ADVECTION-DOMINATED ACCRETION AND THE BLACK HOLE EVENT HORIZON Ramesh Narayan

Energy Equation Accreting gas is heated by viscosity (q + ) and cooled by radiation (q - ). Any excess heat is stored in the gas and transported with the flow. This represents “advection” of energy (q adv ), or “advective cooling” Per Unit Volume

Energy Equation q + = q - + q adv Thin Accretion Disk (Shakura & Sunyaev 1973; Novikov & Thorne 1973;…) Most of the viscous heat energy is radiated Advection-Dominated Accretion Flow (ADAF) (Narayan & Yi 1994, 1995ab; Abramowicz et al. 1995; Chen et al. 1995;…) Most of the heat energy is advected with the gas

Two Kinds of ADAFs Advection dominates under two conditions Radiation is trapped in the gas and cannot diffuse out before gas falls into the BH. “Slim Disk” solution (Abramowicz, Czerny, Lasota & Szuszkiewicz 1988). L  L Edd Gas is very dilute and cannot radiate its thermal energy before it falls into the BH. Radiatively Inefficient -- “RIAF” (Ichimaru 1977; NY 1994,1995; Abramowicz et al. 1995). L  ( )L Edd

Properties of ADAFs/RIAFs: 1 Very hot: T i ~ K/r, T e ~ K (virial, since ADAF loses very little heat) Large pressure: c s ~ v K Geometrically thick: H/R ~ 1 Optically thin (because of low density) Expect Comptonized spectrum: kT  100 keV It is a stable solution Explains low hard state of XRBs Esin et al. (1998,2001)

Properties of ADAFs/RIAFs: 2 Thin disk to ADAF/RIAF boundary occurs at Mdot crit ~ 0.01—0.1 Mdot Edd (for reasonable  ~ 0.1) Location of the boundary is nicely consistent with L acc at which: BH XRBs switch from the high soft state to the low hard state (Esin et al. 1997) AGN switch from quasar mode to LINER mode (Lasota et al. 1996; Quataert et al. 1999; Yuan & Narayan 2004) Yuan & Narayan (2004)

BH Accretion Paradigm: Thin Disk + ADAF Narayan (1996); Esin et al. (1997) Slim Disk State? (Kubota, Makishima) Narayan & Quataert (2005) Slim Disk RIAF

Properties of ADAFs/RIAFs: 3 By definition, an ADAF has low radiative efficiency Roughly, we expect a scaling (Narayan & Yi 1995) Extreme inefficiency of Sgr A* and other quiescent SMBHs is explained (N, Yi & Mahadevan 1995) Quiescent XRBs explained (N, McClintock & Yi 1996; N, Barret & McClintock 1997) Narayan & Yi (1995) Esin et al. (1997)

Properties of ADAFs: 4 Winds and Jets Narayan & Yi (1994, Abstract): … the Bernoulli parameter is positive, implying that advection- dominated flows are susceptible to producing outflows … We suggest that advection-dominated accretion may provide an explanation for … the widespread occurrence of outflows and jets in accreting systems Narayan & Yi (1995, Title): “Advection-Dominated Accretion: Self-Similarity and Bipolar Outflows” Strong outflows confirmed in numerical simulations ADAFs  JETS, WINDS

Recent Developments Fender, Belloni & Gallo (2003): paradigm on accretion flows and jets steady jets found in hard state hysteresis SMBH accretion and galaxy formation Effect of “feedback” on galaxy formn & SMBH growth “radio mode” of accretion It all comes down to ADAFs-outflows-jets

BH Accretion Paradigm: Thin Disk + ADAF + Jet Narayan 1996; Esin et al. (1997)Fender, Belloni & Gallo (2003) ADAF model provides theoretical underpinning for jet paradigm Hysteresis in low-high-low state transitions not yet understood

Jean-Pierre, We Need You!! ADAFs have always been strongly attacked Now ADAFs are being forgotten Things were okay so long as Jean-Pierre was our spokesman ! Then he lost interest in ADAFs … Now, the ADAF-Bashers are running wild The ADAF clan is getting absolutely killed Jean-Pierre, please come back!

Are Black Hole Candidates Really Black Holes? We know that BH candidates are Compact: R  few R S Massive: M  3M  (not neutron stars) But how sure are we that they are really BHs? Can we find independent evidence that our BH candidates actually possess Event Horizons ? This is a basic and important question

Accretion and the Event Horizon Accretion flows are very useful, since inflowing gas reaches the center and “senses” the nature of the central object X-ray binaries have an additional advantage --- we can compare NS and BH systems

Signatures of the Event Horizon Differences in quiescent luminosities of XRBs (Narayan, Garcia & McClintock 1997; Garcia et al. 2001; McClintock et al. 2003;…) Differences in variability power spectra of XRBs (Sunyaev & Revnivtsev 2000) Differences in Type I X-ray bursts between NSXRBs and BHXRBs (Narayan & Heyl 2002; Tournear et al. 2003; Remillard et al. 2006) Differences in X-ray colors of XRBs (Done & Gierlinsky 2003) Differences in thermal surface emission of NSXRBs and BHXRBs (McClintock, Narayan & Rybicki 2004) IR flux of Sgr A* (Broderick & Narayan 2006, 2007)

Basic Idea Accretion releases energy ~(GM/R) per gram accreted Typically, 50% is released in the disk, L disk ~0.1(Mdot)c 2, and 50% at the stellar surface, L star ~0.1(Mdot)c 2 For a given Mdot, predicts some difference in luminosity between a NS and a BH Usually, luminosity difference is modest (order unity) L disk L star

Enter ADAFs ! A low-mdot ADAF/RIAF system has L disk  0.1(Mdot)c 2 Most of the gravitational energy is stored in the gas as thermal energy and released only when the gas hits the stellar surface With stellar surface: L = L disk + L star ~ 0.2 (Mdot)c 2 With event horizon: L = L disk  (Mdot)c 2 Expect huge deficit in L :- L  (Mdot)c 2 – robust test possible But we need an independent estimate of Mdot  NS control sample L disk L star

Surface

The First Study Narayan, Garcia & McClintock (1997) Considered NS and BH XRB transients in quiescence -- very sub-Eddington accn Found evidence for a large luminosity difference But only a few systems … Much better evidence now

Better Way to Plot the Data Our original idea was too simple – we just compared luminosity swings But different SXTs will have different mdot values in quiescence Better to plot Eddington-scaled quiescent luminosities vs orbital period Lasota & Hameury (1998) Menou et al. (1999)

Transient XRBs in quiescence have ADAFs (N, M & Yi 96) Binary period P orb determines Mdot in quiescence (Lasota & Hameury 1998; Menou et al. 1999) At each P orb, we see that L/L Edd is much lower for BH systems. True also for raw L values. (Garcia et al. 2001; McClintock et al. 2003; …)

Transient XRBs in quiescence have ADAFs (N, M & Yi 96) Binary period P orb determines Mdot in quiescence (Lasota & Hameury 1998; Menou et al. 1999) At each P orb, we see that L/L Edd is much lower for BH systems. True also for raw L values. (Garcia et al. 2001; McClintock et al. 2003; …)

Independent Confirmation Radiation from the surface of a star is expected to be thermal X-ray spectra of BH XRBs in quiescence have power-law shape We can set a stringent limit on thermal component in the BH system XTE J (McClintock et al. 2004)  no surface McClintock et al. (2003)

Can Strong Gravity Provide a Loophole? Suppose our BHs do have surfaces, but at a radius VERY SLIGHTLY outside the horizon (gravastar, dark energy star): R star =R S +  R Extreme relativistic effects are expected: Radiation may take forever to get out Surface emission may be redshifted away Emission may be in particles, not radiation Surface may not have reached steady state None of these can explain the observations

Takes Forever for Signals to Get Out In terms of time as measured by an observer at infinity, the world line of infalling matter never crosses the horizon, and Signals emitted by the matter take “forever” to get out

But How Much Extra Delay? The extra delay relative to the Newtonian case is TINY At most it is 10 ms (for  R ~ Planck scale) --- no big deal

Gravitational Redshift Looks serious, especially if redshift is large But energy has to be conserved A calculation shows that L loc exceeds Mdot c 2 by precisely a factor (1+z) 2 such that L  = Mdot c 2

Avery Broderick’s Argument L L/(1+z) 2 L L

Summary ADAFs are found all over the place >99% of BHs in the universe have ADAFs ! Strong connection between ADAFs and Jets ADAFs provide compelling evidence for the existence of BH Event Horizons Jean-Pierre played a major role in all this