Gas outflows and star formation in type 2 AGNs - AGN feedback in action? Jong-Hak Woo (Seoul National Univ.) & Hyun-Jin Bae, Donghoon Son, Marios Karouzos Muse view of OIII outflows in NGC 7582 (Juneau+16 in prep.) OIII gas kinematics (Woo+16a)
Gas outflows AGN feedback? How common? Ionized gas outflows detected over galaxy scales (~1-10 kpc). (e.g., Nesvadha+07, Harrison+14, Karouzos+16a, Husemann+16). Suppress or trigger SF? AGN feedback mechanism? (e.g., Greene+12, Liu+13, Cresci+15, Vilar-Martin+16, Karouzos+16b).
Gas outflows AGN feedback? How common? Ionized gas outflows detected over galaxy scales (~1-10 kpc). (e.g., Nesvadha+07, Harrison+14, Karouzos+16a, Husemann+16). Suppress or trigger SF? AGN feedback mechanism? (e.g., Greene+12, Liu+13, Cresci+15, Vilar-Martin+16, Karouzos+16b). Important to investigate outflows in BH-galaxy evolution The M-sigma relation of active and inactive galaxies (e.g., Woo et al. 2015) σ * (km/s)
Gas outflows AGN feedback? How common? Ionized gas outflows detected over galaxy scales (~1-10 kpc). (e.g., Nesvadha+07, Harrison+14, Karouzos+16a, Husemann+16). Suppress or trigger SF? AGN feedback mechanism? (e.g., Greene+12, Liu+13, Cresci+15, Vilar-Martin+16, Karouzos+16b). Important to investigate outflows in BH-galaxy evolution Outflow demography based on a large sample is needed to understand outflows as a population (e.g., Boroson+05, Komossa 08, Zhang+11, Crenshaw+10, Wang+11, Mullaney+13).
Gas outflow studies at SNU 1. Outflow census: how strong, how common, relation to AGN energetics (Bae & Woo 14, Woo+16a) 2. Outflow model simulations: intrinsic properties of outflows - Talk by H.-J. Bae (Bae & Woo 16) 3. Connection to star formation: positive or negative feedback? (Woo+16b in prep.) 4. IFU follow-ups of AGNs with/without strong outflows (Karouzos+16a, Karouzos+16b, Bae+16b in prep.)
Statistical sample of AGNs and SF galaxies at z<0.3 We selected AGNs and SF galaxies with well-defined emission lines from SDSS (A/N>5). type N pure AGNs ~23,000 composite obj. ~16,000 SF galaxies ~69,000 For each AGN, we subtract stellar population model, and measure systemic velocity and stellar velocity dispersion (σ * ). Flux density 140 50 20 120 10 100 80 60 40 40 30 0 485048554860486548704875 00016 (z=25159142) 120 100 8 80 6 60 4 40 2 20 0 0 498049905000501050205030 628062906300631063206330 200 150 100 50 0 6540 6560 6580 6600 20 0 4000 5000 6000 7000 Rest-frame wavelength
[OIII] λ5007 traces ionized gas outflows. ~44% of AGNs have a broad wing in OIII, representing outflows. - Outflows are common. Hβ [OIII] We measure flux-weighted velocity (1 st moment), then calculate vel. shift (V OIII ) w.r.t systemic velocity. We measure vel. dispersion (σ OIII 2 nd moment), and compare with stellar VD. flux (10-16 erg/s/cm 2 /A ) H +[NII] [SII] wavelength (A )
OIII VD (σ OIII ) correlates with AGN luminosity. OIII kinematics are related to AGN energetics. For SF galaxies, no strong trend between L OIII and σ OIII Woo+16b (to be submitted)
σ OIII is larger than σ *, indicating non-virial motion. σ OIII correlates with stellar VD (σ * ), indicating galaxy gravitational potential plays a role (e.g., Nelson & Whittle 1995). σ OIII > σ *, indicating presence of non-gravitational kinematics (i.e., outflows). stellar VD log σ * (km/s) Woo+16a,b
Non-grav. component is comparable to grav. component For AGNs without OIII wing, gas and stellar kinematics are similar. For AGN with OIII wing, σ OIII is larger than σ * by a factor of 1.3-1.4, indicating σ non-gr ~ σ gr. single Gaussian [O III] double Gaussian [O III] < log (σ [O III] /σ * ) > 0.1 0.1 Seyfert only AGN AGN + composite 38 39 40 41 42 log L [O III] (erg s 1 ) σ OIII 2 = σ non-gr 2 + σ gr 2 = σ non-gr 2 + σ * 2 Woo+16a
Outflow fraction dramatically increases with AGN energetics Fraction of AGN with OIII wing dramatically increases with L and L/ L Edd, indicating that outflows are prevalent among high-l AGNs. Fraction of double Gaussian [O III] Seyfert only AGN AGN + composite Woo+16a 39 40 41 42 43 3 2 1 log L [O III] (erg s 1 ) log (L bol /L Edd )
Gas outflows are prevalent in AGNs. σ [O III] /σ * > detectable outflows Over all luminosity range, at least 50% of AGNs have outflows (i.e., σ OIII > σ * ). ~50% of high-l AGNs have extreme outflows (i.e., σ OIII > 1.4 σ * ). Fraction σ [O III] /σ * > 1.4 Seyfert only AGN AGN + composite extreme outflows The majority of high-l AGNs show strong outflow signature (see also Boroson+05, Crenshaw+10, Wang+11, Mullaney+13). 38 39 40 41 42 3 2 1 log L [O III] (erg s 1 ) log (L bol /L Edd ) Woo+16a
VVD (vel.-vel. dispersion) diagram reveals extreme gas kinematics VVD distribution is clearly different between AGNs and SF galaxies. σoiii (km/s) VOIII (km/s) Woo+16b in prep.
Luminosity dependence of VVD distribution VVD distribution expands with AGN luminosity. Outflow kinematics are connected to AGN radiation (AGN-driven). σoiii (km/s) VOIII (km/s) Woo+16b
Gas outflows do not depend on radio luminosity σ OIII increases with L radio, reflecting grav. potential. However, once normalized by σ *, there is no correlation. For most AGNs, ionized gas outflows are not directly related to radio activity. V OIII shows similar trend. Woo+16a V OIII (km/s) log σ OIII / σ * log σ OIII (km/s) log L 1.4GHz (erg/s) log (L 1.4GHz /L opt )
3-D outflow simulations (talk by H.-J. Bae) Bi-cone + dust disk (intrinsic velocity, opening angle, inclination of the cone, inclination of dust disk) Projected 2-D V OIII & σ OIII maps (calculated through each line of sight) Calculate flux-weighted V OIII & σ OIII Log flux velocity velocity dispersion Dust disk: -60 degree Bae & Woo 16 ApJ, in press
MC simulations of VVD diagram (talk by H.-J. Bae) Using random distribution of orientation angle of the bicone, dust plane, and intrinsic velocity, we simulated the VVD diagram. Dust extinction plays an important role. The intrinsic velocity ranges from ~200 to ~2,000 km/sec. Bae & Woo 16, ApJ, in press
AGN have lower SFR compared to SF galaxies (optical & X-ray) AGNs have lower SSFR than main sequence galaxies. AGNs feedback? Evolutionary sequence? stellar mass (M sun ) Shimizu+15 Ellison+16
SSFR correlates with outflow kinematics AGNs with outflow signatures have similar SSFR compared to SF galaxies. AGNs with no outflows have much lower SSFR. Woo+16b to be submitted
AGNs with no outflow have lower SSFR. log σ [O III] /σ * > 0.3 strong outflows All mass SF galaxies AGNs with strong outflows have similar SSFR compared to SF galaxies. No suppression, no enhancement. AGNs with no outflows show much lower SSFR. Woo+16b to be submitted Normalized number < log σ [O III] /σ * < 0.3 weak outflows log σ [O III] /σ * < no outflows -2.0-1.5 - -0.5 0.5 log SSFR (yr -1 )
All mass Low-Eddington AGNs have lower SSFR log L bol /L Edd > -2 high-edd. SF galaxies High-Eddington AGNs have similar SSFR compared to SF galaxies. Low-Eddington AGNs have much lower SSFR. Normalized number -3 < log L bol /L Edd < -2 Intermed. mid-ir selected AGNs = high low-edd. log L bol /L Edd < -3 Eddington AGNs Woo+16b to be submitted -2.0-1.5 - -0.5 0.5 log SSFR (yr -1 )
Delayed AGN feedback or gas depletion? high Eddington + strong outflow + MS SFR vs. low Eddington + no/weak outflow + low SFR All mass log σ [O III] /σ * > 0.3 All mass log L bol /L Edd > -2 Delayed AGN feedback? 1. Gas supply: AGN and SF coexist (Netzer 09) 2. Outflows impact on ISM after dynamical time scale (10 6-7 yrs) 3. Decrease SFR and AGN activity (low Eddington, no outflows) Normalized number < log σ [O III] /σ * < 0.3 log σ [O III] /σ * < Normalized number -3 < log L bol /L Edd < -2 log L bol /L Edd < -3 Transition due to gas depletion -2.0-1.5 - -0.5 0.5 log SSFR (yr -1 ) -2.0-1.5 - -0.5 0.5 log SSFR (yr -1 )
Gemini GMOS-IFU Follow up Hα OIII (Karouzos+16a;16b, submitted) OIII shows outflow kinematics while Hα shows rotation.
Gemini GMOS-IFU Follow up Hα OIII (Karouzos+16a;16b, submitted) SF ring is present at the edge of outflow region.
Summary Gas outflows are prevalent among local AGNs, based on 1) the presence of a broad wing component in OIII, and 2) based on the non-gravitational component (σ OIII > σ * ). In particular, the majority of high-l AGNs show strong outflows. OIII vel. & vel. disp. increase with AGN luminosity, suggesting that outflows are linked to AGN energetics. Ionized gas outflows are not related to radio activity. AGNs with strong outflows follow the SF main sequence while AGNs with no outflows have much lower SSFR, implying delayed AGN feedback.