MNRAS 439, 3835–3846 (2014) doi:10.1093/mnras/stu234 Advance Access publication 2014 March 3 Starburst–AGN mixing – I. NGC 7130 Rebecca L. Davies,1‹ Jeffrey A. Rich,2 Lisa J. Kewley1,3 and Michael A. Dopita1,4 1Research School of Astronomy and Astrophysics, Australian National University, Cotter Road, Weston, ACT 2611, Australia 2Carnegie Observatories, 813 Santa Barbara Street, Pasadena, CA 91101, USA 3Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA 4Astronomy Department, King Abdulaziz University, PO Box 80203, Jeddah, Saudi Arabia Accepted 2014 January 31. Received 2014 January 27; in original form 2013 October 21 Downloaded from ABSTRACT We present an integral field spectroscopic study of the luminous infrared galaxy NGC 7130, a known starburst–active galactic nucleus (AGN) composite galaxy. We employ standard emission-line ratio diagnostics and maps of velocity dispersion and velocity field to investigate how the dominant ionizing sources change as a function of radius. From the signatures of both http://mnras.oxfordjournals.org/ star formation and AGN activity we show that NGC 7130 is a remarkably clean case of starburst–AGN mixing. We find a smooth transition from AGN-dominated emission in the centre to pure star-forming activity further out, from which we can estimate the radius of the extended narrow line region to be 1.8 ± 0.8 kpc. We calculate that the fraction of [O III] luminosity due to star formation and AGN activity is 30 ± 2 and 70 ± 3 per cent, respectively, and that the fraction of Hα luminosity due to star formation and AGN activity is 65 ± 3 and 35 ± 2 per cent, respectively. We conclude with a discussion of the importance and potential of starburst–AGN mixing for future studies of the starburst–AGN connection. at The Australian National University on August 31, 2014 Key words: galaxies: active – galaxies: individual: NGC 7130 – galaxies: Seyfert. more fundamental which is yet to be uncovered (e.g. Alexander & 1 INTRODUCTION Hickox 2012, review; Mullaney et al. 2012). One of the longest standing mysteries in astrophysics is the nature Naively, one might suggest that the scaling relations simply fall of the relationship between star formation and active galactic nu- out of the common requirement of cold gas to fuel both star for- cleus (AGN) activity in composite galaxies. For many years, it has mation and AGN activity. This explanation is problematic, as gas been well established that there is a strong relationship between must lose ≈99.9 per cent of its angular momentum in order to move the accretion activity of supermassive black holes and the evolution from a ≈10 kpc stable orbit in a star-forming disc to the zone of of their host galaxies (Cattaneo, Haehnelt & Rees 1999; Haehnelt influence of an AGN on a ≈10 pc scale (Jogee 2006). However, it & Kauffmann 2000; Cid Fernandes et al. 2001; Kauffmann et al. is well established both theoretically and observationally that black 2003; Granato et al. 2004; LaMassa et al. 2012). The mass of a hole activity must impact star formation. Black hole feedback can supermassive black hole has been found to scale with properties heat material in the interstellar medium of a galaxy (e.g. Sijacki & of the host galaxy such as velocity dispersion (M–σ relation), the Springel 2006), or drive powerful outflows resulting in significant stellar mass in the bulge (MBH–M relation) and the luminosity of gas redistribution (e.g. Rich et al. 2010; Alatalo et al. 2011; De- the bulge (MBH–L relation; e.g. Magorrian et al. 1998; Ferrarese & Buhr, Quataert & Ma 2011; Rich, Kewley & Dopita 2011). These Merrit 2000;Gebhardtetal.2000;Tremaineetal.2002; Marconi processes deplete the galaxy’s cold gas reservoir, resulting in the & Hunt 2003; Gultekin et al. 2009; Bennert et al. 2011; McConnell regulation or quenching of star formation (e.g. Di Matteo, Springel &Ma2013). Although these relationships indicate that by z = 0, & Hernquist 2005;Crotonetal.2006; Hopkins et al. 2006, 2008; the building of the stellar mass and black hole mass in galaxies is in Sijacki et al. 2007; Booth & Schaye 2009; McCarthy et al. 2010). some way correlated, it is unclear whether such a correlation is the Black hole feedback directly impacts the results of galaxy evolution result of a connection between black hole and star formation activ- simulations, which consistently indicate that AGN feedback is piv- ity on time-scales relevant to the formation of stars (facilitated by otal for preventing overproduction of stellar mass in galaxies (e.g. processes such as mergers, starburst-driven winds and AGN-driven Croton et al. 2006; Vogelsberger et al. 2013). outflows; e.g. Yuan, Kewley & Sanders 2010; Rafferty et al. 2011), However, the finding that approximately 30–50 per cent of Seyfert or whether these scaling relations are simply pointing to something 2 nuclei are associated with young stellar populations (e.g. Gonza- lez Delgado, Heckman & Leitherer 2001; Storchi-Bergmann et al. 2001;CidFernandesetal.2004; Gonzalez Delgado & Cid Fernan- E-mail: [email protected] des 2005; Sarzi et al. 2007) seems to suggest that AGN activity C 2014 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society 3836 R. L. Davies et al. may enhance star formation, and/or that circumnuclear star forma- et al. 2006), NGC 5135 (Bedregal et al. 2009), IRAS 19254−7245 tion may enhance AGN activity. Silverman et al. (2009) found that (Reunanen, Tacconi-Garman & Ivanov 2007) and HE 2211−3903 the vast majority of AGN hosts at z ≤ 1 have star formation rates (Scharwachter¨ et al. 2011). IFU data also provide the information (SFRs) higher than that of non-AGN hosts of similar mass, and that required to study how the contribution of the starburst and the AGN the incidence of AGN activity increases with decreasing stellar age. vary with distance from the central AGN, probing the radius of the All this evidence suggests that star formation itself is an impor- AGN extended narrow-line region (ENLR). tant source of viscosity in the circumnuclear gas, assisting some In this paper, we capitalize on the strengths of IFU data to carry of the gas to fall into the AGN, while driving the rest into states out a detailed study of the ionization sources and kinematics in of higher angular momentum. In order to gain greater insight into NGC 7130. NGC 7130 is a LIRG, observed using the Wide Field the starburst–AGN connection, it is necessary to study the relation- Spectrograph (WiFeS) on the Australian National University (ANU) ship between black hole and star formation activity on relatively 2.3-m telescope as part of the Great Observatory All-Sky LIRG instantaneous time-scales. However, it is notoriously difficult to ac- Survey (GOALS; Armus et al. 2009). LIRGs are particularly ap- curately calculate quantities such as SFRs in starburst–AGN com- propriate targets for study of the starburst–AGN connection as they posite galaxies. In such galaxies, the photometric and spectroscopic commonly contain both starbursts and AGN. The prevalence of dual indicators of star formation will be contaminated by emission from energy sources in LIRGs is likely to be a direct result of their his- Downloaded from the AGN, so that it is only possible to derive upper limits on SFRs, tory as mergers of molecular gas-rich spirals, which are thought to with an extremely large margin for error. be the trigger for powerful bursts of star formation (e.g. Barnes & Several studies have attempted to combat this issue by calculating Hernquist 1992; Sanders & Mirabel 1996). During merger events, the relative contribution of AGN and star formation in starburst– gaseous dissipation funnels material towards the centre of mass of AGN composite galaxies. Heckman et al. (2004) compute the AGN coalescing systems, leaving them with extremely large reservoirs of fractions for galaxies from the Sloan Digital Sky Survey (SDSS; gas. This gas rotates in the gravitational field of the galaxy, and can http://mnras.oxfordjournals.org/ York et al. 2000) Data Release 4 (DR4) by comparing the positions be converted into a rotating stellar disc by means of a starburst (e.g. of galaxies on the [N II]/Hα versus [O III]/Hβ diagnostic diagram to Barnes & Hernquist 1992; Mihos & Hernquist 1994, 1996) and/or synthetic composite galaxy templates with different AGN fractions. feed AGN accretion processes. Davies et al. (2007) estimate the contribution of the AGN to the line Evidence for the starburst–AGN composite nature of NGC 7130 emission of a sample of Seyfert galaxies by calculating the equiv- has been found in the IR, optical, ultraviolet (UV) and X-ray (Gon- alent width of a CO absorption feature and comparing this to the zalez Delgado et al. 1998;CidFernandesetal.2001; Gonzalez known value for globular clusters. Imanishi et al. (2011) constrain Delgado, Heckman & Leitherer 2001;Continietal.2002; Spinoglio, the contribution of the AGN in the infrared (IR) by analysing the Andreani & Malkan 2002;Levensonetal.2005). In this paper, we strength of polycyclic aromatic hydrocarbon (PAH) features in the use this galaxy to show that IFU data can be utilized to create spatial at The Australian National University on August 31, 2014 3–20 µm range. All of these methods assume that AGN fractions maps which allow a clean separation of regions of star formation, can be reliably calibrated across the AGN sequence. However, vari- AGN and composite ionization from both sources, and permit the ations in ionization parameter, metallicity, hardness of the radiation identification of outflows and/or feedback that may be present in the field and electron density across the active galaxy population in- galaxy. We demonstrate that there is a strong relationship between troduce significant uncertainty in the derived AGN fractions when the ionization state of the gas and distance from the centre of the information from multiple galaxies is combined (e.g.