Kinetic Luminosity of Quasar Outflows and Its Implications to AGN Feedback Nahum Arav Virginia Tech Collaborators: Max Moe Jay Dunn Doug Edmonds Benoit Burguet Kirk Korista Chris Benn Manuel Bautista Eric Hallman Elisa Costantini Kentaro Aoki

AGN Feedback and cosmological structure formation

Credit: Andrey Kravtsov

Cluster cooling flows: Wu et al ;Ciotti & Ostriker 2001; Bower et al ; Vernaleo & Reynolds 2006; Rafferty et al. 2006; Thacker et al. 2006; Puchwein et al 2008 Soker et al 2010

Enrichment of the intergalactic and intracluster medium Furlanetto & Loeb 2001; Cavaliere 2002 Yuexing et al.; 2006; D'Odorico et al. 2006; Li, et al. 2007

Evolution of the host galaxy Scannapieco & Oh 2004; Di Matteo et al. 2005; Hopkins et al. 2005,2006; Springel et al. 2005; Menci et al. 2006; Haiman et al Somerville et al. 2008, et al. 2008)merville, et al. 2008)

Growth of super-massive black holes Silk & Rees 1998; Blandford & Begelman 1999, 2004; King 2003 ; Cattaneo et al. 2005; Hopkins et al. 2007; Ostriker et al. 2010

Self Regulation Gravitationally insignificant. However, the black hole binding energy can be larger than the galaxy’s! As soon as the central BH accretes large quantities of gas so as to significantly increase its mass, it releases large amounts of energy that would suppress further accretion onto it. In short, the BH growth is self-regulated. This energy may be sufficient to curtail the growth of the galaxy and to heat up the cluster gas. This is called AGN Feedback.

In some quasars we see evidence of mass ejection.

C IV 10,000 km/s O VI

Do quasar outflows contribute significantly to feedback processes? It will all depend on the: Mass flux and Kinetic luminosity of observed outflows.

AGN NHNH R v f

Kinetic luminosity of absorption outflows Up until recently Exceptions: De Kool + (3 objects); Hamann 3c191

kinetic luminosity of comp C in the SDSS outflow ergs/s = Solar masses/yr AGN feedback models need kinetic luminosity ~5%LBOL Less than 40% error

How do we go from the spectrum to measuring the kinetic luminosity?

From absorption troughs to kinetic luminosity Reliable measurements of N ion cannot use EW, tau ap Photoionization modeling to convert N ion to N H and U Distance of the Outflow from the Central Source:Distance of the Outflow from the Central Source: Number Density via Troughs from metastable levels Fe II* UV1, UV2…; Si II* 1264, 1533…

For full coverage 2587

Star disk: r~ cm QSO emission source: r~ cm SN remnant NGC 2736

assuming solar abundances For absolute abundances determination in AGN outflows see: Gabel + (2006); Arav + (2007)

Physical parameters of comp C in the SDSS outflow A clear violation of the momentum limit QSO

What about the solid angle subtended by the wind?

N(S IV*)/N(S IV)~

4.5<log(n H )<5.2 R~1kpc

S IV results Since S is 10 times less abundant than C, our results are consistent with a picture that all C IV BAL has S IV in the spectra, putting a good fraction of them at R~kpc scale. Imaging Knowing which outflows are at a few kpc scale should allow us to image them with HST at z~2

full What is the full kinetic luminosity of the SDSS outflow? Two more outflow components (highest velocity one is much closer) What about the hot phase? In Seyferts (where we can see it) the hot phase is dominant

AGN v V=4000 km/s observer cm - 3

Consequences for AGN feedback A 5000 km/s wind shocks the surrounding medium to ~10 9 K, with t cool ~t hubble, leading to very efficient PdV work. Over 10 8 years quasar duty cycle, such kinetic luminosity (5x10 45 ergs/s) will yield a total kinetic energy of ergs. Enough to inflate the largest observed X-ray bubbles.

Intracluster chemical enrichment (Eric Hellman, Harvard)

COS=FUSE+STIS on Turbo Increased sensitivity allows us to choose the sample for our experiment rather than the experiment for the sample The wonderful diagnostic filled spectral region short ward of 1000A

Summary Quasar outflows are a major component of AGN feedback, reaching kinetic luminosities of a few percent of L BOL, with mass flux of hundreds of solar masses per year. Due to their larger opening angle and higher mass fluxes, absorption outflows may be more efficient for AGN feedback processes than AGN jets.