Volume-limited limited sample

Transcription

1 Masses of Early-Type Galaxies from Atlas 3D Michele Cappellari

2 Mass DF degeneracy SAURON Data Model Cap ppellari+07 Both orbital distribution (DF) and kinematics are 3D Given gravitational potential one finds DF to fit kinematics Only constraints for models is DF > 0 Large variations in potential shape produce equally good fits to full LOSVD (Gerhard+98; Valluri+04; de Lorenzi+09) Little hope to study central dark matter via dynamics alone But total mass robustly determined (e.g. Thomas+ (e.g. Thomas+07)

3 M/L from dynamics versus population Key is to remove observational biases Selection bias Homology or virial assumption bias Single Stellar Population (SSP) approximation This project Mass-selected selected large volume-limited limited sample Dynamical axisymmetric models Include detailed 2D surface brightness (Multi-Gaussian Expansion by Emsellem+94 94) Allow for orbital anisotropy Accurate stellar population modeling Use multi band SDSS colors and line-strength Allow for multiple populations

4 Complete survey of ETGs Volume-limited limited sample M K < (from 2MASS) (> M ) D < 42 Mpc δ 29 < 35º b > 15º Atlas 3D observes all E/S0s 263 early type galaxies Multi wavelength data set SAURON stellar + gas kinematics CO single dish (IRAM 30m) CO interferometer (CARMA) HI interferometer (WSRT) 8 band photometry (SDSS + ING + 2MASS)

5 The Team PIs: Michele Cappellari (Oxford), Eric Emsellem (ESO), Davor Krajnović (Oxford), Richard McDermid (Gemini) CoIs / Students: (France, Germany, Spain, The Netherlands, UK, USA) Kathey Alatalo, Roland Bacon, Leo Blitz, Maxime Bois, Frederic Bournaud, Martin Bureau, Roger Davies, Tim de Zeeuw, Jesus Falcon-Barroso Barroso, Sadegh Khochfar, Harald Kuntschner, Raffaella Morganti, Thorsten Naab, Tom Oosterloo, Marc Sarzi, Nicholas Scott, Paolo Serra, Remco van den Bosch, Glenn van de Ven, Gijs Verdoes-Kleijn Kleijn, Lisa Young, Anne-Marie Weijmans

6 Jeans Anisotropic Models (JAM) V rms = (V 2 +σ 2 ) ½ SAURON JAM model Isotropic model stellar kinematics σ z = 0.85 σ R σ z = σ R V 2-integrals Jeans models have DF = f(e,l z ) σ z = σ R But real axisymmetric ETGs galaxies have σ z < σ R (Cappellari+07, Thomas+09) We now allow for σ z σ R DF = f(e,l z,i 3 ) (Cappellari 08) Just two parameters (i,σ z /σ R ) fit shape of both V rms ( rms and V!

7 Good fits to fast rotators ETGs Data Vrms Model Vrms Data V Model V 21 of 263 JAM 2-parameters dynamical models... Data Vrms Model Vrms Data V Model V

8 Simple ETGs structure from Atlas 3D 90% of local ETGs are fast rotators (Emsellem+07) (80% of luminous mass) A single parameter σ z /σ R captures most of their dynamics (but majority are barred) Rotation dictated by galaxy shape and σ ϕ = σ R Similar to classic hypothesis of faint ETGs isotropic rotators (σ z = σ R = σ ϕ ) (Davies+83; Kormendy & Bender 96) What formation process produces this regular structure? κ = 1.02 ± rms Definition: ratio of observed projected angular momentum versus the one of a model with σ ϕ =σ R κ= 1 (Cappellari08) (cfr. Satoh 80)

9 Notable exceptions Counterrotating disks like NGC4550 (Rubin+92; Rix+ κ 0.2 from JAM fit ; Rix+92) Explained by flipping major fraction of orbits Rare (2%) event in ETGs 10% of ETG are weakly triaxial slow rotators (Illingworth 77; Binney 78; Davies+83; Kormendy & Bender 96; Emsellem+07; Cappellari+07) Data V rms Model V rms Data V Model V Data V rms Model V rms Data V Model V Single component Two counterrotating JAM disks JAM

10 (M/L)-σ relation from Atlas 3D Best fit Cappellari+06 Full Atlas 3 D sample 109 ETGs with SBF distances General agreement with previous determinations (Cappellari+06; van der Marel & van Dokkum 07) Slope explains FP tilt (J (Jørgensen+ rgensen+96) Relation continues down to σ < 100 km/s Observed rms scatter 30% with SBF distances (Tonry+01; Mei+07) Decreases to 19% rms for 39 ETG with SBF in Mei+ Mei+07

11 Two populations models (Drory+04) Use high S/N SAURON spectra within 1Re: line strength indices Hβ, Fe5015, Mgb Use SDSS (+INT) ugriz and 2MASS JHK Essential to constrain contribution of possible young population via 2-components population models (e.g. Drory+04; Schawinski+07) SSP age biased towards age of young component and metallicity of old one (e.g. Serra & Trager 07) SSP M/L underestimates M/L of composite population

12 (M/L) pop versus (M/L) JAM : 1-SSP (M/L) pop < (M/L) JAM requires bottom-light IMF normalization (Kroupa 01; Chabrier 03) Like spirals (Bell & de Jong 01) No trend for (M/L) pop < 1 Star forming ETGs (M/L) JAM differences at 30% DM (M/L) pop > 2 Similar result with different codes (Bruzual & Charlot 03; Oneto-one All ETGs Maraston 05; Vazdekis 99). Population models from Serra, McDermid + in prep

13 (M/L) pop versus (M/L) JAM : 2-SSP 2-SSP models bring (M/L) pop in better agreement with (M/L) JAM Different trends above and below (M/L) pop 2 Still many outliers 30 30% DM Oneto-one All ETGs

14 Dark matter fractions within 1R e Many outliers disappear with accurate distances Assuming universal IMF Median DM = 26% Clear difference for old (>8Gyr) massive ETGs 30% DM Oneto-one scatter up to 50% DM ETGs with SBF dist

15 Interpreting constant-m/l models M/L K 8 of 28 models of S0 and Spirals Williams+09 see poster JAM models including NFW halo and extended kinematics Degeneracy between stellar M/L and halo mass Constant-M/L models give accurate total M/L Justifies using ratio of JAM and population M/L to measure DM fraction (within 1Re) if IMF is constant Williams+ find median DM of 16% similar to our result

16 Conclusions Dynamical and stellar population models for 263 local ETGs from Atlas 3 D project 90% of local ETGs are fast rotators Remarkably simple dynamical structure Mass dominated by stars within 1R e (median DM=26%) (Gerhard+01; Rusin+03; Cappellari+06; Koopmans+06; Thomas+07 07; Bolton+08; Williams+09) Clear difference for old (>8Gyr) massive ETGs: larger scatter and up to 50% DM fraction Related to assembly epoch? (see Thomas+ (see Thomas+09)