Decoding The Rosetta Stone of Galaxy Formation The Milky Way with SDSS and Future Large Scale Surveys Mario Juric Institute for Advanced Study, Princeton with Zeljko Ivezic, Nick Bond, Brani Sesar, Robert Lupton… and the SDSS Collaboration Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Big Picture :: Galaxy and Structure Formation on ≤ MW scales • Mapping with SDSS • Taking the measure of the Milky Way • Realizing the dream of near-field cosmology Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH LCDM Structure Formation I. Known knowns: Structure formation on large scales Hierarhical s. f. Abundant substructure … II. Known unknowns: Small scale structure formation Baryon behavior Disk formation Disk survival Nature of DM … The Galaxy: Laboratory for II. -> I. A complete cosmo+galform theory must predict its properties Bullock et al. Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Milky Way • Thin disk Thick disk (Pseudo)bulge Stellar halo Exponential or sech2 disk models Power laws or de Vaucouleurs profiles for the halo Components trace the DM dominated potential They are a product of Milky Way formation and evolution Dissecting the Milky Way with SDSS Program 1. Directly measure the distribution ©(~ r;~ v; [F e=H ]) of stellar number density, kinematics, and metallicity in a representative volume of the Galaxy. 2. Use the distributions to learn about [Gg]alaxy formation, evolution, interactions with environment, and the distribution of dark matter. Sloan Digital Sky Survey Imaging and Spectroscopic Survey ~8,000 deg2 to ~21.5 mag 5 bands (ugriz: UV-IR), 0.02 mag < 0.1 arcsec absolute astrometry ~50M, mostly main sequence, stars R=2000 spectrograph (390<λ/nm<600) RV to ~10 km/s Stellar parameters for for >280k stars SEGUE I/II SDSS DR6 Imaging Sky Coverage (Adelman-McCarthy et al. 2008, ApJS, 175, 297) An excellent tool for Galactic structure studies Accurate m’band photometry: distance and metallicity estimates Accurate astrometry: proper motions Large area and faint flux limit: representative volume Numerous (MS) stars: reduced uncertainties Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH How we do it ~3M ~150M obsvs. u, g, r [Fe/H] Teff g, r, i a, d (circa 2002) a, d (circa 1955) Mr (absolute magnitude) ~50M X, Y, Z (position) ma, md (proper motion) POSS (USNO-B catalog) Mapping the Milky Way with SDSS ~30M Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Volume limited 3D distributions of r, [Fe/H], ml, mb in 19 r-i color bins (spectral types ~F8-M5) ~8000 deg2 E.g: r-i=0.1-0.15 (late F) 20 kpc (density) 8 kpc (Fe/H, m) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Dissecting the Data Cube R Z Y X Y Z R Mapping the Milky Way with SDSS X Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Density Maps: D < 3kpc (Late K, M stars) Right: X-Y maps of number density distribution at Z=+/900 and +/-600 pc for 1 < r-i < 1.1 stars (~M dwarfs) “Face-on view” of the Galaxy Mapping the Milky Way with SDSS Bottom: Density contours around the axis of symmetry Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Density Maps: D > 3kpc (F, G, early K) Mapping the Milky Way with SDSS Right: X-Y maps of number density distribution at Z=5, 4, 12, 10 kpc for .1 < r-i < .15 stars (~F/G SpT) Signatures of overdensities Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Edge-on View (R-Z density distribution) Right: f(R, Z) density distribution (“edge-on view”) 1 kpc (bottom right; M dwarfs) to ~20kpc (top left; F dwarfs) scales Map Analysis Summary: Mapping the Milky Way with SDSS Smooth, axisymmetric, background consistent with exponentials (disk) and power laws (halo) Overlaid by localized overdensities: clumps and streams Most major overdensities at D > 3kpc Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Disk Model Fit M-dwarfs (D < 2 kpc) excellently fit by two exponentials Best fit: Mapping the Milky Way with SDSS Z0 = 25 pc H1=245 pc, H2=740 pc L1=2.15 kpc, L2=3.3 kpc f=13% Reduced c2=1.6 Uncertainties and covariances easily seen in c2 plots (left) Same values obtained when allowing the scales to vary in adjacent color bins Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Halo Fits 10kpc < D < 20kpc Power law Clearly aspherical, oblate qH = 0.6 Normalization: fH ≤ 0.5%, Poorer fit (reduced c2~3) Mapping the Milky Way with SDSS nH = 2.8 Indicative of large scale departures from simple power law (dual halo) Or clumpiness of the halo (Bell et al. 2008) Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Why Should You Care? Chen et al. (2001; SDSS turnoff stars, ~280 deg2) 12 Juric et al. (2008; ~6500 deg2) Significant disagreements in prior measurements 10 8 Siegel et al. (2002; ~15deg2) Significant disagreements about the percentage of mass contained in the thick disk IMPORTANT, as these fits are EXTRAPOLATED to the rest of the Galaxy when building dynamical models Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Value of Wide Area I: Breaking the Degeneracy NGP line of sight only: H1=260pc H2=1000pc f=4% Mapping the Milky Way with SDSS H1=245pc H2=750pc f=13% Two substantially different fits describe the NGP line of sight equally well A number of prior studies are NGP-only Wide area survey is necessary to break the degeneracy In our case, the fit always converged to the same minimum (or did not converge at all) Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Value of Wide Area II: Seeing/Avoiding the Substructure Vermin of the Galaxy If unrecognized, overdensities will influence the fits The only way to identify them is with a wide area survey Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Distribution of Metallicity in the Milky Way Ivezic et al. (2008) No radial metallicity gradient Features in density space ↔ features in metallicity space Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Vertical Variation of Metallicity Distribution Function (MDF) 1) Clear disk/halo separation • 2) Mapping the Milky Way with SDSS Halo: Gaussian MDF Vertical metallicity gradient in the disk Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Disk MDF Detection of metallicity gradient ¹ D (Z ) = ¹ 1 + ¢ ¹ exp(¡ jZ j=H ¹ ) dex H ¹ = 1:0kpc; ¹ 1 = ¡ 0:78; ¢ ¹ = 0:35 Disk metallicity distribution: Mapping the Milky Way with SDSS Approximately fitted with a gaussian with Z-dependent mean, F([Fe/H]) ~ G(mD, s=0.16) Best fit by an asymmetric distribution with a slight low-metallicity tail Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Asymmetric Disk MDF Disk Disk Halo Halo Halo Halo Disk Disk ©disk ([F e=H ]) / G(¹ D ; ¾= 0:11) + 1:7 £ G(¹ D ¡ 0:1; ¾= 0:21) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Adding Kinematics: Proper motions towards the NGP Easy to interpret: vf= mb×D is the rotational velocity Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Halo/Disk Components, Disk Rotational Velocity Lag Top panels: small dots are individual stars, large symbols are the median values Top left: disk stars show clear rotational velocity lag Top right: halo stars vf ~ 220 km/s, no significant rotation Disk Halo Halo Disk Mapping the Milky Way with SDSS Halo Disk Bottom left: disk velocity lag not linear Bottom right: halo velocity dispersion increase consistent with being due to photometric errors only Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Rotational velocity distribution functions 0.8 < Z/kpc < 1.2 2 < Z/kpc < 3 Mapping the Milky Way with SDSS 1.5 < Z/kpc < 2 5 < Z/kpc < 7 1. Disk: Asymmetric rotational velocity distribution with Z-gradient 2. Halo: Unimodal, Gaussian velocity distribution of fixed dispersion and mean Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Revisiting the Thin/Thick Disk Dichotomy Vertical density distribution exhibits a break Metallicity distribution exhibits a metal-weak tail Rotational velocity distribution exhibits a lowvelocity tail Is there evidence for a model of the density, metallicity, and velocity distributions as a superposition of two distinct populations? Metal-rich, kinematically cold, thin disk Metal-poor, kinematically warm, thick disk? Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Metallicity-rotational Velocity Correlation Ivezic et al. (2008) Left: Expected rotational velocitymetallicity correlation at 1kpc < Z < 1.2kpc Right: Observed metallicity-rotational velocity correlation Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Thick Disk Formation: Major merger, heating, or accreted? We find no evidence for the thick disk being a clearly distinct component Abundance dichotomy May pose a problem for formation by a single major merger We cannot differentiate between the other two possibilities Caveat: Model admittedly simple. Will be revisited in the context of dynamical models of the Galaxy (work in progress) Usually used to argue for “catastrophic” thin/thick disk formation Not so: see Schonrich & Binney (arXiv: 0809.3006v1) Thin disk/thick disk… The Disk. Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Disk Formation with Radial Migration Roškar et al (2008) simulation: gas accretion only, SNe feedback, no mergers. Analysis: Loebman et al. (arXiv: 0810.5158v1). Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Next Decade: Decoding the Rosetta Stone of Galaxy Formation Disclaimer: “This part contains statements which may be deemed to be "Forward-Looking Statements" within the meaning of Section 27A of the Securities Act of 1933… Statements that are not historical facts, including statements about our beliefs and expectations, are forward-looking statements. These statements are based on current plans, estimates and projections, and therefore you should not place undue reliance on them.” Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Milky Way Halo Belokurov et al (2007) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Dark Side is Really Dark Simon & Geha (2007) More, potentially closer, ultrahigh M/L dwarfs left to be uncovered, providing targets for DM annihilation searches. A project for DES. Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Future: Disk is Where the Action Is Disks are the least well understood results of structure formation Disk formation Disk survival Old disks? Reformed after z=1? Disk structure When? Where? How? Disk dark matter? This is where this program is going We will soon be able to observationally constrain all of these questions Surprisingly little being done to actually do it (compared to the efforts devoted to the halo) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Mass Distribution of the Milky Way Disk Measuring the vertical force field Kz(R,φ) Constraining the mass scale length of the disk Constraining the DM distribution (e.g. Dehnen & Binney 1998, BT 2007) The Sun SEGUE Imaging Footprint D~2kpc Galactic Plane (edge-on view) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Juric et al. (2008) Abundant Disk Substructure Monoceros stream (Newberg et al. 2002; not shown here, more later) Two additional disk substructures R=6.5kpc, Z=1.5kpc: ~20% over the background “Thick disk asymmetry” of Larsen & Humphreys (1996) R=9.5kpc, Z=0.8kpc: ~50% over the background Faint kinematic/metallicity signature (Bond et al., in prep) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Juric et al. (2008) Abundant Disk Substructure Need further follow-up (wider survey) to trace its full extent 20-40 such streams in the disk (crude extrapolation) These are merger remnants + results of secular evolution: simulations should be able to predict their properties Need a statistical description to compare with simulations Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Evidence of Mergers: Monoceros Stream Kazantzidis et al. (2007) simulations Juric et al. (2008) “observations” Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Monoceros Stream: Evidence of Accretion Disk only Mapping the Milky Way with SDSS Disk w. Mon. Ivezic et al. (2008) Monoceros stream (Newberg et al. 2002) clearly distinct in metallicity space Metal poor compared to the disk, but metal rich compared to the halo ([Fe/H] = -0.95 dex) Strong evidence for external origin (merger remnant, as opposed to disk flaring or excitation) Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH North-South Asymmetries: “Disk Buckling?” Morrison, Rockosi, Juric, ... Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Why Now? SDSS: “I’m not dead!” Kinematics SEGUE I/II, APOGEE (2011 onwards) Upcoming: SDSS-III (now): 2,000 deg2 of additional imaging Pan-STARRS PS1 (2009): 30,000 deg2 to r=23, deeper w. coadds SkyMapper (2009): 20,000 deg2 to r=21.6, deeper w. coadds DES (2011?): 5,000 deg2 to r=25.6 GAIA (2013): 40,000 deg2 to r=20, mas proper motions LSST (2015): 20,000 deg2 to r=24, to r=27 with coadds RAVE (2010?): 1M radial velocities In the next 5 years, the Galaxy will be mapped in unprecedented detail. The major challenge for the future is converting that data into useful information. Boon for Galactic structure studies We need a framework for systematic and quantitative comparison with simulations Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH The Galaxy with Pan-STARRS and LSST RR Lyrae limit MS stars limit Pan-STARRS 3p Survey at z=2.1kpc above the plane Bullock & Jonhston (2005), Ivezic et al. and the LSST collab. (2007) Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH SDSS III / APOGEE :: Unveiling the History of the Milky Way H-band spectra of ~100,000 red giants selected from 2MASS R~20,000, typical S/N~100 Radial velocities to ~ 0.5 km s-1 Individual abundances of ~10 chemical elements, including O, C, N, Fe, Si Increase number of high-res, high-S/N spectra by factor of 100! Bright time observations, 2011-2014, with new, 300-fiber, cryogenic spectrograph Detailed chemical and kinematic mapping of all Galactic stellar populations, including the inner Galaxy (high AV regions) Chemical evolution and enrichment Merger remnants in phase space Chemical fingerprinting => “Growing” the Milky Way’s merger tree Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH Instead of a Summary: The Galaxy Within a Decade (an optimist’s view) Knowledge of dynamical state and formation history of a galaxy: The mass of the DM halo (GAIA, HVSs) Census of (visible) DM subhaloes (Pan-STARRS, DES, LSST) Dark matter distribution (SDSS III, GAIA, LSST) Dynamical model (SDSS III, Pan-STARRS, GAIA, LSST) Disk formation, in situ vs. accretion (SDSS III, GAIA, LSST) Tracing merger remnants (SDSS III, GAIA, Pan-STARRS, LSST) All of these, esp. the insights on disk build-up, generalize to other galaxies Necessary for complete, beginning-to-end, understanding of structure formation in the universe. Mapping the Milky Way with SDSS Mario Juric <mjuric@ias.edu>, Tuesday, January 6th, 2009. Ohio State, Columbus, OH