 2 v ρ c 0 , $$ \begin{aligned} \phi (r = 0) = -\frac{4}{3}cg(c)\pi Gr_{\rm v}^2 v\rho _{\rm c}^0, \end{aligned} $$(13)

where c = 10 is the concentration parameter, g(c) = 1/(ln(1 + c)−c/(1 + c)) is a function of the concentration parameter, v = 200 is the virial overdensity, ρc0 = 277.5 M⊙/kpc3 is the present critical density, and rv = (3Mh/4πvρc0)1/3 is the virial radius. Finally, using the expression defined in Manzano-King et al. (2019) we obtain the escape velocity:

v esc 2 = 2 | ϕ ( r = 0 ) | . $$ \begin{aligned} v_{\rm esc}^2 = 2|\phi (r = 0)|. \end{aligned} $$(14)

In Fig. 4 the velocity of the outflow is compared with vesc. While W80 is the most common value when defining the outflow velocity, vout has recently been also adopted as outflow velocity estimator (e.g. Manzano-King et al. 2019; Bohn et al. 2021; Zheng et al. 2023). According to both W80 and vout, 8252−1902, 8442−1901 and 8982−3703 would have outflows consistent with the high enough velocity needed to escape the galaxy and the dark matter halo. Also, in Fig. 4, we show a colourbar with the kinetic energy rate of the galaxies. The three dwarf galaxies whose outflows are able to escape are among the five with the highest kinetic energy rate. For these three galaxies, the outflowing gas could contribute to AGN feedback by removing gas from the central region or could cause feedback to the galactic halo by heating the intergalactic medium (Smethurst et al. 2021).

thumbnail Fig. 4.

Comparison of the ratios of the two velocity indicators W80 and vout with vesc. The grey dashed lines represent the 1:1 relation. Three out of the 13 MaNGA dwarf galaxies with AGN outflow candidates would have outflows fast enough to become unbound from the galaxy and from its dark matter halo.

3.4. Driver behind these outflows

Although we have primarily based our analysis on the selection of AGN spaxels to detect AGN outflows, it was necessary to engage in a deeper study in order to understand the origin of the outflows detected.

3.4.1. Considering stellar winds from massive stars or SNe as a potential driver

Massive stars are able to inject energy and momentum to the interstellar medium through outflows coming from stellar winds and SNe explosions. These stellar processes are commonly believed to be the main source of feedback in dwarf galaxies (e.g. Veilleux et al. 2005; Martín-Navarro & Mezcua 2018). In this section, we investigate whether the 13 MaNGA dwarf galaxies with outflows could be driven by stellar processes by computing the expected energy driven by stellar wind outflows in massive stars or by SNe outflows.

For stellar winds in massive stars, the kinetic energy rate can be derived as d E / d t = 1 2 d M / d t v out 2 $ \mathrm{d}E/\mathrm{d}t=\frac{1}{2}\mathrm{d}M/\mathrm{d}t\,v_{\mathrm{out}}^2 $ (Rosen 2022), where dM/dt is the mass rate, which typically ranges from 10−5 to 10−4 M⊙ yr−1 (Bally 2008). The outflow velocity in massive stars rarely surpasses 100 km s−1 (e.g. Churchwell 1997; Commerçon et al. 2022). Consequently, the kinetic energy rate is expected to be in the range of 1034 − 35 erg s−1, which is lower than our results. Thus, we can safely rule out massive stars from the plausible driving mechanisms.

For SNe, the kinetic energy rate can be calculated as ∼7 × 1041(SFR/M⊙ yr−1) (Veilleux et al. 2005). We calculated the SFR by taking the total Hα and Hβ emission lines from the MaNGA DAP and using the equation log(SFR) = log(LHα)−41.27 (Kennicutt & Evans 2012), after having previously corrected LHα from extinction using the Calzetti et al. (2000) extinction law. A range of SNe kinetic energy rate that goes from dE/dt ∼ 1040 erg s−1 to ∼1042 erg s−1 is obtained. The expected SNe kinetic energy rate is one to six orders of magnitude higher than the kinetic energy rate of the 13 MaNGA dwarf galaxies with outflows. Consequently, in terms of energy, the 13 outflows detected could be driven by SNe as the kinetic energy rate of the SNe is able to explain the presence of outflows up to 1040 − 42 erg s−1. However, the SFR might be overestimated in those cases where the AGN dominates (see Fig. 3). This comes from the fact that the Hα emission line comes mostly from the AGN. To overcome this problem, we derived the SFR from the SED fitting (see Sect. 2.2). The SFR obtained from the SED fitting is similar or even higher with those obtained from the Hα emission line, which is in agreement with the results obtained from Siudek et al. (2024). Alternatively, we also used a SFR indicator based on the [OII]λ3726 emission line (SFR(OII); e.g. Kennicutt 1998; Hopkins et al. 2003; Kewley et al. 2004). We corrected the [OII]λ3726 emission line for extinction following Calzetti et al. (2000) extinction law and correct for AGN contribution by subtracting ten per cent of the total [OIII]λ5007 luminosity (Vietri et al. 2022). The SFR(OII) obtained is typically one order of magnitude lower than that obtained from the Hα emission line and from the SED fitting. However, the SNe kinetic energy rate computed from the SFR(OII) still remains several orders of magnitude higher than the kinetic energy rate of the 13 MaNGA dwarf galaxies with outflows.

3.4.2. AGN fraction and W80 velocity

In this section, we look at the AGN fraction, fAGN = nagn/(nagn + nSF), with nagn as the number of spaxels classified as AGN and nSF the number of spaxels classified as SF. A Mann-Whitney U test was performed to check whether the W80 velocity of the 12 dwarf galaxies with AGN outflow candidates (8442−1901 is excluded as all the spaxels with S/N ≥ 10 are composite6, see Sect. 2.1) comes from different distributions in cases where the AGN fraction is taken into account. We find that when the AGN fraction of > 0.45, while the W80 velocities are drawn from a different distribution at a ∼99% confidence level, with a p-value of ∼0.004 (see Fig. 5). These velocities are higher (from 231 km s−1 to 566 km s−1) than those corresponding to the galaxies with an AGN fraction < 0.45 (from 205 km s−1 to 227 km s−1). Eight out of the 12 MaNGA dwarf galaxies have an AGN fraction > 0.45, six of them are classified as AGN, and two as SF-AGN (see Table 2). Although the galaxy 8442−1901 was not included in this analysis, its W80 velocity (396 km s−1) is consistent with those from the eight galaxies with AGN fraction > 0.45. Therefore, since AGN outflows are expected to be faster than those driven by stellar processes (Liu et al. 2020; Wylezalek et al. 2020), the AGN emerges as the main candidate to act as the driving mechanism of the outflows observed in these nine galaxies.

thumbnail Fig. 5.

Distribution of the W80 velocity for the dwarf galaxies with AGN outflow candidates with AGN fraction ≤0.45 (blue), and with AGN fraction > 0.45 (orange).

3.4.3. Study of the SF spaxels

We also studied the detection of outflows in the SF spaxels of our sample of 13 dwarf galaxies following the same procedure for the individual spaxels as that described in Sect. 2.1. We excluded the galaxy 8442−1901 as it presents only composite spaxels. For the six dwarf galaxies classified as SF-AGN (see Table 2), the W80 is consistent within the errors when comparing the SF spaxels against the AGN spaxels. On the contrary, For the six dwarf galaxies classified as AGN, four of them (10223−3702, 12487−3704, 8252−1902 and 8982−3703) have an AGN fraction fAGN ∼ 1 so, although we were not able to carry out this analysis, the absence of spaxels ionised by SF suggests that for these three galaxies the AGN must drive the outflow. The two remaining galaxies (8655−6102 and 9889−1902) present lower W80 velocities in the SF spaxels than in the AGN spaxels, even when accounting for their errors. This indicates that for these two galaxies the AGN could be the main candidate for driving the outflows.

Regarding the kinetic energy rate we find consistent values, within the errors, for those obtained from SF spaxels compared to those obtained from AGN spaxels. However, for the galaxy 9889−1902, the kinetic energy rate calculated from the AGN spaxels is one order of magnitude higher. This makes 9889−1902 a strong candidate for hosting AGN outflows, as they are expected to be more energetic than SF outflows (Aravindan et al. 2023).

3.4.4. AGN bolometric luminosity and kinetic energy rate

The AGN bolometric luminosity (Lbol) is given by Lbol = 1000 × L[OIII] (Moran et al. 2014), where L[OIII] is the sum of the [OIII]λ5007 luminosities of those spaxels classified as AGN or composite in the BPT diagrams (Mezcua & Domínguez Sánchez 2024). The AGN bolometric luminosity Lbol for the 13 MaNGA dwarf galaxies with AGN outflow candidates (see Table 3), is four to six orders of magnitude higher than the kinetic energy rate of the outflow, implying that the AGNs in these 13 galaxies are capable of driving these outflows (Liu et al. 2020).

Table 3.

AGN luminosity for the MaNGA sample of 13 dwarf galaxies.

In summary, for the seven MaNGA dwarf galaxies classified as AGN, six are more likely to have outflows driven by the AGN, as indicated by their AGN fraction being higher than that of the six dwarf galaxies classified as SF-AGNs (with their Lbol being greater than the kinetic energy rate and a higher W80 velocity in the AGN spaxels than in the SF spaxels). For the remaining dwarf galaxy classified as an AGN (8442−1901), the outflow is also likely to be driven by the AGN based on its high W80 velocity and its bolometric luminosity. In the case of the six MaNGA dwarf galaxies classified as SF-AGNs, two (10226−1901 and 11754−3701) have an AGN fraction and W80 value consistent with their having been drawn from the same distribution as the five AGN galaxies. For the remaining four SF-AGNs, it is difficult to distinguish the driving mechanism of the outflow, therefore both SNe and AGNs are possible candidates for driving these outflows.

At this point, we divided the sample of 13 MaNGA dwarf galaxies with AGN outflow candidates in two sub-samples: nine MaNGA dwarf galaxies consistent with AGN outflows (the seven AGNs and two SF-AGNs: 10226 and 11754−3701) and four MaNGA dwarf galaxies with uncertain outflows (the remaining four SF-AGNs). We emphasise that for all the 13 dwarf galaxies we cannot rule out the possibility of SNe as the driving mechanism in this scenario, based on the fact that the SNe kinetic energy rate exceeds the outflow kinetic energy rate in all cases.

3.5. Mass-loading factor

The mass-loading factor is a parameter used to compare the amount of gas that is ejected by the outflow to that consumed by SF (Harrison & Ramos Almeida 2024). It is defined as the ratio of the outflow mass rate compared with the SFR, η = (dM/dt)/SFR. In the MaNGA sample of nine dwarf galaxies with AGN outflows, we find that the mass-loading factor for all of them is η ≪ 1 for all three SFR indicators: the one obtained from the Hα emission line, the one derived from SED fitting, and the SFR(OII). This implies that the amount of gas used for SF processes is greater than the one ejected by the outflow, therefore these outflows are unlikely to quench the SF in the host galaxy (Carniani et al. 2024). However, the mass-loading factor can be misleading due to the fact that the spatial scale when calculating the energetic properties of the outflows and the scale needed to obtain the integrated SFR are different; namely, scales of a few kiloparsecs versus galaxy scales (Harrison & Ramos Almeida 2024), except for the galaxies 11754−3701 and 11826−12702, where Rout > 5 kpc.

3.6. Comparison with other works on dwarf galaxies

In this section, we compare the outflow velocity (as vout and W80) and luminosity obtained for the sample of 13 MaNGA dwarf galaxies with AGN outflow candidates with other studies. For the nine MaNGA dwarf galaxies with AGN outflows the W80 velocity ranges from 231 to 566 km s−1 and vout from 107 to 532 km s−1. In the case of 9889−1902 and 7992−6102, their values for W80 (333 km s−1 and 226 km s−1, respectively) and vout (206 km s−1 and 120 km s−1, respectively) could be enhanced due to the merger process in which they are involved. For the four MaNGA dwarf galaxies with uncertain outflows, the W80 values range from 205 to 226 km s−1 and vout from 96 to 132 km s−1. These results were compared with those of other dwarf galaxies with AGN outflow candidates. By using IFU spectroscopy, Liu et al. (2020) studied 8 out of the 29 dwarf galaxies previously studied in Manzano-King et al. (2019) using longslit spectroscopy. The authors detected AGN outflows with W80 ranging 140−980 km s−1, considering a [OIII]λ5007 Å line profile with up to three Gaussians components (i.e. C1, C2, and C3), where C2 and C3 show strong evidence of outflows. Bohn et al. (2021) also studied the same sample as Manzano-King et al. (2019) but focusing on the [Si VI] 1.9630 μm line instead of the [OIII]λ5007 Å and obtaining an outflow velocity range of 155 ≤ vout (km s−1)≤770 and 108 ≤ W80 (km s−1)≤1350. Zheng et al. (2023) found AGN outflows with a velocity of vout = 471 km s−1 for SDSS J0228−0901. The velocities of the 13 MaNGA dwarf galaxies with AGN outflow candidates are therefore consistent with those previously studied.

Regarding the luminosity, we compared the L[OIII] derived from the sum of the AGN and composite spaxels against the total [OIII] luminosity values from the literature; thus, we had to take certain considerations for both dwarf and massive galaxies in this context.

For the dwarf galaxies reported in Liu et al. (2020), the AGN spaxels contribute at least ∼95% of the [OIII] luminosity, which ranges from ∼1039 erg s−1 to ∼1041 erg s−1, making our results comparable. For the MaNGA sample of nine dwarf galaxies with AGN outflows, we obtain L[OIII] that ranges from ∼1038 erg s−1 to ∼1041 erg s−1. These results agree with those in Liu et al. (2020). For the four dwarf galaxies with uncertain outflows, three of them present L[OIII] ∼ 1037 erg s−1, which is two to four orders of magnitude lower than those from Liu et al. (2020), possibly indicating a different origin.

In the next section, we compare our results with massive galaxies where the AGN dominates, assuming that the total [OIII] luminosity comes mostly from the AGN.

3.7. Comparison with massive galaxies

In Fig. 6, we show the W80 velocity versus L[OIII] for the MaNGA sample of 13 dwarf galaxies with AGN outflow candidates reported here, as well as the dwarf galaxies with AGN outflows studied in Liu et al. (2020), the low-to-intermediate luminosity massive galaxies with AGN outflows from Wylezalek et al. (2020), the low-luminosity active galactic nuclei (LLAGNs) with AGN outflows studied in Heckler et al. (2022), the z ∼ 0.5 radio-quiet quasars with AGN outflows studied in Liu et al. (2013), the z < 0.2 massive type-2 AGN quasars with AGN outflows studied in Harrison et al. (2014), and the z ∼ 2 massive type-1 AGN with AGN outflows studied in Kakkad et al. (2020). The typical velocity of the AGN outflow candidates in dwarf galaxies appears to be < 500 km s−1, while those of massive galaxies often surpasses the 500 km s−1 mark and (in some cases) even 2000 km s−1 (e.g. Kakkad et al. 2020). However, 29 massive but low-intermediate luminosity galaxies from Wylezalek et al. (2020), whose W80 velocities have been recalculated in this paper present velocities of W80 < 500 km s−1. We performed a linear regression on the sample of massive galaxies (Liu et al. 2013; Harrison et al. 2014; Wylezalek et al. 2020; Kakkad et al. 2020), henceforth denoted as the ‘massive sample’. We also fit a linear regression including just the nine MaNGA dwarf galaxies with AGN outflows in addition to those studied in Liu et al. (2020) (henceforth ‘dwarf sample’), along with another one for all the galaxies, massive galaxies, and dwarf galaxies, but excluding the four MaNGA dwarf galaxies with uncertain outflows. Comparing the results of all the regression lines (see Table 4), the one that includes all the galaxies and the one representing the massive sample have similar values for both the slope and the correlation coefficient, based on the fact that the regression lines almost overlapped. They have higher correlation coefficient than the one for the dwarf sample. These results could suggest that the AGN outflows in dwarf galaxies can be produced by the same mechanisms as those in massive galaxies, being the dwarf galaxies a scaled-down version of the massive ones (see also Liu et al. 2020). However, the slope of the fit for the dwarf sample is ∼76% lower than the one for the all galaxies sample and for the massive sample (see Table 4). This may indicate that the W80 velocity for the dwarf sample does not follow the same trend as all the galaxies.

Table 4.

Regression lines of the W80 vs L[OIII] correlation.

thumbnail Fig. 6.

W80 velocity vs L[OIII] for the MaNGA sample of 13 dwarf galaxies with AGN outflow candidates (red squares for AGN galaxies, red diamonds for SF-AGN galaxies), the AGN dwarf galaxy sample of Liu et al. (2020) (red triangles), the z < 2 type-2 AGN quasars of Harrison et al. (2014) (grey pluses), the radio-quiet quasars of Liu et al. (2013) (grey crosses), the z ∼ 2 type-1 massive AGN of Kakkad et al. (2020) (grey diamonds), the massive LLAGN of Heckler et al. (2022) (grey circle) and the MaNGA sample of low-intermediate luminosity massive galaxies of Wylezalek et al. (2020) (grey squares). The solid lines are regression fits to the massive sample (grey), the dwarf sample of AGN outflows, which includes the nine MaNGA dwarf galaxies with AGN outflows and those from Liu et al. (2020) (red), and the combination of all the massive galaxies, along with the dwarf sample of AGN outflows (purple) indicated as the shaded area the 1σ error of the regression line.

The kinetic energy outflow rate is a parameter used as a measure of how powerful the outflow is. The values of kinetic energy rate normalised by Lbol (i.e. outflow efficiency) are shown in Table 3. Figure 7 shows, following Liu et al. (2020), dE/dt versus Lbol (left) and (dE/dt)/Lbol versus Lbol (right) for the MaNGA sample of 13 dwarf galaxies with AGN outflow candidates. We also refer to Liu et al. (2020), Wylezalek et al. (2020) (whose kinetic energy rates are derived in this paper), Heckler et al. (2022), and Liu et al. (2013) in addition to the type 1 quasars with AGN outflows of Rupke et al. (2017), the different AGN at different redshifts (z = 0.1 − 3) of Fiore et al. (2017), and ultraluminous infrared galaxies (ULIRGs) with AGN outflows of Rose et al. (2018). The bolometric luminosity was calculated from the L[OIII] using a bolometric correction factor of 1000 (Moran et al. 2014), except for the galaxies in Rupke et al. (2017) and Fiore et al. (2017), whose L[OIII] values are not reported and where the bolometric luminosity has been calculated either from Lbol = 1.15 × LIR (with LIR being the infrared luminosity) and from the fitting of the optical-ultraviolet spectral energy distribution (SED). The kinetic energy rate of the 13 MaNGA dwarf galaxies with AGN outflows candidates is four to six orders of magnitude lower than their bolometric luminosity. This suggests that it would be difficult for the outflow to have a significant impact in the galaxy as a source of feedback (Wang et al. 2024), although there may be additional outflow mass hidden in other phases (Belli et al. 2024) such as warm and cold molecular gas phases. This result is consistent with findings related to the mass-loading factor, where we find that the outflows are unlikely to quench the SF (see Sect. 3.5). Additionally, a linear regression is performed between log(dE/dt) vs log(Lbol), including all the galaxies except for the four MaNGA dwarf galaxies with uncertain outflows. The correlation coefficient is R = 0.9 indicating that both dwarf and massive galaxies behave similar. In addition to this, a Mann-Whitney U rank test was performed on two independent samples (i.e. massive and dwarf galaxies) returning a p-value > 0.3 in both cases. Consequently, the null hypothesis that the (dE/dt)/Lbol of dwarf and massive galaxies is drawn from the same distribution cannot be rejected. These results may indicate that AGN outflows in dwarf galaxies behave in a similar way as that of massive galaxies, in agreement with the linear correlation when comparing the outflow velocity against the [OIII] luminosity in the ‘all galaxies sample’ of Fig. 6. However, the outflow efficiency does not only depend on the AGN luminosity, but also on other factors, such as the orientation of the outflow with respect to the galaxy disk (Harrison & Ramos Almeida 2024). Thus, a more robust study ought to be performed, which is beyond the scope of the current work.

thumbnail Fig. 7.

Ionised outflow kinetic energy rate vs AGN bolometric luminosity (left) and ratio of the kinetic energy outflow rate to the AGN bolometric luminosity (right) for the MaNGA sample of dwarf galaxies with AGN outflow candidates (red squares for AGN galaxies, red diamonds for SF-AGN galaxies), the AGN dwarf galaxy sample of Liu et al. (2020) (red triangles), radio-quiet quasars of Liu et al. (2013) (grey crosses), type 1 quasars of Rupke et al. (2017) (grey diamonds), the ULIRGs of Rose et al. (2018) (grey pluses), MaNGA sample of low-intermediate luminosity massive galaxies of Wylezalek et al. (2020) (grey squares), and the massive LLAGN of Heckler et al. (2022) (grey circle). The purple solid line is a regression fit considering all the massive galaxies, the nine MaNGA dwarf galaxies with AGN outflows, and those from Liu et al. (2020). The shaded area corresponds to the 1σ error of the regression line.

4. Conclusions

Dwarf galaxies are the most abundant type of galaxies and the building blocks of the more massive ones. Although they were commonly assumed to be regulated by SNe feedback, thousands of AGNs have been found in dwarf galaxies in recent decades. Furthermore, recent studies have shown evidence for the presence of AGN outflows and feedback that may have an impact in the growth of the galaxy. In this paper, the presence of AGN outflows in a sample of 2292 dwarf galaxies, characterised by strong AGN signatures (Mezcua & Domínguez Sánchez 2024), has been studied. These galaxies are drawn from the MaNGA survey. The stacked spectrum of all those spaxels classified as AGN or composite are fitted using the GELATO python code. For the galaxies where the broad component in the [OIII]λ5007 Å emission line has been fitted, GELATO was run again through all the AGN and composite spaxels to spatially resolve this outflowing component and study the kinematic and energetic properties. We found 13 new dwarf galaxies with AGN outflow candidates. The main results of this study are:

  • For nine out of the 13 MaNGA AGN dwarf galaxies, there is a misalignment between the gas and the stellar kinematics, which may indicate the presence of potential outflows.

  • The outflow velocity, W80, of the 13 MaNGA AGN dwarf galaxies ranges from 205 to 566 km s−1. This result is lower than the typical values of massive galaxies, which often surpass 500 km s−1, reaching values above 1000 km s−1.

  • For nine out of the 13 MaNGA AGN dwarf galaxies, the AGN is the most likely mechanism to drive these outflows based on the AGN fraction, the W80 velocity, and the Lbol (see Sect. 3.4). For the remaining four galaxies, both SNe and AGNs are possible candidates for driving these outflows.

  • The AGN outflow velocity, W80, in three out of the 13 MaNGA dwarf galaxies with outflows is higher than the dark matter halo escape velocity. For these galaxies, an enrichment of the circumgalactic medium is expected.

  • We find a correlation between outflow velocity, W80 and L[OIII], which extends all the way from massive to dwarf galaxies with AGN outflows. This may indicate that AGN outflows in dwarf galaxies can be a scaled-down version of massive galaxies. However, the slope of the regression line of the dwarf sample is lower than the one including all galaxies, suggesting that the correlation might not be followed by the AGN outflows in dwarf galaxies.

  • The kinetic energy outflow rate for the sample of 13 MaNGA dwarf galaxies with AGN outflow candidates ranges from 1035 to 1039 erg s−1. The relation between kinetic energy rate and the bolometric luminosity seems to follow a linear relationship through all the mass regime. However, we find that 3 out of the 13 MaNGA dwarf galaxies with AGN outflows signatures do not follow this linear relationship, although the origin of the outflow for these galaxies is uncertain. Also, we find that massive and dwarf galaxies are drawn from the same distribution when comparing the outflow efficiency against the bolometric luminosity.

In this work, we provide arguments in favour of AGN outflows in dwarf galaxies behaving similarly to those in massive galaxies. However, we find three galaxies that do not follow the linear trends showed in this work, although the driving mechanism that causes the outflow in these three galaxies is not well known. Therefore, the necessity of new detections of AGN outflows in dwarf galaxies is crucial in order to have a better understanding of them, as well as their impact and role in galaxy evolution. In a future work, we plan to study the stellar population properties and examine the effects of SNe in these 13 dwarf galaxies with AGN outflow candidates to further understand the feedback processes and their impact on galaxy evolution.

2

Low-ionisation emission line region (LINER).

6

8442−1901 is however classified as an AGN by Mezcua & Domínguez Sánchez (2024) as they consider spaxels with a S/N ≥ 3.

Acknowledgments

The authors thank the anonymous referee for insightful comments. The authors acknowledge the feedback received from Rogemar Riffel and Rogério Riffel from the Universidade Federal de Santa Maria and from the Universidade Federal do Rio Grande do Sul respectively and from all the colleagues of the AGN and galaxy evolution group of the Institue of Space Sciences of Barcelona. The authors also acknowledge the help provided by Raphael Hviding from the Arizona University and the Max-Planck Institute for Astronomy. VRM acknowledges support from the Spanish Ministry of Science, Innovation and Universities through the project PRE2022-104649. MM and AE acknowledge support from the Spanish Ministry of Science and Innovation through the project PID2021-124243NB-C22. This work was partially supported by the program Unidad de Excelencia María de Maeztu CEX2020-001058-M. HDS acknowledges the financial support from the Spanish Ministry of Science and Innovation and the European Union – NextGenerationEU through the Recovery and Resilience Facility project ICTS MRR-2021-03-CEFCA and financial support provided by the Governments of Spain and Aragón through their general budgets and the Fondo de Inversiones de Teruel. HDS acknowledges financial support by RyC2022-030469-I grant, funded by MCIN/AEI/10.13039/501100011033 and FSE+. AA acknowledges funding from the MICINN (Spain) through the Juan de la Cierva-Formación program under contract FJC2020-046224-I and support by the European Union grant WIDERA ExGal-Twin, GA 101158446. F.M.-S. acknowledges support from NASA through ADAP award 80NSSC19K1096. M.S. acknowledges support by the State Research Agency of the Spanish Ministry of Science and Innovation under the grants ‘Galaxy Evolution with Artificial Intelligence’ (PGC2018-100852-A-I00) and ‘BASALT’ (PID2021-126838NB-I00) and the Polish National Agency for Academic Exchange (Bekker grant BPN/BEK/2021/1/00298/DEC/1). This work was partially supported by the European Union’s Horizon 2020 Research and Innovation program under the Maria Sklodowska-Curie grant agreement (No. 754510).

References

  1. Akaike, H. 1974, IEEE Trans. Autom. Control, 19, 716 [Google Scholar]
  2. Aravindan, A., Liu, W., Canalizo, G., et al. 2023, ApJ, 950, 33 [Google Scholar]
  3. Arjona-Galvez, E., Di Cintio, A., & Grand, R. J. 2024, A&A, 690, A286 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Baldassare, V. F., Reines, A. E., Gallo, E., & Greene, J. E. 2015, ApJ, 809, L14 [NASA ADS] [CrossRef] [Google Scholar]
  5. Baldwin, J. A., Phillips, M. M., & Terlevich, R. 1981, PASP, 93, 5 [Google Scholar]
  6. Bally, J. 2008, ASP Conf. Ser., 387, 158 [Google Scholar]
  7. Barai, P., & Dal Pino, E. M. D. G. 2018, Proc. Int. Astron. Union, 14, 154 [Google Scholar]
  8. Belfiore, F., Westfall, K. B., Schaefer, A., et al. 2019, AJ, 158, 160 [CrossRef] [Google Scholar]
  9. Belli, S., Park, M., Davies, R. L., et al. 2024, Nature, 630, 54 [NASA ADS] [CrossRef] [Google Scholar]
  10. Bellovary, J. M., Cleary, C. E., Munshi, F., et al. 2019, MNRAS, 482, 2913 [NASA ADS] [Google Scholar]
  11. Bianchi, L., Herald, J., Efremova, B., et al. 2011, Ap&SS, 335, 161 [Google Scholar]
  12. Bianchi, L., Shiao, B., & Thilker, D. 2017, ApJS, 230, 24 [Google Scholar]
  13. Birchall, K. L., Watson, M., & Aird, J. 2020, MNRAS, 492, 2268 [NASA ADS] [CrossRef] [Google Scholar]
  14. Bohn, T., Canalizo, G., Veilleux, S., & Liu, W. 2021, ApJ, 911, 70 [Google Scholar]
  15. Boquien, M., Burgarella, D., Roehlly, Y., et al. 2019, A&A, 622, A103 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Bower, R. G., Benson, A., Malbon, R., et al. 2006, MNRAS, 370, 645 [Google Scholar]
  17. Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 [NASA ADS] [CrossRef] [Google Scholar]
  18. Bundy, K., Bershady, M. A., Law, D. R., et al. 2015, ApJ, 798, 7 [Google Scholar]
  19. Bykov, S., Gilfanov, M., & Sunyaev, R. 2024, MNRAS, 527, 1962 [Google Scholar]
  20. Calzetti, D., Armus, L., Bohlin, R. C., et al. 2000, ApJ, 533, 682 [NASA ADS] [CrossRef] [Google Scholar]
  21. Capelo, P. R., Feruglio, C., Hickox, R. C., & Tombesi, F. 2024, Handbook of X-ray and Gamma-ray Astrophysics (Springer), 4567 [CrossRef] [Google Scholar]
  22. Carniani, S., Venturi, G., Parlanti, E., et al. 2024, A&A, 685, A99 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Casanueva, C. I., Lagos, C. D. P., Padilla, N. D., & Davison, T. A. 2022, MNRAS, 514, 2031 [NASA ADS] [Google Scholar]
  24. Chabrier, G. 2003, ApJ, 586, L133 [NASA ADS] [CrossRef] [Google Scholar]
  25. Chilingarian, I. V., Katkov, I. Y., Zolotukhin, I. Y., et al. 2018, ApJ, 863, 1 [Google Scholar]
  26. Churchwell, E. 1997, ApJ, 479, L59 [Google Scholar]
  27. Cid Fernandes, R. C., Stasińska, G., Schlickmann, M., et al. 2010, MNRAS, 403, 1036 [NASA ADS] [CrossRef] [Google Scholar]
  28. Clavijo-Bohórquez, W. E., de Gouveia Dal Pino, E. M., & Melioli, C. 2024, MNRAS, 535, 1696 [Google Scholar]
  29. Comerford, J. M., & Greene, J. E. 2014, ApJ, 789, 112 [NASA ADS] [CrossRef] [Google Scholar]
  30. Commerçon, B., González, M., Mignon-Risse, R., Hennebelle, P., & Vaytet, N. 2022, A&A, 658, A52 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Croton, D. J., Springel, V., White, S. D., et al. 2006, MNRAS, 365, 11 [NASA ADS] [CrossRef] [Google Scholar]
  32. Cutri, R., Skrutskie, M., Van Dyk, S., et al. 2003, VizieR Online Data Catalog: II/246 [Google Scholar]
  33. Cutri, R., Skrutskie, M., van Dyk, S., et al. 2012, VizieR Online Data Catalog: II/281 [Google Scholar]
  34. Cutri, R. E., Wright, E., Conrow, T., et al. 2021, VizieR Online Data Catalog: II/328 [Google Scholar]
  35. da Silva, P., Steiner, J. E., & Menezes, R. 2017, MNRAS, 470, 3850 [Google Scholar]
  36. Doi, M., Tanaka, M., Fukugita, M., et al. 2010, AJ, 139, 1628 [NASA ADS] [CrossRef] [Google Scholar]
  37. Draine, B., Aniano, G., Krause, O., et al. 2013, ApJ, 780, 172 [NASA ADS] [CrossRef] [Google Scholar]
  38. Falceta-Gonçalves, D., Caproni, A., Abraham, Z., Teixeira, D., & Dal Pino, E. D. G. 2010, ApJ, 713, L74 [Google Scholar]
  39. Filippenko, A. V., & Sargent, W. L. 1989, ApJ, 342, L11 [NASA ADS] [CrossRef] [Google Scholar]
  40. Fiore, F., Feruglio, C., Shankar, F., et al. 2017, A&A, 601, A143 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  41. Fritz, J., Franceschini, A., & Hatziminaoglou, E. 2006, MNRAS, 366, 767 [Google Scholar]
  42. Fukugita, M., Shimasaku, K., Ichikawa, T., et al. 1996, The Sloan Digital Sky Survey Photometric System, Tech. Rep., SCAN-9601313 [Google Scholar]
  43. Gatto, A., Walch, S., Naab, T., et al. 2017, MNRAS, 466, 1903 [NASA ADS] [CrossRef] [Google Scholar]
  44. Greene, J. E., & Ho, L. C. 2007, ApJ, 670, 92 [NASA ADS] [CrossRef] [Google Scholar]
  45. Hao, L., Strauss, M. A., Tremonti, C. A., et al. 2005, AJ, 129, 1783 [Google Scholar]
  46. Harrison, C. M., & Ramos Almeida, C. 2024, Galaxies, 12, 17 [NASA ADS] [CrossRef] [Google Scholar]
  47. Harrison, C., Alexander, D., Mullaney, J., & Swinbank, A. 2014, MNRAS, 441, 3306 [CrossRef] [Google Scholar]
  48. Hasinger, G. 2008, A&A, 490, 905 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  49. Heckler, K. F., Ricci, T. V., & Riffel, R. A. 2022, MNRAS, 517, 5959 [Google Scholar]
  50. Hickox, R. C., & Alexander, D. M. 2018, ARA&A, 56, 625 [Google Scholar]
  51. Holt, J., Tadhunter, C., & Morganti, R. 2008, MNRAS, 387, 639 [NASA ADS] [CrossRef] [Google Scholar]
  52. Hopkins, A. M., Miller, C., Nichol, R., et al. 2003, ApJ, 599, 971 [Google Scholar]
  53. Hviding, R. E., Hainline, K. N., Rieke, M., et al. 2022, AJ, 163, 224 [NASA ADS] [CrossRef] [Google Scholar]
  54. Inoue, A. K. 2011, MNRAS, 415, 2920 [NASA ADS] [CrossRef] [Google Scholar]
  55. Jarrett, T., Cohen, M., Masci, F., et al. 2011, ApJ, 735, 112 [NASA ADS] [CrossRef] [Google Scholar]
  56. Johnston, V. D., Medling, A. M., Groves, B., et al. 2023, ApJ, 954, 77 [NASA ADS] [CrossRef] [Google Scholar]
  57. Kakkad, D., Mainieri, V., Vietri, G., et al. 2020, A&A, 642, A147 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  58. Kauffmann, G., Heckman, T. M., Tremonti, C., et al. 2003, MNRAS, 346, 1055 [Google Scholar]
  59. Kennicutt, R. C. 1998, ARA&A, 36, 189 [NASA ADS] [CrossRef] [Google Scholar]
  60. Kennicutt, R. C., & Evans, N. J. 2012, ARA&A, 50, 531 [NASA ADS] [CrossRef] [Google Scholar]
  61. Kewley, L. J., Dopita, M., Sutherland, R., Heisler, C., & Trevena, J. 2001, ApJ, 556, 121 [NASA ADS] [CrossRef] [Google Scholar]
  62. Kewley, L. J., Geller, M. J., & Jansen, R. A. 2004, AJ, 127, 2002 [Google Scholar]
  63. Kewley, L. J., Groves, B., Kauffmann, G., & Heckman, T. 2006, MNRAS, 372, 961 [Google Scholar]
  64. Kollatschny, W., & Zetzl, M. 2013, A&A, 549, A100 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  65. Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511 [Google Scholar]
  66. Koudmani, S., Sijacki, D., & Smith, M. C. 2022, MNRAS, 516, 2112 [NASA ADS] [CrossRef] [Google Scholar]
  67. Krajnovic, D., Cappellari, M., De Zeeuw, P. T., & Copin, Y. 2006, MNRAS, 366, 787 [CrossRef] [Google Scholar]
  68. Leung, G. C., Coil, A. L., Aird, J., et al. 2019, ApJ, 886, 11 [NASA ADS] [CrossRef] [Google Scholar]
  69. Liu, G., Zakamska, N. L., Greene, J. E., Nesvadba, N. P., & Liu, X. 2013, MNRAS, 436, 2576 [NASA ADS] [CrossRef] [Google Scholar]
  70. Liu, W., Veilleux, S., Canalizo, G., et al. 2020, ApJ, 905, 166 [NASA ADS] [CrossRef] [Google Scholar]
  71. Liu, W., Veilleux, S., Canalizo, G., et al. 2024, ApJ, 965, 152 [Google Scholar]
  72. Łokas, E. L., & Mamon, G. A. 2001, MNRAS, 321, 155 [CrossRef] [Google Scholar]
  73. Manzano-King, C. M., Canalizo, G., & Sales, L. V. 2019, ApJ, 884, 54 [NASA ADS] [CrossRef] [Google Scholar]
  74. Maraston, C., Pforr, J., Henriques, B. M., et al. 2013, MNRAS, 435, 2764 [NASA ADS] [CrossRef] [Google Scholar]
  75. Marleau, F. R., Clancy, D., Habas, R., & Bianconi, M. 2017, A&A, 602, A28 [EDP Sciences] [Google Scholar]
  76. Martín-Navarro, I., & Mezcua, M. 2018, ApJ, 855, L20 [NASA ADS] [CrossRef] [Google Scholar]
  77. Mezcua, M., & Domínguez Sánchez, H. D. 2020, ApJ, 898, L30 [Google Scholar]
  78. Mezcua, M., & Domínguez Sánchez, H. D. 2024, MNRAS, 528, 5252 [NASA ADS] [CrossRef] [Google Scholar]
  79. Mezcua, M., Civano, F., Fabbiano, G., Miyaji, T., & Marchesi, S. 2016, ApJ, 817, 20 [Google Scholar]
  80. Mezcua, M., Civano, F., Marchesi, S., et al. 2018, MNRAS, 478, 2576 [NASA ADS] [CrossRef] [Google Scholar]
  81. Mezcua, M., Suh, H., & Civano, F. 2019, MNRAS, 488, 685 [Google Scholar]
  82. Mezcua, M., Siudek, M., Suh, H., et al. 2023, ApJ, 943, L5 [NASA ADS] [CrossRef] [Google Scholar]
  83. Mezcua, M., Pacucci, F., Suh, H., Siudek, M., & Natarajan, P. 2024, ApJ, 966, L30 [NASA ADS] [CrossRef] [Google Scholar]
  84. Moran, E. C., Shahinyan, K., Sugarman, H. R., Vélez, D. O., & Eracleous, M. 2014, AJ, 148, 136 [NASA ADS] [CrossRef] [Google Scholar]
  85. Moster, B. P., Somerville, R. S., Maulbetsch, C., et al. 2010, ApJ, 710, 903 [Google Scholar]
  86. Moster, B. P., Naab, T., & White, S. D. 2013, MNRAS, 428, 3121 [Google Scholar]
  87. Mullaney, J., Alexander, D., Fine, S., et al. 2013, MNRAS, 433, 622 [NASA ADS] [CrossRef] [Google Scholar]
  88. Navarro, J. F., Frenk, C. S., & White, S. D. 1997, ApJ, 490, 493 [NASA ADS] [CrossRef] [Google Scholar]
  89. Osborne, C., Salim, S., Boquien, M., Dickinson, M., & Arnouts, S. 2023, ApJS, 268, 26 [NASA ADS] [CrossRef] [Google Scholar]
  90. Osterbrock, D. E., & Ferland, G. J. 2006, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei, 2nd edn. (Sausalito: University Science Books) [Google Scholar]
  91. Penny, S. J., Masters, K. L., Smethurst, R., et al. 2018, MNRAS, 476, 979 [NASA ADS] [CrossRef] [Google Scholar]
  92. Pucha, R., Juneau, S., Dey, A., et al. 2025, ApJ, 982, 10 [Google Scholar]
  93. Raimundo, S. I., Malkan, M., & Vestergaard, M. 2023, Nat. Astron., 7, 463 [NASA ADS] [CrossRef] [Google Scholar]
  94. Reines, A. E., Greene, J. E., & Geha, M. 2013, ApJ, 775, 116 [Google Scholar]
  95. Reines, A. E., Condon, J. J., Darling, J., & Greene, J. E. 2020, ApJ, 888, 36 [NASA ADS] [CrossRef] [Google Scholar]
  96. Ricci, T., Steiner, J. E., & Menezes, R. 2014, MNRAS, 440, 2419 [NASA ADS] [CrossRef] [Google Scholar]
  97. Ristea, A., Cortese, L., Fraser-McKelvie, A., et al. 2022, MNRAS, 517, 2677 [Google Scholar]
  98. Romano, M., Nanni, A., Donevski, D., et al. 2023, A&A, 677, A44 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  99. Rose, M., Tadhunter, C., Ramos Almeida, C., et al. 2018, MNRAS, 474, 128 [Google Scholar]
  100. Rosen, A. L. 2022, ApJ, 941, 202 [NASA ADS] [CrossRef] [Google Scholar]
  101. Rupke, D. S., Gültekin, K., & Veilleux, S. 2017, ApJ, 850, 40 [NASA ADS] [CrossRef] [Google Scholar]
  102. Salehirad, S., Reines, A. E., & Molina, M. 2022, ApJ, 937, 7 [NASA ADS] [CrossRef] [Google Scholar]
  103. Salehirad, S., Reines, A. E., & Molina, M. 2025, ApJ, 979, 26 [Google Scholar]
  104. Sanders, R. L., Shapley, A. E., Kriek, M., et al. 2015, ApJ, 816, 23 [NASA ADS] [CrossRef] [Google Scholar]
  105. Sarzi, M., Falcón-Barroso, J., Davies, R. L., et al. 2006, MNRAS, 366, 1151 [Google Scholar]
  106. Sau, S., Chattopadhyay, T., & Ray, P. 2023, New Astron., 100, 101992 [Google Scholar]
  107. Shepherd, D. 2007, Protostars and Planets V (Tucson: University of Arizona Press), 245 [Google Scholar]
  108. Siudek, M., Pucha, R., Mezcua, M., et al. 2024, A&A, 691, A308 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  109. Skrutskie, M., Cutri, R., Stiening, R., et al. 2006, AJ, 131, 1163 [Google Scholar]
  110. Smethurst, R. J., Simmons, B. D., Coil, A., et al. 2021, MNRAS, 507, 3985 [NASA ADS] [CrossRef] [Google Scholar]
  111. Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ, 753, 30 [Google Scholar]
  112. Tanner, R., & Weaver, K. A. 2022, AJ, 163, 134 [NASA ADS] [CrossRef] [Google Scholar]
  113. Trebitsch, M., Volonteri, M., Dubois, Y., & Madau, P. 2018, MNRAS, 478, 5607 [NASA ADS] [CrossRef] [Google Scholar]
  114. Vazdekis, A., Koleva, M., Ricciardelli, E., Röck, B., & Falcón-Barroso, J. 2016, MNRAS, 463, 3409 [Google Scholar]
  115. Veilleux, S., Cecil, G., & Bland-Hawthorn, J. 2005, ARA&A, 43, 769 [NASA ADS] [CrossRef] [Google Scholar]
  116. Vietri, G., Garilli, B., Polletta, M., et al. 2022, A&A, 659, A129 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  117. Wang, B., de Graaff, A., Davies, R. L., et al. 2024, ApJ, accepted [arXiv:2403.02304] [Google Scholar]
  118. Westfall, K. B., Cappellari, M., Bershady, M. A., et al. 2019, AJ, 158, 231 [Google Scholar]
  119. Winiarska, M. W., Raimundo, S. I., Davis, T. A., et al. 2025, MNRAS, 538, 1191 [Google Scholar]
  120. Wright, E. L., Eisenhardt, P. R., Mainzer, A. K., et al. 2010, AJ, 140, 1868 [Google Scholar]
  121. Wylezalek, D., Flores, A. M., Zakamska, N. L., Greene, J. E., & Riffel, R. A. 2020, MNRAS, 492, 4680 [NASA ADS] [CrossRef] [Google Scholar]
  122. Zheng, Z., Shi, Y., Bian, F., et al. 2023, MNRAS, 523, 3274 [NASA ADS] [CrossRef] [Google Scholar]
  123. Zhuang, M.-Y., & Ho, L. C. 2023, Nat. Astron., 7, 1376 [NASA ADS] [CrossRef] [Google Scholar]
  124. Zinchenko, I. 2023, A&A, 674, L7 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]

Appendix A: Effects on S/N threshold selection

Mezcua & Domínguez Sánchez (2024) performed a spaxel classification only considering spaxels with S/N≥3 in the BPT emission lines (Hα, Hβ, [NII]λ6583, [SII]λ6718, [SII]λ6732, [OIII]λ5007, and [OI]λ6300). This is something commonly adopted in BPT studies (e.g. Reines et al. 2013 for SDSS; Wylezalek et al. 2020 for MaNGA; Salehirad et al. 2022 for GAMA; and Johnston et al. 2023 for SAMI). However, in our analysis, we consider only spaxels with S/N≥10 to ensure reliable spectral fittings. In Fig. A.1, the spatial distribution of the S/N is shown together with the extension of the outflows of Figs. 2 and 3 but considering a S/N> 3 in the outflow spaxels. No sources with AGN outflows are missed when using S/N≥10 rather than S/N≥3.

thumbnail Fig. A.1.

Spatial distribution of the SNR for the same two examples as in Figs. 2 and 3. The colourbar is set for those spaxels with S/N≥10. Those spaxels with 3≤S/N< 10 are shown in yellow to highlight the number of spaxels that would be missed when applying a threshold of S/N≥10 rather than S/N≥3 to ensure the goodness of the fit. The shaded dark spaxels represent the extension of the AGN outflow in each galaxy.

Appendix B: Outflow, [OIII]λ5007, and stellar velocity maps

Figures B.1-B.6 show the outflow, [OIII]λ5007 and stellar velocity maps of the remaining 11 dwarf galaxies with AGN outflow signatures, as in Figs. 2 and 3.

thumbnail Fig. B.1.

Same as Fig. 2, but for the dwarf galaxy 10226-1902.

thumbnail Fig. B.2.

Same as Fig. 2, but for the dwarf galaxies 11022-12702 and 11754-3701.

thumbnail Fig. B.3.

Same as Fig. 2, but for the dwarf galaxies 11826-12702 and 12487-3704.

thumbnail Fig. B.4.

Same as Fig. 2, but for the dwarf galaxies 7992-6102 and 8252-1902.

thumbnail Fig. B.5.

Same as Fig. 2, but for the dwarf galaxies 8442-1901 and 8657-6104.

thumbnail Fig. B.6.

Same as Fig. 2, but for the dwarf galaxies 8982-3703 and 9889-1902.

All Tables

Table 1.

Position angles.

In the text
Table 2.

Properties of the outflows for the MaNGA sample of 13 dwarf galaxies.

In the text
Table 3.

AGN luminosity for the MaNGA sample of 13 dwarf galaxies.

In the text
Table 4.

Regression lines of the W80 vs L[OIII] correlation.

In the text

All Figures

thumbnail Fig. 1.

Top left and right: Spectra of the galaxies 8982−3703 and 12487−3704, respectively. The dark line is the stacked spectrum of the AGN spaxels. The green line is the best fit of the stellar continuum combining SSPs. The blue line is an additional power-law component, the yellow line represents the fitted emission lines, and the red line is the sum of all these components that compose the total model. The residuals after subtraction of the total model from the data are shown at the bottom. Bottom-left and right: Zoom-in on the [OIII]λ5007 Å doublet region of the stacked spectrum. The [OIII] line is fitted by a narrow (blue) and a broad (green) component.
In the text
thumbnail Fig. 2.

MaNGA analysis for one of the dwarf galaxies with AGN outflow candidates (8655−6101). Top left: Spatial distribution of the BPT-classified spaxels (red: AGN; green: composite; blue: SF; orange: LINER). The ‘N’ shows the number of AGN and composite spaxels used in the analysis. Top right: SDSS composite image. The pink hexagon shows the IFU coverage. Middle: Spatial W80 (left) and vout (right) outflow velocity distribution, where we can see the extension of the outflow. Empty squares mark the IFU coverage and grey squares those spaxels with a continuum S/N > 1. Bottom: MaNGA [OIII]λ5007 Å velocity map (left) and stellar velocity (right).
In the text
thumbnail Fig. 3.

Same as Fig. 2 but for the dwarf galaxy 10223−3702.
In the text
thumbnail Fig. 4.

Comparison of the ratios of the two velocity indicators W80 and vout with vesc. The grey dashed lines represent the 1:1 relation. Three out of the 13 MaNGA dwarf galaxies with AGN outflow candidates would have outflows fast enough to become unbound from the galaxy and from its dark matter halo.
In the text
thumbnail Fig. 5.

Distribution of the W80 velocity for the dwarf galaxies with AGN outflow candidates with AGN fraction ≤0.45 (blue), and with AGN fraction > 0.45 (orange).
In the text
thumbnail Fig. 6.

W80 velocity vs L[OIII] for the MaNGA sample of 13 dwarf galaxies with AGN outflow candidates (red squares for AGN galaxies, red diamonds for SF-AGN galaxies), the AGN dwarf galaxy sample of Liu et al. (2020) (red triangles), the z < 2 type-2 AGN quasars of Harrison et al. (2014) (grey pluses), the radio-quiet quasars of Liu et al. (2013) (grey crosses), the z ∼ 2 type-1 massive AGN of Kakkad et al. (2020) (grey diamonds), the massive LLAGN of Heckler et al. (2022) (grey circle) and the MaNGA sample of low-intermediate luminosity massive galaxies of Wylezalek et al. (2020) (grey squares). The solid lines are regression fits to the massive sample (grey), the dwarf sample of AGN outflows, which includes the nine MaNGA dwarf galaxies with AGN outflows and those from Liu et al. (2020) (red), and the combination of all the massive galaxies, along with the dwarf sample of AGN outflows (purple) indicated as the shaded area the 1σ error of the regression line.
In the text
thumbnail Fig. 7.

Ionised outflow kinetic energy rate vs AGN bolometric luminosity (left) and ratio of the kinetic energy outflow rate to the AGN bolometric luminosity (right) for the MaNGA sample of dwarf galaxies with AGN outflow candidates (red squares for AGN galaxies, red diamonds for SF-AGN galaxies), the AGN dwarf galaxy sample of Liu et al. (2020) (red triangles), radio-quiet quasars of Liu et al. (2013) (grey crosses), type 1 quasars of Rupke et al. (2017) (grey diamonds), the ULIRGs of Rose et al. (2018) (grey pluses), MaNGA sample of low-intermediate luminosity massive galaxies of Wylezalek et al. (2020) (grey squares), and the massive LLAGN of Heckler et al. (2022) (grey circle). The purple solid line is a regression fit considering all the massive galaxies, the nine MaNGA dwarf galaxies with AGN outflows, and those from Liu et al. (2020). The shaded area corresponds to the 1σ error of the regression line.
In the text
thumbnail Fig. A.1.

Spatial distribution of the SNR for the same two examples as in Figs. 2 and 3. The colourbar is set for those spaxels with S/N≥10. Those spaxels with 3≤S/N< 10 are shown in yellow to highlight the number of spaxels that would be missed when applying a threshold of S/N≥10 rather than S/N≥3 to ensure the goodness of the fit. The shaded dark spaxels represent the extension of the AGN outflow in each galaxy.
In the text
thumbnail Fig. B.1.

Same as Fig. 2, but for the dwarf galaxy 10226-1902.
In the text
thumbnail Fig. B.2.

Same as Fig. 2, but for the dwarf galaxies 11022-12702 and 11754-3701.
In the text
thumbnail Fig. B.3.

Same as Fig. 2, but for the dwarf galaxies 11826-12702 and 12487-3704.
In the text
thumbnail Fig. B.4.

Same as Fig. 2, but for the dwarf galaxies 7992-6102 and 8252-1902.
In the text
thumbnail Fig. B.5.

Same as Fig. 2, but for the dwarf galaxies 8442-1901 and 8657-6104.
In the text
thumbnail Fig. B.6.

Same as Fig. 2, but for the dwarf galaxies 8982-3703 and 9889-1902.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.