Unveiling a Supermassive Black Hole at the Center of Our Galaxy Andrea Ghez University of California Los Angeles Collaborators (UCLA/Caltech/Keck) E. E. Becklin, G. Duchene, S. Hornstein, D. Le Mignant, J. Lackey/Lu, K. Matthews, M. Milosavljevic , M. Morris, S. Samir, B. T. Soifer, A.Tanner, D. Thompson, N. Weinberg, S. Wright Image courtesy of 2MASS Key Questions Is there a supermassive black hole at the center of our Galaxy? Is it associated with the unusual radio source Sgr A*? Why is it so dim (10-9 LEd)? What is the distance to the Galactic center (Ro) Is there a halo of dark matter surrounding the black hole? When and where are the stars born? Does the black hole influence the appearance / evolution of the stars? Original Case of Central Black Holes Active Galactic Nuclei (AGN) Emit energy at an enormous rate Radiation unlike that normally produced by stars or gas Variable on short time scales Contain gas moving at extremely high speeds CENTRAL ACCRETING BLACK HOLES Cyg A Jets ~105 pc (galaxy 1/10 this size) Do “normal” (non-active) galaxies have “quiet” black holes? Milky Way is Best Place to Answer this Question Pro - Closer (8 kpc) Con - Obstructed View (dust) Optical light: 1 out of every 10 billion photons emitted makes it to us (invisible!) Near Infrared light: 1 out of every 10 photons emitted makes it to us (visible!) Overview of Galactic Center (~100 pc) From Genzel 1994 Overview of Galactic Center (~ 10 pc) SgrA* CircumNuclear Disk (CND) Sgr A* Is unusual radio source Sgr A* coincident with black hole? Non-thermal emission Compact Low-velocity VLA: J.-H. Zhao Dynamical Proof of Black Hole Need to show mass confined to a small volume Rsh = 3 x MBH km (MBH in units of Msun) Use gas/stars as test particles BH F = -G Mencl m/ R II. Stellar Cluster (r a r -2) I. Black Hole (Velocity Dispersion)1/2 a r-1/2 (Velocity Dispersion)1/2 r Enclosed Mass r stars BH r Enclosed Mass stars r Gas Radial Velocity Measurements Gave 1st Hint of Dark Matter Contribution from Luminous Matter Evidence for Dark Matter • HI rotation along Galactic Plane Plot from Genzel 1994 VLA 6 cm image of mini-spiral` (eg. Rougoor & Oort 1960; Ooort 1977; Sinha 1978) • Circumnuclear disk/ring rotation (e.g., Gatley et al. 1986; Guesten et al. 1987) • Ionized streamers in mini-spiral (e.g., Serabyn & Lacy 1985; Serabyn et al. 1987) Dark Matter Confirmed with Stellar Radial Velocity Measurements Contribution from Luminous Matter Evidence for Dark Matter • Integrated stellar light (e.g., McGinn et al. 1989; Sellgren et al. 1990) • Individual Stars (OH/IR, giants, He I) (e.g., Linquist et al. 1992; Haller et al. 1995; Genzel et al. 1996) However, Inferred Dark Matter Density was too Small to Definitively Claim a Black Hole Black Hole Alternatives Clusters of dark objects permitted with the inferred density of ~109 Mo/pc3 Fermion Ball High spatial resolution techniques needed to make further progress. 6” Two Independent High Resolution Imaging Studies Keck (10-meter) 1995 - present 0.”045 Ghez et al. 1998, 2000 Gezari et al. 2002 Tanner et al. 2002 Hornstein et al 2002 Ghez et al. 2003a,b,c Keck Telescopes on Mauna Kea Hawaii NTT La Silla NTT (3.6-meter) 1992 - 2001 0.”15 Eckart & Genzel 1996, 2002 Genzel et al 1997, 2000 VLT (8-meter) 2002 - present 0.”056 Schodel et al. 2002, 2003 Eisenhauer et al. 2003 Genzel et al. 2003a,b VLT Atacama, Chile Diffraction-Limited Images Have Been Obtained with 2 Methods: Speckle & Adaptive Optics (AO) Light from science target Light from reference star a dna ™emiTkciuQ rosserpmoced FIG .erutcip siht ees ot dedeen era Beam Splitter Deformable Mirror Science Camera Computer Wavefront Wavefront sensor Sensor…. AO allows deeper images & spectra! The Shack-Hartmann Wavefront Sensor Lenslet Array Subaperture Focal Spots Incoming Wave 2-Dimensional Detector Spot Deviation Run at 100-500 Hz & deformable mirror has ~300 segments Tremendous Progress Has Been Made With High Angular Resolution Techniques on Large Telescopes 6 QuickTime™ and a GIF decompressor are needed to see this picture. Motions on the Plane of the Sky Easily Measured 200 stars tracked, only central 1”x1” shown 1" DEC RA Proper Motion Measurements Increased Dark Matter Density (x103), Which Ruled Out Clusters of Dark Objects ~1 milli-arcsec astrometric accuracy Eckart & Genzel 1997 & Ghez et al. 1998 (shown) Black Hole Case Strengthened by Acceleration Measurements Ghez et al. 2000 (shown), Eckart et al. 2002 Accelerations provided first measurement of dark mass density that is independent of projection effects r = 3 a2-d / (4 G R2-d3) Dark mass density increased by 10x (~ 1013 Mo/pc3) leaving only fermion balls as BH alternative. Center of attraction coincident with Sgr A* (±30 mas) Minimum orbital period of 15 yrs for S0-2 inferred Proper Motions Now Permit Complete Astrometric Orbital Solutions QuickTime™ and a GIF decompressor are needed to see this picture. 1" Orbits Increase Dark Mass Density By x104, Making Black Hole Hypothesis Hard to Escape S0-16 has smallest periapse passage Rmin = 90 AU = 1,000 Rs * Dark Mass Density Velocities: 1012 Mo/pc3 Accelerations: 1013 Mo/pc3 Orbits: 1017 Mo/pc3 * Fermion ball hypothesis no longer works as an alternative for all supermassive black holes m ~ 50kev c-2 Mass fermion ball < 2x108 Mo Ghez et al. 2002, 2003 (shown); Schoedel et al. 2002, 2003 Independent solutions for 3 stars (those that have gone through periapse) * Milky Way is now the best example of a normal galaxy containing a supermassive Simultaneous Orbital Solution is More Powerful than Independent Orbital Solutions S0-2 Improves Estimate of Black Hole’s Properties S0-16 S0-19 Mass: 3.7±0.4 x 106 (Ro/8kpc)3 Mo Position: ±1.5 mas Adds Estimate Black Hole’s Velocity on the Plane of the Sky Velocity: 30 ±30 km/s Orbits Improve Localization of Black Hole in IR Reference by an Order of Magnitude, Assisting Searches for IR Emission Associated with Black Hole SiO masers used to locate Sgr A* position in IR frame (±10 milli-arcsec) Reid et al. 2003 IRS 7 IRS 10ee Sgr A* 0.1” Dynamical Center pinpointed to ±1.5 milli-arcsec (12 AU) 1” At 3.8 mm, Stellar and Dust Emission are Suppressed, Facilitating the Detection of Sgr A* QuickTime™ and a GIF decompressor are needed to see this picture. Keck AO L’(3.8 mm) images (Ghez et al. 2003, ApJLett, in press, astro-ph/0309076) NIR results fromVLT (Genzel et al. 2003, Nature) Factor of 4 Intensity Change Over 1 week and Factor of 2 Change in 40 minutes Similarity of Flaring Time-scales Suggests IR and X-ray Originate From Same Mechanism Chandra / Baganoff et al. 2001 Flaring from non-thermal tail of high energy electrons Models Physical Process Markoff et al 2001 Yuan et al. 2003 Shocks Magnetic reconnection Emission Mechanism IR Synchrotron X-Ray Self-Synchrotron Compton or synchrotron IR variability suggests electrons are accelerated much more frequently than previously thought Simultaneous Orbital Solution Allows a Larger Number of Orbits to be Determined Black hole’s properties fixed by S0-2, S0-16, & S0-19 M, Xo, Yo, Vx, Vy Less curvature needed for full orbital solution for other stars P, To, e, i, w, W Need only 6 kinematic variables measured (Rx, Ry, Vx, Vy, Ax , Ay) Eccentricities Are Consistent with an Isotropic Distribution While there are many highly eccentric systems measured, there is a selection effect We only measure orbits for stars with detectable acceleration (> 2 mas/yr2) Lower Limit on Semi-Major Axis > ~1000 AU Apoapse Distance > ~2000 AU No selection effect against detecting K<16 mag with A<1000 AU Possible Bias in Distribution of Apoapse Directions Other angle - inclination - appears random With Only Imaging Data, StellarType (age/mass) is Degenerate Based on 2 mm brightness (K = 13.9 to 17; Mk = -3.8 to -0.9) two expected possibilities Late-Type (G/K) Giant (cool & large; old & low mass) Early-Type (O/B) Dwarf / Main-Sequence Star (hot & small; young & high mass) Stellar-Type Degeneracy Easily Broken with Spectroscopy Late-Type (G/K) Giant Deep Carbon Monoxide (CO) absorption lines Early-Type (O/B) Dwarf Weak Hydrogen ( Brg) absorption lines Weak Helium (He) absorption lines Local Gas Makes it Difficult to Detect Weak Brg, Unless Star has Large Doppler Shift Local Gas S0-2 S0-1 1” Local Gas has strong Brg emission lines Effects ability to detect stellar Brg absorption lines if |Vz| < ~300 km/s For OB stars these are the strongest lines, which are already quite weak ~a few Angstroms For low Vz sources, lack of CO is evidence that they are young Brg in OB Stars in Sgr A* Cluster Detected as They Go Through Closest Approach Example of S0-2: Vz = +1100 to -1500 km/sec EW(Br g) = 3 Ang EW (HeI) = 1 Ang Vrot = 170 km/sec Digression: Addition of Spectra Also Provide a Direct Measure of Galactic Center Distance (Ro) NTT/VLT Keck Keck VLT Digression: Ro is now largest source of mass (spin…) uncertainty Ghez et al 2003 (Keck) Eisenhauer et al. 2003 (NTT/VLT) 1, 2, 3s contours The Majority of Stars in the Sgr A* Cluster are Identified as OB Stars Through Their Lack of CO Lack Individual spectra: Gezari et al. 2002 (shown, R=2,000), Lu et al (2004) Genzel et al. 1997 (R=35) Integrated spectra: Eckart et al 1999 & Figer et al. 2000 Presence of OB Stars Raises Paradox of Youth OB stars Have hot photospheres (~30,000 K) Are young (<~10 Myr) & massive (~15 Mo), assuming that they are unaltered by environment The Problem • Existing gas in region occupied by Sgr A* cluster is far from being sufficiently dense for self-gravity to overcome the strong tidal forces from the central black hole. Black Hole Are These Old Stars Masquerading as Youths? Possible Forms of “Astronomical Botox” Need to make stellar photosphere hot Heated (tidally?) by black hole (e.g., Alexander & Morris 2003) • No significant intensity variations as stars go through periapse Stripped giants (e.g., Davies et al. 1998) Accreting compact objects (e.g., Morris 1993) Merger products (e.g., Lee 1994, Genzel et al. 2003) Are Stars Young & Formed In-Situ? Past Gas Densities Would Have to Have Been Much Higher What densities are needed? ~1014 cm-3 at R= 0.01 pc (apoapse distance of S0-2) Mechanism for enhancing past gas densities Accretion disk (e.g., Levin & Beloborodov 2003) Colliding cloud clumps (e.g., Morris 1993, Genzel et al. 2003) Are Stars Young, Formed at Larger Radii, & Efficiently Migrated Inwards? HST/Figer At larger radii, tidal forces compared to gas densities are no longer a problem At 30 pc, young stellar clusters observed Arches and Quintuplet (e.g., Figer et al. 2000, Cotera et al. 1999) Massive (104 M ) & Compact (0.2 pc) Migration Inwards is Difficult, Due to Short Time-scales & Large Distances Ideas Massive binaries on radial orbits experience three body exchange with central black hole (Gould & Quillen 2003) Cluster migration (Gerhard et al. 2000, Kim & Morris 2003, Portegies-Zwart et al 2003, McMillan et al. 2003) Need very central condensed cluster core Variation on cluster migration - clusters with intermediate mass black holes, which scatter young stars inward (Hansen & Milosavljevic 2003) From New Scientist Only Cluster Shuttled Inward with Intermediate Black Hole Reproduces Orbital Properties, but Where are They? Distribution of Semi-major Axes Directions of Apoapse Vectors Orbital limit on reflex motion (< 30 km/s) limits IMBH to 2x105 (R / 16,000 AU)1/2 Mo Conclusions Dramatically improved case for black hole QuickTime™ and a GIF decompressor are needed to see this picture. First detection of IR emission from accreting material Central 1”x 1” The Future More orbits (# ~ t3) Ro to 1% (may allow a recalibration cosmic scale distance ladder) Deviations for Keleperian orbits! Dark matter density increased to 1017 Mo/pc3 with orbits, making the Milky Way the best example of a normal galaxy containing a supermassive black hole More variable than X-ray If from non-thermal tail of e-, shocks/reconnections happening more frequently than previously thought Direct measure of distance to GC (Ro) Raised paradox of youth Majority of stars in Sgr A* cluster appear to be young Low present-day gas densities & large tidal forces present a significant challenge for star formation (none of present theories entirely satifactory) Dynamical insight from orbits A Few Introductions Are Necessary Units Why did people think there might be a black hole at the center of our Galaxy? Galactic center environment Astronomical Units Distances & Angles Angles: • 1 arc-second [”]= 1/3600 degree • Atmosphere limits angular resolution of most observations to 1” R=1AU 1 pc = 206265 AU ~ 3 x 1013 km ~ 3 light years D=1pc Distances: • Astronomical Unit (AU) = Earth-Sun distance • 1 parsec [pc] = distance at which 1 AU subtends 1” Distance to Galactic Center = 8,000 pc Q=1” Sgr A* Cluster Stars Amplifying a Problem Originally Raised by the He I Emission Line Stars He I Emission-Line Stars Massive (20-100 Mo) post-mainsequence stars formed within the last 8 Myrs Located at distances from the black hole of 0.1 - 0.5 pc, which is 10x further than the Sgr A* cluster stars Formation problem Required gas densities are not as severe, but still not found at 0.1 pc Bright He I emission-line stars OB stars in Sgr A* cluster