; Galaxy formation and evolution
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

Galaxy formation and evolution

VIEWS: 52 PAGES: 38

  • pg 1
									                                      Image: Martin Hardcastle




         The Duty Cycle of Radio Galaxies
          in the context of AGN feedback




Stanislav Shabala
Paul Alexander
Summer Ash
Julia Riley
                 Background
• Growth of galaxies
                            Häring & Rix 2004




  and SMBHs linked
• Radio source - X-ray
  gas interaction
• Required to address
  key problems:                                 Baugh 2005


  – Cooling flow problem
  – Bright end of optical
    luminosity function
  – Cosmic downsizing
      Radio source feedback

• Jet formation
• Shock heating from expanding radio
  cocoon
• Buoyant bubbles
• Intermittency
  – Feedback self-regulated
  – Duty cycle
                Overview
• Constraining the local duty cycle
  – Radio/optical sample of local galaxies
  – Radio source models
  – How often do radio AGNs switch „on‟ and
    „off‟?
  – What triggers these processes?
• Radio source feedback in galaxy evolution
  – How important is it?
  – Can we reproduce observed properties?
                  Observed Sample
• Optical survey
   – Sloan Digital Sky Survey (SDSS) DR2
   – Volume/luminosity limited subsample with 0.005 < z ≤ 0.1 and Mr < -
     20.45
• Radio surveys
   – at 1.4-GHz
   – NRAO VLA Sky Survey (NVSS)
       • resolution of 45”, complete to 3.4 mJy
       • sensitive to more diffuse sources, but positions less accurate
   – Faint Images of the Radio Sky at Twenty Cm (FIRST)
       • resolution of 5”, complete to 1 mJy
       • capable of detecting fainter sources, more accurate positions
• Catalogue matching
   – Optically selected sample
   – 1191 radio loud sources (of 44,360 optical) with 0.03 < z ≤ 0.1
       • previous works only luminosity limited
   – AGNs and star forming regions
 AGN vs Star Forming Galaxies

• Optical vs radio AGN
  identification
   – Emission-line diagram
     (Baldwin+ 81)
      • [O III] 5007/Hβ – [N II]
        6583/Hα plane
   – 4000 Å break (Best+ 05)
      • Dn(4000) – Lradio/M* plane
• Adopt 4000 Å break
  (radio) diagnostic

• Independence of optical and radio AGN activity
   -cf. Rawlings & Saunders „91
Radio Luminosity Function

    Radio         Emission-line
  Bivariate Luminosity Function




• Radio sources re-triggered more frequently in
  massive hosts
       FR-II dynamical model
• Collimated jet terminates at hotspot and inflates
  cocoon due to backflow
• Overpressured cocoon expands
• Solution self-similar




               From Kaiser & Alexander (1997)
P-D tracks
         Source parameters
• Jet power and age from position in P-D
  plane
• Source size and luminosity from
  combination of FIRST/NVSS
  – Extended
    • Size from visual inspection
    • Integrated flux
  – Compact sources
    • Single-survey sources
    • FIRST flux if in both surveys
Source distribution in P-D plane
              Jet power




• Higher jet power for more massive hosts
        Source ages




• Massive galaxies host older radio sources
       Fitting the Bivariate LF




• Quiescent timescales from radio loud fraction
                  Timescales

log (M* / Msun)   log Qjet / W   ton,median / yr   toff / yr
    > 11.76         35.6 ± 1.0       5 × 106        2 × 107

  11.56 – 11.76     35.5 ± 0.7       4 × 106        5 × 107

  11.36 – 11.56     35.3 ± 0.7       3 × 106        1 × 108

  11.16 – 11.36     35.3 ± 0.7       1 × 106        1 × 108




• Radio sources are active for longer, and inactive
  for shorter periods in massive hosts
     Radio source intermittency
                                      
                                      M cool
• Cold gas availability         ton 
                                      Q jet
• Cooling rate      
                    M cool  M BH  M* (MBH – Mbulge relation)
                                 1.5  1.7




                 Q jet  M *
                               0.6
• Jet power

    ton  M *
              1.1



• Consistent with observations
 Fuel depletion causes AGN to “switch off” (Shabala+ 08)
• Shorter “off” times for massive hosts
            The story so far…
• Construct a flux- and volume-limited radio-
  optical sample
  – Sensitive to both compact and extended structure
• Emission-line and radio AGN activity
  independent
• Split up objects by stellar mass
  – Derive jet powers and ages
• Radio sources in more massive hosts are re-
  triggered more frequently and longer-lived
  – consistent with cold gas being required to fuel the
    AGN
   Galaxy formation and evolution
• Semi-analytic models
  – DM simulations + analytic baryonic prescriptions
  – Gas cooling and star formation
• Low-mass galaxies
  – Reionization and supernovae feedback
• Massive galaxies
  – AGN feedback
  – Radio source/ICM interaction complicated
     • Different “coupling efficiencies” of jet to ICM required for
       different galaxy masses (Best+ 07)
   radio source models should be considered
             Structure growth
• Average growth of DM halo
  – Given current mass
  – From fits to Millennium Simulation (van den Bosch „02)
  – Implicitly includes mergers
• Baryons accreted at virial radius along with DM
  – Follow DM
 Gas cooling and star formation
• Radiative cooling (Sutherland & Dopita „93)
• Cold gas transported to disk after a dynamical time
• Stars form from cold gas if there is sufficient cold gas in
  the disk (Kauffmann „96; Croton+ 06)
                      V  R 
   M crit  7.5 107  vir -1  vir  Msun
                                    
                      km s  kpc 
• Star formation rate is
               M  M crit 
   M *   SF  cold
                           
               t           
                 dyn, disk 
   – SF ~ 0.1
Feedback: reionization and SNe
• UV flux from first generation of stars and quasars
  – Limits baryonic fraction accreted onto halo (Gnedin
    „00)
• Instantaneous recycling approximation (De
  Lucia+ 04)
  – Fraction of stars immediately returned to cold gas
• Energy injection
  – Heating and ejection
  – Parametrised by halo
  – Gas returned within a few tdyn
               Black Hole growth
• Fraction acc of cold disk gas is accreted onto
  central BH
• Excess angular momentum removed by jet
• Accretion disk properties depend on rate of
  fuelling
   –    m  mcrit  standard thin disk (Shakura & Sunyaev
         
       „73)
        • Radiatively efficient (“quasar mode”)
        • Weak jet
   –     m  mcrit  Advection Dominated Accretion Flow
            
       (ADAF; Narayan & Yi „94)
        • Radiatively inefficient
        • Powerful radio jets (“radio mode”)
        Radio source feedback
• Thin disk  ADAF transition when
  instantaneous accretion rate falls below critical
  value
   – Actual accretion rate from cold gas reservoir is at the
     Eddington rate (Meier „01)
• Dynamical radio source model of Kaiser &
  Alexander „97
   – Jet inflates cocoon
   – Supersonic expansion and shock heating
     Radio source intermittency
• Jet intermittency
   – Weak jets disrupted in dense galaxy core (Alexander „02)
                                   2
                      R                     core      
     Q jet ,min  10  core 
                  37
                      2.5 kpc         1.51022 kg m -3  W
                                                         
                                                       

   – Powerful jets inflate cocoons that heat gas at large radii
    central cooling and change in accretion solution
    termination of jet
• ADAF  thin disk transition
     mcrit ,up ~  ADAF  0.1
                  2


   – Consistent with observations of X-ray binaries
• Thin disk  ADAF transition
   – Growing hole at centre of thin disk
   – Free parameter mcrit ,down
                      
            Model parameters

• Supernovae feedback: halo
• AGN feedback
  – Accretion onto BH: acc
  – Transition from thin disk to ADAF: mcrit ,down
                                       
• Constrain these using observational data
             Feedback modes




• Reionization and SNe feedback dominant for Mhalo < 1012 Msun
• AGN feedback at higher masses
        Supernovae feedback
• Stellar mass function
  – From luminosity function, assuming an Initial Mass
    Function (IMF)
  – Use “diet Salpeter” IMF (Bell+ 03)
     • Less low-mass stars
      lower mass-to-light ratio
     • cf. top-heavy Bottema „97 IMF
• Convolve stellar mass of a given z = 0 halo with
  halo mass function (Jenkins+ 01)
• Low-mass end constrains halo
Supernovae feedback




     Best fit halo ~ 0.01
                Black Hole accretion
•   MBH – Mbulge relation (Magorrian+ 98, Häring & Rix „04)
     – Fraction of stars in bulge from observations of local SDSS galaxies (Benson+ 07)




•   Best fit acc ~ 0.02
•   Expect BH and bulge to grow in step, even when AGN feedback is included
Onset of radio mode




   Best fit   
              mcrit ,down ~ 0.04
MBH – Mbulge relation
    Star Formation Rate Density




•   Good match to z ~ 1.5 - 2
     – Derived star formation rates are overestimates for z > 2
     – Switch to top-heavy IMF? (Baugh+ 05, Bower+ 06)
Evolution of the stellar mass function
     z=0          z = 0.5        z = 1.0




    z = 1.5       z = 2.0        z = 2.6
Top-heavy IMF at high z
   z = 1.0        z = 1.5




   z = 2.0        z = 2.6
             Cosmic downsizing
               AGN                             no AGN




• AGN feedback suppresses star formation in massive galaxies at the
  present epoch
       Galaxy bimodality




• Massive galaxies are red and dead
                         Summary
• Local massive galaxies host more powerful jets
   – active for longer, and spend less time in quiescent phase
• Local radio sources are fuelled by accretion of cooled
  hot gas
• Intermittent radio source feedback
   – Radio jet in ADAF phase
   – Jet terminated by instabilities or switch to thin disk phase
   – BH and spheroid grow in step
• Recover star formation histories to z ~ 2
   – High-mass end of the stellar mass function
   – Cosmic downsizing
   – Change to a top-heavy IMF at z > 2

								
To top