Outflows, Shocks and Coronal Line Emission in a Radio-Selected AGN in a Dwarf Galaxy
M. Molina, A. E. Reines, J. E. Greene, J. Darling, J. J. Condon
DDraft version February 1, 2021
Typeset using L A TEX twocolumn style in AASTeX63
Outflows, Shocks and Coronal Line Emission in a Radio-Selected AGN in a Dwarf Galaxy
Mallory Molina, Amy E. Reines, Jenny E. Greene, Jeremy Darling, and James J. Condon eXtreme Gravity Institute, Department of Physics, Montana State University, Bozeman, MT 59717, USA Department of Astrophysical Sciences, Princeton University, Princeton, NJ 08544, USA Center for Astrophysics and Space Astronomy, Department of Astrophysical and Planetary Sciences, University of Colorado, 389 UCB,Boulder, CO 80309-0389, USA National Radio Astronomy Observatory, Charlottesville, VA, 22903, USA (Received December 10, 2020; Revised January 20, 2020; Accepted January 27, 2020)
ABSTRACTMassive black holes (BHs) in dwarf galaxies can provide strong constraints on BH seeds, howeverreliably detecting them is notoriously difficult. High resolution radio observations were recently usedto identify accreting massive BHs in nearby dwarf galaxies, with a significant fraction found to be non-nuclear. Here we present the first results of our optical follow-up of these radio-selected active galacticnuclei (AGNs) in dwarf galaxies using integral field unit (IFU) data from Gemini-North. We focus onthe dwarf galaxy J1220+3020, which shows no clear optical AGN signatures in its nuclear SDSS spec-trum covering the radio source. With our new IFU data, we confirm the presence of an active BH viathe AGN coronal line [Fe X ] and enhanced [O I ] emission coincident with the radio source. Furthermore,we detect broad H α emission and estimate a BH mass of M BH = 10 . M (cid:12) . We compare the narrowemission line ratios to standard BPT diagnostics and shock models. Spatially-resolved BPT diagramsshow some AGN signatures, particularly in [O I ]/H α , but overall do not unambiguously identify theAGN. A comparison of our data to shock models clearly indicates shocked emission surrounding theAGN. The physical model most consistent with the data is an active BH with a radiatively inefficientaccretion flow (RIAF) that both photoionizes and shock-excites the surrounding gas. We concludethat feedback is important in radio-selected BHs in dwarf galaxies, and that radio surveys may probea population of low accretion-rate BHs in dwarf galaxies that cannot be detected through opticalsurveys alone. Keywords: active galaxies – dwarf galaxies – low-luminosity active galactic nuclei – radio jets – blackholes INTRODUCTIONSupermassive black holes (BHs) are known to be ubiq-uitous in the nuclei of massive galaxies (e.g., Kormendy& Richstone 1995; Kormendy & Ho 2013), but theirinitial formation conditions have been lost to merger-driven growth over cosmic time (Volonteri 2010; Natara-jan 2014). The proposed theories for the formation ofthe initial BH “seeds in the early Universe include rem-nants from Population III stars, which would create BHswith M BH ∼ M (cid:12) (Bromm & Yoshida 2011). Al-ternatively, the direct collapse of gas (Loeb & Rasio Corresponding author: Mallory [email protected] M BH ∼ –10 M (cid:12) .While the first BH “seeds” at high redshift are toosmall and faint to currently detect, nearby dwarf galax-ies can place strong constraints on BH seed masses (seeReines & Comastri 2016; Greene et al. 2020, and refer-ences therein). Dwarf galaxies are known to have rela-tively quiet merger histories compared to more massivegalaxies (Bellovary et al. 2011), and recent work hasshown that dwarf galaxies may experience stunted BHgrowth due to supernova feedback (Angl´es-Alc´azar et al. a r X i v : . [ a s t r o - ph . GA ] J a n M BH (cid:46) ) is expected to berelatively close to the initial seed mass.Unfortunately detecting BHs in dwarf galaxies is dif-ficult, due to their smaller masses and lower luminosi-ties. In principle, there are numerous ways to identifyaccreting BHs, in many different wavelength regimes(see Ho 2008, for a complete review). In the opticalregime, broad H α emission and/or narrow-line ratiosconsistent with AGN photoionization have been used toidentify low-mass AGNs in dwarf galaxies (e.g., Reineset al. 2013). This work often relies on the Baldwin,Phillips and Terlevich diagrams (BPT diagrams; Bald-win et al. 1981), which differentiate between the harderspectral energy distributions (SEDs) created by AGNsand those of star forming regions. However, these dia-grams are not as effective at identifying low-luminosityAGNs (LLAGNs), and low ionization nuclear emissionregions (LINERs; see Ho 2008; Kewley et al. 2019, fora review). LINERs with LLAGNs in massive galaxieswill have a relatively larger contribution from shocksthan higher luminosity AGNs, thus creating a combina-tion of power sources that contribute to the observedemission (Molina et al. 2018). Similarly, radiatively in-efficient accretion flows (RIAFs), which require a lowaccretion rate, have a different SED that could removeany observable broad-line emission (Trump et al. 2011).Furthermore, optical AGN indicators may be diluted bythe host galaxy in these low-metallicity systems, evenwithout a significant amount of ongoing star formation(Moran et al. 2002; Groves et al. 2006; Stasi´nska et al.2006; Cann et al. 2019). Given the low luminosity na-ture of the BHs in dwarf galaxies, any combination ofthese scenarios can easily hide AGN activity.Infrared (IR) data, particularly the Wide-field InfraredSurvey Explorer (WISE; Wright et al. 2010) colors arealso used to identify AGNs in more massive galaxies(Jarrett et al. 2011; Mateos et al. 2012; Stern et al.2012), but can be confused with star formation in dwarfgalaxies (Hainline et al. 2016; Condon et al. 2019, La-timer et al. in prep). Similarly, almost all AGNs, evenLLAGNs, produce radio emission (see Ho 2008, and ref-erences therein), which is not affected by dust attenu-ation. However, the radio emission from BHs in dwarfgalaxies could be similar with that from H II regions orsupernova remnants or supernovae (e.g., Reines et al.2008; Chomiuk & Wilcots 2009; Johnson et al. 2009;Aversa et al. 2011; Kepley et al. 2014; Varenius et al.2019). Therefore, caution must be used when identi-fying AGNs in dwarf galaxies using either radio or IRdata. Reines et al. (2020) conducted a radio survey usingthe NSF Karl G. Jansky Very Large Array (VLA), andidentified 13 dwarf galaxies with radio emission indicat-ing AGNs, some of which were outside of the nucleus ofthe galaxy. For all 13 AGN candidates, they proved theobserved radio emission was too luminous and compactto be explained by stellar processes, such as thermal H II regions, and individual or populations of supernovae andsupernova remnants (see Section 5 of Reines et al. 2020).While Reines et al. (2020) found that the AGN candi-dates showed enhanced [O I ]/H α emission in their SloanDigital Sky Survey spectra (SDSS Blanton et al. 2017),most of them did not have clear optical signatures ofAGN photoionization. Given the SDSS spectra have a3 (cid:48)(cid:48) diameter, they have a significant contribution fromthe host galaxy which could contaminate the AGN sig-nal. Therefore, higher-resolution optical data is crucialto confirm these BH candidates.In this work, we present the first of the optical inte-gral field unit (IFU) follow-up of the Reines et al. (2020)sample taken with the Gemini Multi-Object Spectro-graph on Gemini-North (GMOS-N). Our main goals areto confirm the presence of an AGN and study its impacton the host galaxy. In this paper, we present the resultsfor SDSS J122011.26+302008.1 (ID 82 in Reines et al.2020), hereafter referred to as J1220+3020. The DarkEnergy Camera Legacy Survey (DECaLS; Dey et al.2019) optical image and the VLA radio emission forthis galaxy are shown in Figure 1, and the galaxy prop-erties are listed in Table 1. The Reines et al. (2020)VLA observation for J1220+3020 was consistent witha point source at a resolution of ∼ . (cid:48)(cid:48)
18, which corre-sponds to ≈
100 pc. The radio source in J1220+3020is in the nucleus of the galaxy, resides in one of thebrighter objects in the sample, and was observed for thefull time requested . We therefore will use J1220+3020as a case study to test the methodology of our anal-ysis, and compare the properties of radio-selected andoptically-selected AGN.We will make full use of our IFU data by first study-ing 2–D emission-line maps, especially [O I ]. We willidentify any strong morphological features of the gasand discuss their physical implications. After that, wewill create resolved 1–D spectra for the radio source andsurrounding regions, and compare them to optical diag-nostics to determine their dominant excitation mecha-nism. Given the previously described issues with stan-dard BPT diagrams, we will also explore shock models Due to the Covid-19 Pandemic, our observing program was notcompleted.
H Activity and Outflows in a Dwarf Galaxy Table 1.
Properties of J1220+3020 a Property Value UnitsGalaxy R.A. b b ∗ /M (cid:12) ) 9.4 ... M g c − . ± .
02 mag g − r c . ± .
01 mag r d S . e . ± .
15 mJyRadio Source R.A. f f f S ± µ Jylog(L ) 20 . ± .
03 W Hz − α . − . ± . g ... α . − . ± . g ...Point Source? h True ... a All reported values are from Reines et al. (2020), and assume h = 0 . b The coordinates for the galactic nucleus as defined by the NASA-Sloan Atlas (NSA; Blanton et al. 2011). c Magnitudes are both K-corrected and corrected for foregroundGalactic extinction. dr is the Petrosian 50% radius. e The 1.4 GHz detection is from the VLA Faint Images of the Ra-dio Sky at Twenty centimeters (FIRST) Survey, with errors cal-culated using the procedure described in Condon (1997). f The position of the 9 GHz radio emission as detected by Reineset al. (2020). The “Radio Source Offset” is the difference be-tween the NSA galactic nucleus coordinates and the radio sourcecoordinates. g The spectral indices assume S ν ∝ ν α . We also present the newcalculation of α . , but note that the two observations were takenyears apart, and that the 1.4 GHz flux measurement could havecontamination from star formation in the host galaxy. h The radio emission in J1220+3020 is consistent with a pointsource, where the fitted 2–D Gaussian model has a major axisfull-width at half-maximum (FWHM) of 0 . (cid:48)(cid:48)
18 and a minor axisFWHM of 0 . (cid:48)(cid:48) from Allen et al. (2008), which have been successfullyused to explain LINER-like emission in more massivegalaxies (Molina et al. 2018).In Section 2 we describe the observations and data re-duction process. The 2–D spatial emission is presented Figure 1.
Top:
DECaLS grz -band images of dwarf galaxyJ1220+3020. The red cross indicates the position of the radiosource, the white dashed circle indicates the position of theSDSS fiber, and the blue box is the 2 . (cid:48)(cid:48) × . (cid:48)(cid:48) Bottom: (cid:48)(cid:48) white scale bar above the radio source for reference.The emission is consistent with a point source of size 0 . (cid:48)(cid:48) × . (cid:48)(cid:48)
17. The properties of J1220+3020 are listed in Table 1. in Section 3, and the resolved 1–D emission, includingemission-line fitting, is discussed in Section 4. We com-pare our data to BPT diagrams and shock models inSection 5. We consider these results and discuss poten-tial physical models for the system in Section 6 and sum-marize our findings and conclusions in Section 7. In thispaper, we assume a ΛCDM cosmology with Ω m = 0 . Λ = 0 . = 70 km s − Mpc − . OBSERVATIONS AND DATA REDUCTION2.1.
Description of Observations
The GMOS-N/IFU data for J1220+3020 were takenbetween February and July 2020. The GMOS-N/IFUuses a hexagonal lenslet array with projected diametersof 0 . (cid:48)(cid:48) (cid:48)(cid:48) × (cid:48)(cid:48) field of view (FOV).Each lenslet is coupled to a fiber, which redirects thelight to GMOS. The IFU can be used in either one-slitor two-slit mode. The one-slit mode will record roughlyhalf the number of fibers as the two-slit mode, creatinga 2 . (cid:48)(cid:48) × . (cid:48)(cid:48) β to [S II ] λλ , R ∼ ◦ based on telescope observing constraints and theposition of the radio source in the galaxy.The CCDs on GMOS-N have two ∼ . (cid:48)(cid:48) ∼
100 ˚A gaps in the IFUspectra. We therefore took four 900 s observations,with two at each central wavelength setting of 580 nmand 585 nm. These two settings were chosen to cap-ture all the lines of interest, avoid any emission linesfalling within the chip gaps, and to fill out the contin-uum within the chip gaps.Given the lack of point sources within the FOV of thescience frames, we cannot assess the seeing on a frame-to-frame basis. However, all four frames used in thedata reduction meet the following criteria: 85% imagequality, 50% cloud cover, 80% background and ‘any’ wa-ter vapour. In practice, this means the seeing is likelywithin 1 (cid:48)(cid:48) .In addition to the science frames, two 300 s standardstar observations of Hz 44 were taken within the sameobservation window, using two different central wave-length settings: 580 nm and 600 nm. These two settingscorrespond to the lowest and highest central wavelengthsettings for all 10 objects in our Gemini observing pro-gram. We then use these two standard star observationsto create a master sensitivity function that can be ap-plied to all of our science observations. These framesare processed in a manner similar to the science data asdescribed in Section 2.2.2.2.
Basic Data Reduction
The data were reduced according to the GMOS IFU-1data reduction tutorial (Gemini Observatory & AURA2016; Rupke 2014), which uses commands from the gemini package in pyraf (Green 2012). We give a briefdescription of the process below.Before reducing the science frames, we reduce thestandard star observations to create the sensitivity func-tions. We first create a master bias for the standard https://gmos-ifu-1-data-reduction-tutorial-gemini-iraf.readthedocs.io/en/latest/ star. We then correct one of two flats for bias and over-scan, and use it to verify the mask definition file (MDF),which maps the fibers in the observation to their correctbundle, and masks out dead fibers. After defining theMDF, we trace the light illumination across the detectorusing the same flat.We use that reduced flat to extract the arc and cal-culate the wavelength solution, which is used to fullyreduce the flat. This process includes modelling andsubtracting scattered light between the bundles, correct-ing for quantum efficiency and calculating the responsefunction.After correcting the flat, we begin processing the stan-dard star observation itself. We remove the bias andoverscan, subtract scattered light and attach the MDF.The standard frames are then corrected for quantum ef-ficiency, rectified to enforce the same wavelength scale,and then sky-subtracted. After this is completed, thereduced standard star frame is used to calculate the sen-sitivity function. This process is repeated for both stan-dard star observations. The two individual sensitivityfunctions are combined to make the master sensitivityfunction, which can be applied to all science frames.Finally, the science frame goes through the same pro-cessing as the standard star, with the addition of cos-mic ray rejection using L.A.-cosmic (van Dokkum 2001).We then use the sensitivity function calculated from thestandard star observation to calibrate our science data.2.3. Combining Exposures into Final Data Cube
The basic data reduction process is completed for eachindividual exposure, creating 4 data cubes. Each datacube has a plate scale of 0 . (cid:48)(cid:48) − in the two spa-tial directions, and ∼ . − in the spectral direc-tion. Due to the use of two different wavelength settingsas discussed in Section 2.1, the two pairs of observationsdiffer in their wavelength ranges by ∼
50 ˚A. In orderto combine the exposures into the final data cube, werebin all four individual cubes, using a flux-conservingmethod, to have the same wavelength scale and range.The final data cube has an observed wavelength rangeof 4133–7232 ˚A, with 0.5 ˚A pixel − . By matching thewavelength range, we truncate the blue and red endsof the spectra with central wavelengths of 580 nm and585 nm, respectively. However, this step does not im-pact any emission lines of interest. H Activity and Outflows in a Dwarf Galaxy
IRAF task imcombine is then used to mediancombine the aligned images (National Optical Astron-omy Observatories 1999). The error bar for each pixelin the final IFU data cube is the standard deviation ofthe four values of that pixel, one from each exposure.2.4. Calculating WCS and Astrometry Correction
The world coordinate system (WCS) for a GMOS IFUdata cube is presented as physical coordinates relativeto the central pixel. As precise astrometry is vital toproperly identify the optical emission associated withthe radio source, we performed an astrometry correctionon our GMOS observations in a two step process.In the first step, we convert the WCS from the phys-ical system provided by the data reduction pipelineto an FK5 coordinate system using the information inquestion 3 of the data reduction section of the GMOSIFU question page . The process is relatively straight-forward given that our PA is 90 ◦ . The only deviation inour calculation from the process described on the GMOSIFU question page is that the GMOS detector plate scalehas been updated from 0 . (cid:48)(cid:48) − to 0 . (cid:48)(cid:48) − tomatch our observations.After constructing the initial WCS, we confirm theastrometry by comparing the 2–D image of the GMOSobservation to the r -band image of the galaxy fromSDSS. The r -band image is chosen to match the regionof the spectrum where the GMOS signal is strongest,i.e. λ (cid:38) r -band photometry of SDSSare aligned to perform the final astrometry correction.Given the larger pixel size of SDSS, the astrometry iscorrect to within 0 . (cid:48)(cid:48) α and [O I ] λ σ x = σ y = 1 spatial pixel.While [O III ] λ IRAF is distributed by the National Optical Astronomy Observa-tories, which are operated by the Association of Universities forResearch in Astronomy, Inc., under cooperative agreement withthe National Science Foundation. http://ast.noao.edu/sites/default/files/GMOS FAQ/GMOSIFU.html the measurements of [O III ] using one spatial pixel unre-liable. Additionally, we do not include velocity disper-sion maps as most of the line widths are comparable tothe instrumental resolution. Finally, we do not correctfor reddening because we are interested in qualitativeglobal trends. Since H α and [O I ] are relatively close toeach other, any changes due to reddening will be ap-proximately the same for both lines.We calculated these maps by fitting a combination ofGaussians to both the H α and [O I ] λ α or 2 for [O I ], then the line is con-sidered detected and a second fit using two Gaussian isperformed. We accept the 2-Gaussian fit if the reduced χ is lower by at least 20%. For the pixels with spec-tra that displayed blending between the [N II ] doubletand H α , all three lines were fit simultaneously. In somecases, a third Gaussian was needed to fully describe theemission from H α . When more than one Gaussian wasneeded, the centroid is defined as the peak of the ob-served emission line.The results of this process are shown in Figure 2, withthe position of the radio source shown as the black cross.The velocity maps are calculated with respect to thecentroid of the emission line at the position of the radiosource. We note a small mismatch in the fibers, whichcan be clearly seen in the H α α flux along with a knot of blue-shifted H α gas. However,its velocity offset is approximately equal to the error inthe velocity maps, ( σ v ≈
10 km s − ). Additionally, theradio source appears to be in the center of an elongatedfeature of enhanced [O I ] emission, which could indicatean outflow. Unlike H α we do not see any trends in thevelocity of the [O I ] gas. However, we do see broadened[O I ] emission associated with the elongated, enhancedfeature, as discussed in Section 4.3.2.We note that while these 2–D maps are useful for ex-amining the spatial extent of emission lines within thegalaxies, the individual pixels are not resolved. There-fore, we avoid creating 2–D emission line maps of differ-ent components in these emission lines as they may notbe physically meaningful. ◦ ◦ . . . . . Right Ascension D e c li n a t i o n H α − f λ ( e r g/ s / c m / ˚A ) ◦ ◦ . . . . . Right Ascension D e c li n a t i o n [O I] 0 . . . . . . − f λ ( e r g/ s / c m / ˚A ) ◦ ◦ . . . . . Right Ascension D e c li n a t i o n H α − − V e l o c i t y ( k m / s ) ◦ ◦ . . . . . Right Ascension D e c li n a t i o n [O I] − − V e l o c i t y ( k m / s ) Figure 2.
Top: α (left) and [O I ] λ − f λ , with the color bar on the right of the image. A black pixel represents spectra where the line was not detected. Theshifted rows are due to a small mismatch of fibers in the data reduction, but this effect does not significantly impact our 1-Dresolved measurements. The flux was calculated via fitting a combination of Gaussians, as described in Section 3. We showthe shape of the point spread function via the black aperture in the H α flux panel. The black cross in each panel marks theposition of the radio source. The H α emission peaks near the radio source, and there is an elongated feature of enhanced [O I ]emission centralized on the radio source. This [O I ] feature could be indicative of an outflow. Bottom:
Same as Top except forthe velocity. The velocity is measured in km s − , and presented relative to that at the radio source. We have smoothed thevelocity maps with a 2–D Gaussian, where σ x = σ y = 1 spatial pixel. The radio source appears to be the origin for a blue-shiftedknot of H α gas, but this offset is consistent with the measured errors in velocity ( σ v ≈
10 km s − ). Meanwhile there is no cleartrend in the [O I ] emission.4. RESOLVED 1–D SPECTRA4.1.
Region definitions
In order to accurately measure the emission from theradio source, we must spatially bin the data to createresolved 1–D spectra. We binned the data using twodifferent methods: the first focuses on the integratedemission, while the second explores spatially resolvedregions. The positions of these regions within the IFUFOV in are shown Figure 3, and the aperture definitionsare provided in Table 2.The first method, shown in the left panel, consists ofconcentric, cumulative apertures. Aperture A isolatesthe region of enhanced [O I ] emission seen in Figure 2.We used the matplotlib contour command (Hunter 2007)to quantify the extent of the enhanced [O I ] emission,shown as the magenta contour, which is includes all pix-els with flux f i where f i ≥ . f max . We confirmed byeye that the contour encompassed the enhanced [O I ]emission feature seen in Figure 2. We then define Aper-ture A to match the general shape of the [O I ] feature, and use the major axis of that aperture to define themajor axis of the [O I ] emission.Meanwhile, apertures B and C are cumulative circu-lar apertures of increasing radius centered on the ra-dio source. The largest aperture shown is the measuredSDSS DR8 spectrum. Each aperture in method 1 in-cludes all of the light within the defined area. If there isan AGN at the position of the radio source, these aper-tures will illustrate whether its light can be hidden byincreasing the contamination from the host galaxy.The second method is shown on the right panel, andinvolves resolved regions around the radio source. Aper-ture 7 is the same as aperture B in method 1. We usedthe major axis of aperture A from method 1 to defineapertures 1 and 4 in method 2. In other words, apertures1 and 4 measure the emission that is directly outside ofthe radio source aperture, and in the direction of the po-tential [O I ] outflow. We then filled in the gaps betweenregions 1 and 4 using apertures that were spatially re-solved and outside the radio source aperture. H Activity and Outflows in a Dwarf Galaxy ◦ ◦ . . . . Right Ascension D e c li n a t i o n A B CSDSSMethod 1 0 . . . . . . . − f λ , c o ll a p s e d ( e r g/ s / c m / ˚A ) ◦ ◦ . . . . . Right Ascension D e c li n a t i o n Method 2 7 2 3456 1 0 . . . . . . . − f λ , c o ll a p s e d ( e r g/ s / c m / ˚A ) Figure 3.
The collapsed IFU image of J1220+3020, showing the different 1–D regions used in the analysis of this work. Theblack cross is the radio source, and the magenta dot-dash contour shows the position of the enhanced [O I ] emission within thegalaxy, as seen in Figure 2. The definitions of this contour and its major axis are given in Section 4.1. All the regions aredesignated by a letter or a number, and will be referenced in the analysis in Section 6. Left:
First aperture definition method,with 3 total apertures that all overlap. The first is the elliptical aperture around the enhanced [O I ] feature. Then there areapertures with 1 (cid:48)(cid:48) and 2 (cid:48)(cid:48) diameters centered on the radio source. The aperture of the SDSS spectrum is shown as a red dashedline for reference. Right:
Second aperture definition method. Aperture 7 is identical to aperture B in the first method, and isa 1 (cid:48)(cid:48) aperture around the radio source. Apertures 1 and 4 follow the direction of the major axis of the enhanced [O I ] feature,while the remaining apertures fill in the space between these two regions. We note that the resolution of the GMOS IFU datais 1 (cid:48)(cid:48) , while the radio source is a point source of size0 . (cid:48)(cid:48) × . (cid:48)(cid:48)
17. Therefore, all apertures except for apertureA in method 1 are resolved, and cover a much largerphysical area than the radio source. We show the re-solved 1 (cid:48)(cid:48) spectrum of the BH candidate (aperture B inmethod 1 and aperture 7 in method 2) in Figure 4.4.2.
Continuum subtraction and emission-linemeasurements of 1–D spectra
After extracting the 1–D spectra, we subtract the con-tinuum and measure the emission lines; the narrow lineflux measurements used in our analysis are reported inTable 3. As there is no clear detection of the stellarcontinuum, we fit a third-order polynomial to the con-tinuum for all spectra. We then fit groups of lines in-dividually with the code with pyspeckit (Ginsburg &Mirocha 2011), which is a wrapper around the pythonpackage MPFIT, and is designed for fitting astronomi-cal spectra. We employ a fitting process similar to thatof Reines et al. (2013). We show the fits to all of thestrong lines and a table of the line component widthsfor aperture B, i.e., the spectrum of the radio source, inAppendix A.We begin by fitting the [S II ] λλ II ] dou-blet is then fit with a two-component Gaussian model,where the width, relative position, and height ratio for the two-component model constrained to be the samefor each line. We select the two-component model if thereduced χ value is at least 10% lower than that of thesingle Gaussian model. We found that the only spectrathat required a two-component model for the [S II ] dou-blet were those that included emission from the radiosource. This includes the unresolved aperture A and re-solved apertures B and C in method 1. Examples of oneand two-component Gaussian fits for [S II ] are shown inFigure 5.After choosing the appropriate the [S II ] model, we useit to constrain the [N II ]+H α complex. We again employa linear fit to the continuum and use the [S II ] model asa template for the narrow line emission in both the [N II ]doublet and H α . For the [N II ] doublet, we require theemission lines to have the same width as the [S II ] modelin velocity space, constrain the separation as defined bytheir laboratory wavelengths, and fix the relative fluxas [N II ] λ / [N II ] λ α is a bit harder toconstrain via [S II ] as it is a recombination line, not aforbidden line. If [S II ] is best-fit by a single Gaussianmodel, we use it to constrain the narrow-line emissionof H α , but allow the width of H α to increase by as muchas 25%. If [S II ] is best-fit by a two-component Gaussianmodel, we constrain the H α narrow emission to have thesame profile and width (in velocity space) as [S II ]. Insome cases, another broader component was needed tofit the H α emission line. We accepted the fit with thebroad component if the reduced χ was smaller by atleast 20%. Since H α is a recombination line, we did notassume this third component was necessarily a “broad”component from the broad line region of an AGN. A Table 2.
Aperture Definitions for 1–D Spectra
Method a Number R.A. Dec. a b b θ Area Description(deg) (deg) (arcsec) (arcsec) (deg) (kpc )1 A 185.04695 30.33563 0.35 0.18 90 0.059 Captures Enhanced [O I ] emission1 B 185.04694 30.33564 0.50 ... ... 0.234 1 (cid:48)(cid:48) Aperture centered on Radio Source1 C 185.04694 30.33564 1.00 ... ... 0.936 2 (cid:48)(cid:48)
Aperture centered on Radio Source2 1 185.04716 30.33584 0.50 ... ... 0.234 Follows semi-major axis of enhanced [O I ] emission2 2 185.04686 30.33590 0.46 ... ... 0.198 Off-set from radio source in NNW direction2 3 185.04663 30.33573 0.50 ... ... 0.234 Off-set from radio source in NW direction2 4 185.04670 30.33545 0.50 ... ... 0.234 Follows semi-major axis of enhanced [O I ] emission2 5 185.04702 30.33540 0.52 0.38 0 0.253 Off-set from radio source in SSE direction2 6 185.04726 30.33557 0.50 ... ... 0.234 Off-set from radio source in SE direction2 7 c (cid:48)(cid:48) Aperture around Radio Source a The apertures in Method 1 are cumulative, and have total areas of 0.197 arcsec ,0.78 arcsec , and 3.14 arcsec , respectively. b If The aperture is circular, “a” represents the radius. If it is an ellipse, then “a” is the semi-major axis. c Aperture 7 in method 2 is identical to aperture B in method 1. f ( e r g c m s ˚A a r c s e c ) [N II]+H ↵ [S II][O I]H [O III]H He I [Ar III]1-D spectrum of Radio Source f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I] Figure 4.
The resolved continuum-subtracted, 1 (cid:48)(cid:48) spectrum of the radio source, or aperture B in method 1 and aperture 7 inmethod 2, with the strong lines labeled. The spectrum is a combination of three different CCD chips, with the chip gaps at ∼ ∼ I ] λ H Activity and Outflows in a Dwarf Galaxy f ( e r g / s / c m / ˚ A / a r c s e c ) Method 1, Aperture C f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I] f ( e r g / s / c m / ˚ A / a r c s e c ) Method 2, Aperture 5Method 2, Aperture 5 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I] Figure 5.
Top:
Example of two-component fit to the [S II ]lines, using the aperture C from method 1 (2 (cid:48)(cid:48) aperture). Thedata are in black, and the residuals are shown below the fit.The red line is the final fit to the emission lines, while theblue dashed lines show the individual components and thebaseline fit. This spectrum includes emission from the radiosource. Bottom:
Same as the top, but for a one-componentfit to the [S II ] lines, using the fifth aperture from method 2.This spectrum is not coincident with the radio source. more complete discussion of this issue is presented inSection 4.3.We then apply the same process used for the [N II ]doublet to the [O I ] doublet, with the assumed flux ratioof [O I ] λ / [O I ] λ I ] emission lines. Additionally, the [Fe X ] λ β and [O III ] doublets separatelyfrom the rest of the emission lines. These lines fallin the noisier part of the GMOS spectra. We there- fore fit them independently, using no constrains on H β .The [O III ] λλ III ] λ / [O III ] λ β wasnot detected. Due to the higher level of noise on theblue end of the spectrum, we accept fits to H β down toa S / N = 2. Below that threshold, we calculate the 3- σ upper limit, using the [O III ] λ Features of the Line-Emitting Gas
Broad Hydrogen Balmer Emission
Given that H α is a recombination line, it is typicallyemitted physically closer to the power source than [S II ].As a result, H α will potentially have a broader profile,and will likely have a different emission-line shape thanthe forbidden narrow lines. As we are interested in usingBPT diagrams in our analysis, which relies on the non-broad line region emission, we must be cautious whendifferentiating between the broad and narrow compo-nents. Reines et al. (2013) accounted for this issue byallowing the width of H α to increase by as much as25% compared to that of the [S II ] doublet, but giventhe higher spectral resolution of our data, we alloweda broader component to be included if needed. We ac-cepted that component if the reduced χ was smaller byat least 20%.The only two spectra in method 1 that required anextra component for H α were apertures A (unresolvedaperture around enhanced [O I ] feature) and B (resolved,1 (cid:48)(cid:48) aperture of radio source). In both cases, the full widthat half maximum, or FWHM (cid:38)
900 km s − which areclearly broad-line components. We also measure the H α flux in a 1 (cid:48)(cid:48) annulus directly outside the radio source,and do not detect this broad component. We thereforeconclude that this broad emission is associated with theradio source. The fit for aperture B is shown in the toppanel of Figure 6. We note that the H α broad compo-nent in aperture B (which is the same as aperture 7) isonly 5% of the total observed H α emission, and there-fore removing this feature will not significantly affect theemission line ratios used in our analysis.Conversely for method 2, all apertures required an ex-tra component to fully describe the H α emission. How-ever, only aperture 7 (which is the same as aperture B inmethod 1) had a FWHM greater than 500 km s − , be-fore correcting for instrumental broadening. We there-fore assume that this extra H α component in apertures1–6 are narrow emission, and include them in our anal-ysis. We show an example of the secondary H α compo-nent for aperture 1 in the bottom panel of Figure 6.0 − − − f λ ( − e r g / s / c m / ˚ A ) Method 1, Aperture B6540 6550 6560 6570 6580 6590 6600Rest Wavelength (˚A) − − f λ ( − e r g / s / c m / ˚ A ) Method 2, Aperture 1
Figure 6.
Top:
The fit to the H α broad component formethod 1, aperture B, the 1 (cid:48)(cid:48) aperture around radio source.The data are shown in black, the full model is shown in red,the narrow line components are shown by the blue dashedlines and the H α broad component is the solid orange line.The residuals are shown below the fit. This broad componenthas a FWHM ∼
900 km s − after correcting for instrumen-tal broadening. We consider this Gaussian to be broad-lineemission, and do not include it in the narrow-line measure-ment. Bottom:
Same as top, but for method 2, aperture1, a 1 (cid:48)(cid:48) aperture offset from the radio source. The “broad”component has a FWHM ∼
300 km s − after correcting forinstrumental broadening. We therefore do not consider thisGaussian to be broad-line emission, and we do include it inthe narrow-line measurement. We find no evidence for a broad component in H β .However, if the broad component is contributing to theobserved H β emission, it could cause an overestimationof the narrow-line flux. If we assume that the broad H β component contributes the same proportional amount offlux as that in H α ( ∼ β flux falls within the calculated flux uncertaintymaking this effect negligible. 4.3.2. [O I ] and [Fe X ] emission In addition to the broad H α component seen near theradio source, we also detect both a broader [O I ] λ X ] λ X ] λ X ] λ β can be usedto calculate the velocity of jet-driven shocks in AGNs(Wilson & Raymond 1999). The integrated flux of the[Fe X ] coronal line is detected with a S / N = 4 .
3, and thepeak of the line is ≈ σ above the surrounding contin-uum.Meanwhile, LINERs with LLAGNs and strong ra-dio emission are known to have enhanced [O I ] emissionwhich can have different profiles from that of [S II ] (Bal-maverde & Capetti 2014). We note that we only detectthe broader [O I ] and [Fe X ] emission in apertures coinci-dent with the radio source. While we do not see similarkinematic components in the other strong lines, [O I ] isnot as easily created by stellar sources than the otherforbidden lines which makes it less subject to galaxy con-tamination. An example fit to the [O I ]+[Fe X ] complexis shown for method 1, aperture B in Figure 7. While theadditional [O I ] component is broader than those seen inthe other forbidden lines, forbidden line emission cannotoriginate in the broad line region. We therefore do notexclude them from our narrow-line analysis.4.3.3. Overall physical conditions of the gas
After measuring the emission lines, we calculatedthe electron densities of all 1–D spectra using the[S II ] λλ T e = 10 K. All of the spectra had densities withinthe range of n e = 500–1000 cm − , and showed noclear spatial trend. We additionally measured the typ-ical velocity widths of the narrow forbidden lines to beFWHM ∼ − . As the spectral resolution is ∼
100 km s − , the narrow lines are marginally resolved.Finally, the metallicity was calculated using equation 1in Pettini & Pagel (2004). We found the metallicity tobe 12 + log( O/H ) = 8 . COMPARISON OF SPECTRA TO MODELS5.1.
Standard BPT Diagnostics
In order to determine the mechanism powering theobserved emission, we compared the measured emissionlines to commonly used models in the standard BPTdiagrams. As these line ratios are reddening insensitive,we do not correct for dust.
H Activity and Outflows in a Dwarf Galaxy f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I] Figure 7.
Same as Figure 6, but for the [O I ] broademission from method 1, aperture B. We label both the[O I ] λλ X ] λ I ] broad component, shown in orange,cannot be explained by the [S II ] profile template alone. Thewidth of this component is FWHM ∼
700 km s − , aftercorrecting for instrumental broadening. Since the forbiddenline emission cannot originate in the broad line region, wedo not exclude it from our analysis. [Fe X ] λ / N = 4 . ≈ σ above the con-tinuum. We compare the [N II ]/H α , [S II ]/H α and [O I ]/H α vs. [O III ]/H β measurements for all apertures for bothmethods to the diagnostics for gas photoionized by hotstars in Figure 8. The Kewley et al. (2006) theoreti-cal extreme starburst lines are shown in solid red, whilethe Kauffmann et al. (2003) empirical composite line forthe [N II ]/H α diagram is shown as a dashed red line. Wepresent the results for each method below. Method 1 –
The apertures in this method repre-sent the “integrated” measurements, with aper-ture A as the smallest region and aperture C asthe largest. The SDSS measurements from Reineset al. (2020) are also plotted for reference.From Figure 8, we see the [N II ]/H α and [S II ]/H α ratios are consistent with ionization by hot stars.However, all of the [O I ]/H α measurements fromthe GMOS data presented in this work fall in-side the Seyfert locus. This is in contrast to theSDSS observation that has a smaller exposure timeand worse spectral resolution ( R ∼ R ∼ I ]/H α is the cleanest diagnostic to discriminatebetween stellar and non-stellar processes since it issensitive to the hardness of the ionizing radiationfield (Kewley et al. 2006). Method 2 –
This method includes the resolved1 (cid:48)(cid:48) radio source spectrum (aperture 7), and re-solved, off-nuclear spectra (apertures 1-6). Thetwo regions that are directly outside the endpointsof the enhanced [O I ] feature are shown as orangearrows (apertures 1 and 4), while the regions sur-rounding the radio source are shown in blue (aper-tures 2, 3, 5 and 6). The black diamond (aperture7) is the same as aperture B in method 1. We notethat apertures 1 and 4 do not show systematicallydifferent line ratios from apertures 2, 3, 5 and 6.This could imply that the outflow does not reachthe region outside the 1 (cid:48)(cid:48) central aperture.The [O I ]/H α emission is again enhanced in thisdiagram, with almost all points consistent withSeyfert-like line ratios. Similarly, the [N II ]/H α and [S II ]/H α are largely consistent with the hotstars and composite loci. As in method 1, there isno single mechanism that can describe all of theobserved emission.5.2. Shocked Emission Diagnostics
The spectra in J1220+3020 do not clearly produceAGN-like emission as defined by the three standard BPTdiagrams (with the exception of the [O I ]/H α diagram).However, there is a well-detected broad H α componentand [Fe X ] line that are clear indicators of an AGN, cre-ating ambiguity in these diagnostics. As other LLAGNshave been well-described by shock models (see Kewleyet al. 2019, and references therein), we also comparedour data to the shock and shock+precursor models fromAllen et al. (2008). We describe the models and pa-rameter constraints below, and compare the data to themodels in Section 5.2.1.The shock models were produced using the MappingsIII code (Allen et al. 2008; Sutherland et al. 2013)and consist of two physical scenarios: simple shocksand shocks with a precursor. For the simple shocksmodel, the gas is collisionally ionized by the propagat-ing shock. Meanwhile, in models with a precursor, theshock-heated gas is allowed to produce photons thattravel upstream and ionize the gas ahead of the shockfront. The shock with a precursor model is usuallyinvoked for situations where both photoionization andshock excitation are present (e.g., in LINERs; Molinaet al. 2018).2 − . − . . α ) − . − . . . . . . l og ( [ O III ] / H β ) Method 1H II AGNComp.
ABCSDSS − . − . . . α )H IISeyfert LINER − . − . − . . α )H IISeyfert LINER − . − . − . − . α ) − . − . . . . . . l og ( [ O III ] / H β ) H IIAGN Comp.Method 2 − . − . − .
25 0 . α )H IISeyfert LINER − . − . − . α )H IISeyfert LINER Figure 8.
Top:
The [O
III ]/H β vs. [N II ]/H α , [S II ]/H α and [O I ]/H α diagrams, or BPT diagrams, for the apertures from method1. The markers representing the different spectra are defined by the legend in the [N II ]/H α diagram. The SDSS spectrum isfrom Reines et al. (2020). The red solid lines are the extreme starburst lines and the Seyfert lines as defined in Kewley et al.(2006). The dashed line in the [N II ]/H α diagram is the composite line from Kauffmann et al. (2003). The labels indicatethe different classification regions for each diagram: H II , composite and AGN for [N II ]/H α , and H II , Seyfert and LINER for[S II ]/H α and [O I ]/H α . There is clearly enhanced [O I ]/H α emission that cannot be described by star formation. Therefore,there is no clear mechanism among all three diagrams. Bottom:
Same as top, but for method 2. We again see enhanced [O I ]/H α emission not explained by star formation, creating ambiguity in these diagnostics. The Mappings III models cover a wide range of mag-netic field strengths, velocities, densities and metallici-ties. While the radio source could have a magnetic fieldstrength of B (cid:38) µ G, this would only hold within the < . (cid:48)(cid:48) (cid:48)(cid:48) , which is much largerthan the radio source, so the magnetic field strengthcould be very low. We therefore chose to set the mag-netic field strength at 1 µ G, as its grid overlapped with most of the models incorporating magnetic fields of lessthan 10 µ G. In general, AGNs can have a wide range ofradio strengths (see Table 1 of Ho 2008), and the sub-sequent magnetic field strength is dependent on boththe strength and morphology of that radio emission (seeequation 5.109 in Condon & Ransom 2016).In order to set the velocity constraint, we examinedthe widths of the narrow emission lines. The forbiddenlines typically had FWHMs ∼ − . How- H Activity and Outflows in a Dwarf Galaxy I ] line was slightly broader, especially inthe radio source spectrum, which had a separate broadcomponent with FWHM ∼
700 km s − . If shocks areindeed driving the observed emission, the observed lowvelocity widths of the forbidden lines imply that theshock should also have a low velocity. We therefore con-strained the velocities in the models to be within therange v = 100–500 km s − , where the upper limit is thetypical boundary between low- and high-velocity shocks.The density and metallicity parameters are jointly ex-plored in the Allen et al. (2008) models. For all metal-licities that are not solar, a density of n = 1 cm − is assumed. Only the solar metallicity models explorechanges in density, from n = 0 . − cm − . The galaxystudied here has a sub-solar metallicity, determined bythe [N II ]/H α ratio, and a density of n ≈ cm − , givenby the [S II ] λλ n = 1 − − , as well as models with a fixeddensity of n = 1 cm that assume metallicities consistentwith the Large and Small Magellanic Clouds (LMC andSMC).Finally, for the radio source spectrum, we detect theAGN coronal line [Fe X ] λ X ] λ β ra-tio can be used to calculate the jet-driven shock veloci-ties (Wilson & Raymond 1999). After correcting for red-dening, we find the [Fe X ] λ β ∼ .
23, which im-plies a shock velocity in the range v = 200–300 km s − ,which is consistent with the measured velocities seen inthe narrow forbidden lines.5.2.1. Comparison to Shock Models
We over-plot our data on the shock andshock+precursor models for all three BPT diagramsin Figure 9. We note that given the inconsistenciesin density and metallicity between the data and themodels, the [N II ]/H α diagram will likely have moresystematic errors than the other two diagrams.We only show the apertures from method 2 in theseplots as method 1 is focused on integrated light. Wequalitatively discuss each aperture in detail below,and give more quantitative, overall conclusions in Sec-tion 5.3.For reference, all of the apertures in method 2 arespatially resolved. Aperture 7 is the 1 (cid:48)(cid:48) aperture aroundthe radio source, while apertures 1 and 4 are the twonon-nuclear regions that follow the semi-major axis ofthe enhanced [O I ] emission. The remaining apertures (2, 3, 5 and 6) are non-nuclear apertures not associatedwith the enhanced [O I ] emission. Aperture 1:
This emission is consistent with bothshocks and shocks+precursor models in all threediagrams. We therefore do not discriminate be-tween these mechanisms, but note that shocks arecontributing to the observed emission.
Aperture 2:
This emission is technically consis-tent with both the shocks and shocks+precursormodels in all three diagrams. However, the shocksmodel appears to better describe the data in the[O I ]/H α diagram. In order to be explained byshocks+precursor, the gas would need to have adensity less than 1 cm − even with a sub-solarmetallicity. This is inconsistent with our densitymeasurement of n e = 500 − − . We there-fore conclude that the gas in aperture 2 is likelydominated by shocked emission. Aperture 3:
We cannot discriminate between theshock and shock+precursor mechanisms given thelower limit on the [O
III ]/H β ratio. We do notethat shocks are contributing to the observed emis-sion. Aperture 4:
Similar to aperture 1, we cannot dis-criminate between the shock and shock+precursormechanisms but note that shocks are contributingto the observed emission.
Aperture 5:
Similar to aperture 3, we cannot dis-criminate between the shock and shock+precursormechanisms given the lower limit on the [O
III ]/H β ratio. We do note that shocks are contributing tothe observed emission. Aperture 6:
Similar to aperture 5, we cannot dis-criminate between the shock and shock+precursormechanisms given the lower limit on the [O
III ]/H β ratio. We do note that shocks are contributing tothe observed emission. Aperture 7:
This aperture is centered on the ra-dio source, and is consistent with the shocks andshock+precursor models in all three diagrams.However given the detected AGN coronal line[Fe X ] λ α com-ponent (FWHM ≈
900 km s − , not included inthe analysis here), we know that photoionizationis likely contributing to the observed emission. Wetherefore conclude that the mechanism that bestdescribes the gas is a shocks+precursor model.4 ° . ° . ° . . ° . . . . ° . ° . ° . . . . . . . . . log([N II] ∏ Æ ) . . . . . . l og ( [ O III ] ∏ H Ø ) v =
500 km/sv =
200 km/s v = k m / s v = k m / s Z S M C , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - ° . ° . ° . . . ° . . . . . ° . ° . . . . . . . . . log([S II] ∏ Æ ) . . . . . . l og ( [ O III ] ∏ H Ø ) v = k m / s v = k m / s v = k m / s v = k m / s Z S M C , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - ° . ° . ° . . . ° . . . . . . ° . ° . ° . . . . . . . . . log([O I] ∏ Æ ) . . . . . . l og ( [ O III ] ∏ H Ø ) v = k m / s v = k m / s v = k m / s v = k m / s Z S M C , n = c m - Z ⊙ , n = c m - Z ⊙ , n = c m - Z S M C , n = c m - Figure 9.
Comparison of the emission line measurements from method 2 to the Allen et al. (2008) shock and shock+precursormodels. For all diagrams, the simple shock models are shown in magenta and the shock+precursor models are shown in black.The red dashed lines for both the shock and shock+precursor models represent the models assuming the SMC metallicity, whilethe blue dot-dashed line assumes the LMC metallicity. Both the SMC and LMC models assume a density of n = 1 cm − . Theedges of the grids are labeled by the value that defines the line (velocity or density). We note that the velocity always increasesdownwards in the shock models and upwards in the shock+precursor models. We also show standard hot star photoionizationdiagnostics (i.e., Kewley et al. 2006; Kauffmann et al. 2003) in gray. The measurements presented are from method 2, and thesymbol representing each aperture is indicated in the legend in each subplot. We find the simple shocks model does a betterjob at describing the regions outside the radio source (all but aperture 7), and the shocks+precursor models best describe theemission coincident with the radio source (aperture 7). A detailed description of the results is given in Section 5.2.1. TopLeft:
The [N II ]]/H α diagram. We expect this diagram to have the most inconsistency between the data and models due to thestrong dependence on metallicity. However, we find that the shocks and shocks+precursor models still do an adequate job ofdescribing the data. Top Right:
The [S II ]/H α diagram. Both the shocks and shock+precursor models describe the data well. Bottom:
The [O I ]/H α diagram. The simple shocks models describe the data in aperture 2 better than the shocks+precursormodels. The remaining apertures outside the radio source (aperture 1, 3, 5 and 6) can be explained by either the shocks orshocks+precursor models. While, the shocks+precursor and shocks models can both describe the emission coincident with theradio source (aperture 7), the presence of the broad H α and [Fe X ] line strongly indicates the influence of photoionization. Wetherefore conclude that the shocks+precursor model best describes the emission coincident with the radio source. H Activity and Outflows in a Dwarf Galaxy
General Conclusions from Model Comparisons
We find overall that photoionization from hot starsdo not adequately describe the data. After includingcareful constraints on the density, metallicities, and ve-locities of the shock and shock+precursor models, wefind that the apertures associated with the radio sourceand at the two edges of the enhanced [O I ] feature arewell-described by shock+precursor models, while theother apertures are well-described by the simple shocksmodel. The predicted shock velocities from the modelare v ∼ −
300 km s − , which is consistent measuredFWHM values for the narrow-line emission. While wecannot cleanly define the predicted metallicity and den-sity measurements expected from the models, the major-ity of the data lie are consistent with n e = 10 − with Z = Z (cid:12) . Therefore, it is possible for the data tobe consistent with Z < Z (cid:12) and 100 < n e < ,which agrees with the metallicity and density estimatesfrom the nebular emission. This is especially true in the[O III ]/H β vs. [S II ]/H α diagram, which is sensitive toshock excitation (Rich et al. 2010; Molina et al. 2018).We therefore conclude that the data are consistent withthe shock models. THE BH AND ITS ENVIRONMENTWe conclude that an active BH is present inJ1220+3020, due to the strong radio detection fromReines et al. (2020), the broad H α emission, enhanced[O I ] emission and the detection of the AGN coronalline [Fe X ] λ α emission originatesfrom the broad line region of the BH. Under that as-sumption, we use equation 5 from Reines et al. (2013)and find M BH = 10 . M (cid:12) . The predicted M BH forJ1220+3020 given its stellar mass (log[ M ∗ /M (cid:12) ] = 9 . M BH = 10 . –10 M (cid:12) (Reines & Volonteri 2015). Themedian M BH in the dwarf galaxies studied in Reineset al. (2013) was M BH ≈ × M (cid:12) . Therefore ourBH mass measurement is consistent with other BHs ingalaxies of a similar mass range.6.1. Contribution of Shocks
Despite the clear indicators described in Section 5.3,the traditional BPT diagnostics did not identify the ob-ject as “AGN-like” in all three diagrams for SDSS dataand the [N II ]/H α and [S II ]/H α diagrams for GMOS.The difference between these two data sets can in partbe explained by the poorer spectral resolution and largerobserving region of the SDSS data. The larger SDSSaperture lets in more of the host galaxy light which willlikely contaminate the AGN signal (Moran et al. 2002).Additionally, poorer spectral resolution blurs line pro- files, and could potentially hide the weaker broad com-ponents, including those seen in [O I ] λ II ]/H α and[S II ]/H α diagrams, which best explains the ambiguityseen in the traditional BPT diagnostics.We conclude that the best-fitting physical model is aLLAGN that photoionizes the nearby gas, and drivesa wind or other outflow that creates shocked emissionon larger spatial scales. Outside of the outflow’s di-rect influence, turbulent motion of the gas caused bythe outflow could also create shocked emission. Thismodel explains the enhanced [O I ] emission in both the2–D map and 1–D radio source spectrum, the [Fe X ] andbroad H α emission and the observed narrow emission-line ratios. While stellar winds could potentially explainthe observed shock-excited gas, they cannot explain thestrong [Fe X ] and radio emission, which strongly disfa-vors this scenario.6.2. Structure of the Central Engine L bol /L Edd (cid:46) − ) that there exists a radius, R t wherethe collisional cooling timescale is equal to the accretiontimescale. Outside of R t , the accretion structure is a ge-ometrically thin, optically thick disk. Meanwhile inside R t , the accretion disk becomes geometrically thick andoptically thin, as the ions remain at the virial temper-ature while the electrons are cooled by bremsstrahlung,synchrotron, and Compton up-scattering. In fact, thisstructure could remove some of the more traditionalAGN-like narrow-line ratios (Ho 2008; Trump et al.2011). RIAFs are both theoretically predicted and ob-servationally shown to have strong radio outflows (Meier2001; Fender & Belloni 2004). Additionally, increasedradio strength has been observationally tied to decreasedaccretion strength (Mel´endez et al. 2008; Diamond-Stanic et al. 2009).We estimate the bolometric luminosity of the AGNby first converting the broad H α luminosity detected inthe resolved 1 (cid:48)(cid:48) spectrum of the radio source to L (5100˚A)using equation 2 from Greene & Ho (2005), and then ap-plying the L bol = 10 . L (5100˚A) relation from Richardset al. (2006). The Eddington ratio for the BH inJ1220+3020 is L bol /L Edd ∼ .
03, which is consistentwith a RIAF-powered engine.We note that there are usually no broad emissionlines in RIAFs due to the cooler inner disk (Trumpet al. 2011), unlike J1220+3020. However, we onlydetect weak broad H α emission, with L (H α broad ) =4 . × erg s − , and there is a precedent for detectedbroad emission in AGNs with very low accretion rates( L int /L Edd < − ; Ho 2009).We also note that there is a ∼ . M BH estimates (Vestergaard & Peterson 2006;Shen 2013), including the one quoted here. Therefore,the range of M BH = 10 . –10 . M (cid:12) results in an Ed-dington ratio within the range ∼ . M BH estimate. If we have over-estimated the BH mass, thenthe accretion rate would be too high to be explainedby a RIAF engine. Therefore, without a full SED, wecannot definitively claim the central engine has a RIAF.However, the lack of AGN-like narrow emission-line ra-tios, the strong radio emission and the evidence for AGNoutflows, i.e., [Fe X ] and enhanced [O I ], are all consistentwith the RIAF model, and we therefore favor that ex-planation. If there is a RIAF, the BH mass and AGN luminosityestimates could be impacted by the potential suppres-sion of broad emission by the cooler inner disk (Yuan& Narayan 2004; Ho 2008; Trump et al. 2011), but wereiterate that our estimated BH mass is consistent withthose in the same galaxy mass range.6.2.1. Contribution from the Radio Jet
We use the VLA Faint Images of the Radio Sky atTwenty centimeters (FIRST) survey 1.4 GHz detection(Becker et al. 1995) and equation 1 from Cavagnoloet al. (2010) to calculate the jet power. Given thatReines et al. (2020) found that the radio emission isdominated by the AGN candidate in this object, we as-sume there is little contamination by star formation. Wenote that this relation was calculated using AGNs in gi-ant ellipticals with radio luminosities similar to that ofJ1220+2030. If we assume this relation holds, we cal-culate P jet ≈ . × erg s − , which is about 9% ofthe Eddington luminosity assuming the BH mass givenabove. Thus the majority of AGN power is likely drivenby emission from the radio jet, not accretion.As discussed in Cavagnolo et al. (2010), the P jet pre-sented here is actually a calculation for P cav , which isthe mechanical power needed to create the radio cav-ity. However, they note that since the mechanical P cav is significantly larger than the synchrotron power of thejet, they assume P jet = P cav . Finally, this measurementdoes not include the energy channeled into shocks, whichcan greatly exceed P cav (Nulsen et al. 2005).In addition to the low Eddington ratio, the RIAFmodel is associated with very strong radio emission andevidence for outflows or other turbulent motion in thegas creating shocked emission. We see evidence for verystrong radio emission via the larger energy contribu-tion from the radio jet than accretion. While we donot observe large radio outflows, this black hole has a M BH = 10 . M (cid:12) , which would likely not be able to pro-duce large radio lobes. However, there is evidence forturbulent motion and outflows via the [Fe X ] emission.While the coronal [Fe X ] line is typically thought to bepowered by AGN photoionization (e.g., Nussbaumer &Osterbrock 1970; Korista & Ferland 1989; Oliva et al.1994; Pier & Voit 1995), it can also be produced byshock-excited gas from out-flowing winds caused by ra-dio jets (Wilson & Raymond 1999). If we assumethe latter scenario, we can calculate the expected ve-locity of the jet-driven shock using the [Fe X ]/H β ra-tio. The predicted velocity is ∼ −
300 km s − ,which is consistent with the measured line dispersions(FWHM ∼ −
300 km s − ). Given the narrow-lineevidence for shocks and the radio detection from Reines H Activity and Outflows in a Dwarf Galaxy X ] line is likely createdby shocks driven by out-flowing winds from the radiojets associated with the BH.6.3. Comparison to Optically-Selected BHs
We conclude that the engine in J1220+3020 is largelyconsistent with the RIAF model, and that the shockedemission due to out-flowing gas or winds partially ob-scures the optical AGN photoionization signatures. Thisindicates that radio-selected AGN may probe a differentpopulation of active BHs in dwarf galaxies than stan-dard optical diagnostics.Optical diagnostics are known to pick out bright, highaccretion- rate BHs in dwarf galaxies where photoioniza-tion dominates the energy budget (Reines & Comastri2016). As the accretion rate decreases, radio emission in-creases and outflows and winds become more importantto the overall energy budget. Therefore, radio-selectedBHs in dwarf galaxies will likely have lower accretionrates than those found in optical surveys. Given thata significant fraction of supermassive active BHs in thelocal Universe are LLAGNs (Ho 2008), radio surveysmight find a significant population of BHs in local dwarfgalaxies. SUMMARY AND CONCLUSIONSIn this paper we present the first optical follow-up ofthe radio-selected BH candidates discovered by Reineset al. (2020). We focus on J1220+3020, which is a brightnuclear BH candidate, to identify any unique featuresassociated with radio-selected AGNs in dwarf galaxies.We use GMOS-N/IFU data to study look for opticalsignatures of an active BH, and study its 2–D physicalenvironment. Our results are as follows: • There is an elongated feature of enhanced[O I ] λ ≈
700 km s − ) [O I ] compo-nent in the 1–D spectrum of the radio source. Weconclude these are likely signatures of an outflow. • In the 1–D spectrum of the radio source, we de-tect the AGN coronal line [Fe X ] λ α (FWHM ≈
900 km s − ), which both indicatethe presence of an AGN. Using the FWHM andluminoisity of the broad H α emission, we estimatea BH mass of M BH = 10 . M (cid:12) . • We compare our data to both standard BPT di-agnostics and the Allen et al. (2008) shock andshock+precursor models. We conclude that theLLAGN photoionizes the gas in its immediatevicinity, while an outflow or winds associated with the LLAGN shock-excites gas further away anddrives the [Fe X ] λ • We favor the RIAF model as it is consistent withalmost all of the observed data. Finally, weconclude that shocks associated with outflows orwinds significantly contribute to the AGN energybudget. • Sensitive, high resolution radio surveys of dwarfgalaxies can probe a population of BHs with loweraccretion rates than optical surveys.While we do not find any 2–D velocity structure thatwould indicate an an outflow, the enhanced 2–D [O I ]emission and broadened [O I ] emission associated withthat feature, the strong radio and [Fe X ] detection, andthe agreement of shock models with the [Fe X ]/H β ra-tio and the measured forbidden emission-line widths allstrongly support this scenario. Finally, we note theAGN in J1220+3020 would not be identified throughthe [N II ]/H α diagram alone, nor relying solely on stan-dard BPT diagnostics. Therefore by limiting our workto that diagram and set of models, we are likely missinga large portion of the AGN population in dwarf galax-ies. In order to address this problem, future work shouldinclude a the use of shock models, and if possible [O I a community-developed core Python package for Astronomy (AstropyCollaboration et al. 2013, 2018). Software:
APLpy (Robitaille & Bressert 2012), As-tropy (Astropy Collaboration et al. 2013, 2018), GeminiIRAF (Gemini Observatory & AURA 2016), IFSRED(Rupke 2014), IRAF (National Optical Astronomy Ob-servatories 1999), L.A.-cosmic (van Dokkum 2001), Mat-plotlib (Hunter 2007), PyRAF (Green 2012), pyspeckit(Ginsburg & Mirocha 2011)APPENDIX A. EMISSION-LINE FITS FOR THE 1 (cid:48)(cid:48)
RADIO SOURCE SPECTRUMIn this appendix we show the fits to the strong emission lines in the 1 (cid:48)(cid:48) spectrum of the radio source in Figure 10.We specifically focus on the emission lines used in our analysis. H Activity and Outflows in a Dwarf Galaxy − − f λ ( − e r g / s / c m / ˚ A ) [O III] λλ f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I]6280 6300 6320 6340 6360 6380Rest Wavelength (˚A)02468 f ( e r g / s / c m / ˚ A / a r c s e c ) [Fe X][O I][O I] − f λ ( − e r g / s / c m / ˚ A ) [S II] λλ Figure 10.
The fits for the emission lines in the 1 (cid:48)(cid:48) spectrum of the radio source. In each plot, the fit is shown in red, the dataare shown in black and the individual components are shown in blue. The emission lines shown are indicated in each subplot.The broad component of the [O I ] component is shown in orange. We do not do the same for the broad component of H α , as itis not clearly visible in this image. We show a zoomed in plot of the H α broad feature in Figure 6. Allen, M. G., Groves, B. A., Dopita, M. A., Sutherland,R. S., & Kewley, L. J. 2008, ApJS, 178, 20Angl´es-Alc´azar, D., Faucher-Gigu`ere, C.-A., Quataert, E.,et al. 2017, MNRAS, 472, L109Astropy Collaboration, Robitaille, T. P., Tollerud, E. J.,et al. 2013, A&A, 558, A33Astropy Collaboration, Price-Whelan, A. M., Sip˝ocz, B. M.,et al. 2018, AJ, 156, 123Aversa, A. G., Johnson, K. E., Brogan, C. L., Goss, W. M.,& Pisano, D. J. 2011, AJ, 141, 125Baldwin, J. A., Phillips, M. M., & Terlevich, R. 1981,PASP, 93, 5Balmaverde, B., & Capetti, A. 2014, A&A, 563, A119Becker, R. H., White, R. L., & Helfand, D. J. 1995, ApJ,450, 559Begelman, M. C., Volonteri, M., & Rees, M. J. 2006,MNRAS, 370, 289Bellovary, J., Volonteri, M., Governato, F., et al. 2011,ApJ, 742, 13Blandford, R. D., & Begelman, M. C. 1999, MNRAS, 303,L1Blanton, M. R., Kazin, E., Muna, D., Weaver, B. A., &Price-Whelan, A. 2011, AJ, 142, 31Blanton, M. R., Bershady, M. A., Abolfathi, B., et al. 2017,AJ, 154, 28Bromm, V., & Yoshida, N. 2011, ARA&A, 49, 373Cann, J. M., Satyapal, S., Abel, N. P., et al. 2019, ApJL,870, L2Capetti, A., Axon, D. J., & Macchetto, F. D. 1997, ApJ,487, 560Cavagnolo, K. W., McNamara, B. R., Nulsen, P. E. J.,et al. 2010, ApJ, 720, 1066Cecil, G., Morse, J. A., & Veilleux, S. 1995, ApJ, 452, 613Choi, J.-H., Shlosman, I., & Begelman, M. C. 2015,MNRAS, 450, 4411Chomiuk, L., & Wilcots, E. M. 2009, ApJ, 703, 370Condon, J. J. 1997, PASP, 109, 166Condon, J. J., Matthews, A. M., & Broderick, J. J. 2019,ApJ, 872, 148Condon, J. J., & Ransom, S. M. 2016, Essential RadioAstronomyDavies, M. B., Miller, M. C., & Bellovary, J. M. 2011,ApJL, 740, L42Devecchi, B., & Volonteri, M. 2009, ApJ, 694, 302Dey, A., Schlegel, D. J., Lang, D., et al. 2019, AJ, 157, 168Diamond-Stanic, A. M., Rieke, G. H., & Rigby, J. R. 2009,ApJ, 698, 623 Dopita, M. A. 2002, in Revista Mexicana de Astronomia yAstrofisica Conference Series, Vol. 13, Revista Mexicanade Astronomia y Astrofisica Conference Series, ed. W. J.Henney, W. Steffen, L. Binette, & A. Raga, 177–182Dopita, M. A., Koratkar, A. P., Allen, M. G., et al. 1997,ApJ, 490, 202Falcke, H., Wilson, A. S., & Simpson, C. 1998, ApJ, 502,199Fender, R., & Belloni, T. 2004, ARA&A, 42, 317Ferruit, P., Wilson, A. S., Whittle, M., et al. 1999, ApJ,523, 147Gemini Observatory, & AURA. 2016, Gemini IRAF: Datareduction software for the Gemini telescopesGinsburg, A., & Mirocha, J. 2011, PySpecKit: PythonSpectroscopic ToolkitGreen, W. 2012, Society for Astronomical Sciences AnnualSymposium, 31, 159Greene, J. E., & Ho, L. C. 2005, ApJ, 630, 122Greene, J. E., Strader, J., & Ho, L. C. 2020, ARA&A, 58,257Groves, B. A., Heckman, T. M., & Kauffmann, G. 2006,MNRAS, 371, 1559Habouzit, M., Volonteri, M., & Dubois, Y. 2017, MNRAS,468, 3935Hainline, K. N., Reines, A. E., Greene, J. E., & Stern, D.2016, ApJ, 832, 119Ho, L. C. 2008, ARA&A, 46, 475—. 2009, ApJ, 699, 626Hunter, J. D. 2007, Computing in Science Engineering, 9,90Jarrett, T. H., Cohen, M., Masci, F., et al. 2011, ApJ, 735,112Johnson, K. E., Hunt, L. K., & Reines, A. E. 2009, AJ, 137,3788Kauffmann, G., Heckman, T. M., Tremonti, C., et al. 2003,MNRAS, 346, 1055Kepley, A. A., Reines, A. E., Johnson, K. E., & Walker,L. M. 2014, AJ, 147, 43Kewley, L. J., Groves, B., Kauffmann, G., & Heckman, T.2006, MNRAS, 372, 961Kewley, L. J., Nicholls, D. C., & Sutherland, R. S. 2019,ARA&A, 57, 511Korista, K. T., & Ferland, G. J. 1989, ApJ, 343, 678Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511Kormendy, J., & Richstone, D. 1995, ARA&A, 33, 581Laor, A. 1998, ApJL, 496, L71Lodato, G., & Natarajan, P. 2006, MNRAS, 371, 1813Loeb, A., & Rasio, F. A. 1994, ApJ, 432, 52
H Activity and Outflows in a Dwarf Galaxy Lupi, A., Colpi, M., Devecchi, B., Galanti, G., & Volonteri,M. 2014, MNRAS, 442, 3616Mateos, S., Alonso-Herrero, A., Carrera, F. J., et al. 2012,MNRAS, 426, 3271Meier, D. L. 2001, ApJL, 548, L9Mel´endez, M., Kraemer, S. B., Armentrout, B. K., et al.2008, ApJ, 682, 94Molina, M., Eracleous, M., Barth, A. J., et al. 2018, ApJ,864, 90Moran, E. C., Filippenko, A. V., & Chornock, R. 2002,ApJL, 579, L71Narayan, R., & Yi, I. 1995, ApJ, 452, 710Natarajan, P. 2014, General Relativity and Gravitation, 46,1702National Optical Astronomy Observatories. 1999, IRAF:Image Reduction and Analysis FacilityNetzer, H. 2013, The Physics and Evolution of ActiveGalactic NucleiNulsen, P. E. J., Hambrick, D. C., McNamara, B. R., et al.2005, ApJL, 625, L9Nussbaumer, H., & Osterbrock, D. E. 1970, ApJ, 161, 811Oetken, L. 1977, Astronomische Nachrichten, 298, 187Oliva, E., Salvati, M., Moorwood, A. F. M., & Marconi, A.1994, A&A, 288, 457Penston, M. V., Fosbury, R. A. E., Boksenberg, A., Ward,M. J., & Wilson, A. S. 1984, MNRAS, 208, 347Pettini, M., & Pagel, B. E. J. 2004, MNRAS, 348, L59Pier, E. A., & Voit, G. M. 1995, ApJ, 450, 628Portegies Zwart, S. F., Baumgardt, H., Hut, P., Makino, J.,& McMillan, S. L. W. 2004, Nature, 428, 724Reines, A. E., & Comastri, A. 2016, PASA, 33, e054Reines, A. E., Condon, J. J., Darling, J., & Greene, J. E.2020, ApJ, 888, 36Reines, A. E., Greene, J. E., & Geha, M. 2013, ApJ, 775,116 Reines, A. E., Johnson, K. E., & Goss, W. M. 2008, AJ,135, 2222Reines, A. E., & Volonteri, M. 2015, ApJ, 813, 82Rich, J. A., Dopita, M. A., Kewley, L. J., & Rupke,D. S. N. 2010, ApJ, 721, 505Richards, G. T., Lacy, M., Storrie-Lombardi, L. J., et al.2006, ApJS, 166, 470Robitaille, T., & Bressert, E. 2012, APLpy: AstronomicalPlotting Library in PythonRupke, D. S. N. 2014, IFSRED: Data Reduction forIntegral Field SpectrographsSabra, B. M., Shields, J. C., Ho, L. C., Barth, A. J., &Filippenko, A. V. 2003, ApJ, 584, 164Shen, Y. 2013, Bulletin of the Astronomical Society ofIndia, 41, 61Stasi´nska, G., Cid Fernandes, R., Mateus, A., Sodr´e, L., &Asari, N. V. 2006, MNRAS, 371, 972Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ,753, 30Stone, N. C., K¨upper, A. H. W., & Ostriker, J. P. 2017,MNRAS, 467, 4180Sutherland, R., Dopita, M., Binette, L., & Groves, B. 2013,MAPPINGS III: Modelling And Prediction inPhotoIonized Nebulae and Gasdynamical ShocksTrump, J. R., Impey, C. D., Kelly, B. o. C., et al. 2011,ApJ, 733, 60van Dokkum, P. G. 2001, PASP, 113, 1420Varenius, E., Conway, J. E., Batejat, F., et al. 2019, A&A,623, A173Vestergaard, M., & Peterson, B. M. 2006, ApJ, 641, 689Volonteri, M. 2010, A&A Rv, 18, 279Wilson, A. S., & Raymond, J. C. 1999, ApJL, 513, L115Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al.2010, AJ, 140, 1868Yuan, F., & Narayan, R. 2004, ApJ, 612, 724 T a b l e . N a rr o w E m i ss i o n L i n e F l u x e s f o r J + E m i ss i o n L i n e F l u x a M e t h o d N u m b e r H β λ [ O III ] λ [ O III ] λ [ O I ] λ [ O I ] λ [ F e X ] λ [ N II ] λ H α λ [ N II ] λ [ S II ] λ [ S II ] λ A b ± ± ± ± . ± . . ± . ± ± ± ± ± B ± ± ± ± ± . ± . . ± . ± . ± . ± ± C ± ± ± ± ± ... ± . ± ± ± ± ± ± ± . ± . . ± . ... . ± . ± . ± . . ± . . ± . ± ± ± . ± . . ± . ... . ± . ± . ± . . ± . . ± . < ± ± . ± . . ± . ... . ± . ± ± . . ± . . ± . < ± ± . ± . . ± . ... . ± . ± . ± . . ± . . ± . < . ± . ± . ± . . ± . ... . ± . ± . ± . . ± . . ± . < ± ± ± . ± . ... . ± . ± . ± . . ± . . ± . c ± ± ± ± ± . ± . . ± . ± . ± . ± ± N / A S D SS d ± ± ± ± ± ... ± ± ± ± ± a F l u x e s a r e p r e s e n t e d i nun i t s o f − e r g s − c m − , a nd a r e n o t c o rr ec t e d f o rr e dd e n i n g . b M e t h o d , a p e r t u r e A i s n o t s p a t i a ll y r e s o l v e d . c T h i s Sp ec t r u m i s t h e s a m e a s m e t h o d , a p e r t u r e B . d T h e S D SS m e a s u r e m e n t s p r e s e n t e dh e r e w e r ec a l c u l a t e db y R e i n e s e t a l. ( ) , a ndu s e t h e (cid:48)(cid:48) S D SS D R s p ec t r u mm