PNe as mass tracers Dark-to-luminous properties of early-type galaxies Nicola R. Napolitano Kapteyn Institute Groningen (NL) ESO workshop: PNe beyond the Milky Way. Garching May 2004
CollaboratorsN. DouglasKapteyn Institute PN.S ConsortiumK. KuijkenUn. Leiden M. ArnaboldiINAF-OAT M. CapaccioliINAF-OAC A. RomanowskyUn. Nottingham M. MerrifieldUn. Nottingham H. MerrettUn. Nottingham K. FreemanMt. Stromlo O. GerhardUn. Basel PNe as mass tracers Dark-to-luminous properties of early-type galaxies ESO workshop: PNe beyond the Milky Way. Garching May 2004
Observed trends of the M/L in early type galaxies Inner stellar kinematics (long-slit spectroscopy) PNe kinematics
When we measure =M/L ratios in galaxies we basicly deal with dark-to-luminous mass distribution assuming a radially constant stellar mass-to-light radius, * (from stellar population synthesis models) At every radius, R, (in B band for example): dark-to-luminous mass fraction In presence of a significant(?) dark halo, this quantity must grow with the distance from the center: OK! HOW MUCH in order to be consistent with CDM(=NFW+c vir )? Can we predict the radial run of the M/L in galaxies?
NFW 97 M DM =M DM (M vir, c vir ) Hernquist (1990) M lum =M lum (M l,tot,,R e ) M DM =M DM (M vir )M lum =M lum (M l,tot ) Stellar syntesis mod Shen et al. (2003) Bullock et al. (2001) Predicted M/Ls in CDM framework Formation efficiency Baryon fraction (CMB) Bennett et al. 2003
Predicted M/Ls in CDM framework M l,tot = 1.1x10 11 M sol sf =0.1 (Padmanabhan et al. 04) sf =0.2 (Fukugita et al. 98) sf =0.6 (Guzik&Seljak 02 Marinoni&Hudson 02) M/Ls increase more quickly for decreasing efficiencies and...
MASS Predicted M/Ls in CDM framework M l,tot = 1.1x10 11 M sol
MASS Predicted M/Ls in CDM framework M l,tot = 1.4x10 11 M sol
Predicted M/Ls in CDM framework M l,tot = 5x10 11 M sol MASS …and M/Ls increase more quickly for increasing masses Are these trends observed in early-type galaxies?
Predicted M/Ls in CDM framework M l,tot = 5x10 11 M sol MASS NGC 1399 M/L gradients
Predicted M/Ls in CDM framework Log M l,tot sf =0.6 sf =0.2 sf =0.1 sf M/L gradients
sf =0.6 sf =0.2 sf =0.1 Predicted M/Ls in CDM framework Log M l,tot Plateau Spread distribution Gerhard et al. 01 Magorrian&Ballantyne01
sf =0.6 sf =0.2 sf =0.1 Predicted M/Ls in CDM framework Log M l,tot
Napolitano et al Equilibrium? Log M l,tot
Inconsistent with sf < 0.6 at 95% s.l. Low concentrations c=5±2 Log M l,tot Baryon physics in galaxy luminous regions not accounted in simulation [adiabatic collapse (de Jong et al. 2003, Dutton et al. 2003)?]; hydrodynamical simulations (Meza et al. 2003, Wright et al. 2004) found DM not to be dominant up to 5R e in low-luminosity compact galaxies Halo stripping?
Conclusions We have a “toy-model” to make predictions of the M/L trend in the CDM framework, to be compared with future PNe observations AND other dynamical tracers (GCs, Xrays, lensing). Thanks to the PN.S statistical samples of PN radial velocities are reaching the precision which will allow to discriminate the efficiency of the baryon cooling in galaxies. We have shown on a sample of 21 galaxies that the M/L gradients by PN and stellar kinematics are generally in agreement with the CDM expectations with some “critical points”: a) nearly L * galaxies are in conflict with “reasonable” formation efficiency unless we do not assume low-concentration halos (at odd with collisionless CDM simulations, but qualitatively expected in adiabatic collapse picture); b) low-luminosity/gradients galaxies are tendentially more consistent with very high formation efficiency ( sf 0.6) while high-luminosity/gradients galaxies are spread on all the range of allowed efficiencies ( sf = ). The turn-off point is around M l,tot =1.2x10 11 M sun, i.e. around L * luminosities (M B ~ -20.2). More (accurate) M/L estimates are needed in order to confirm these results….
FUTURE... …The PN Spectrograph
Predicted M/Ls in CDM framework M l,tot = 1.4x10 11 M sol Precisions of 20% at 5-6 R e in order to discriminate models Such a precisions are expected to be reached with the PN.S
Dichotomy of ETs in DM properties as well as in luminous properties? faintbright boxydisky core power-law Such a dichotomy is observed in X-ray properties (Pellegrini 1999) Turn-off point M B ~ L * galaxies
MASS Predicted M/Ls in CDM framework M l,tot = 1.1x10 11 M sol NGC 3379
MASS Predicted M/Ls in CDM framework M l,tot = 1.4x10 11 M sol NGC 5128
Predicted M/Ls in CDM framework M l,tot = 5x10 11 M sol MASS NGC 1399 …and M/Ls increase more quickly for increasing masses Are these trends observed in early-type galaxies?
sf =0.6 sf =0.2 sf =0.1 Predicted M/Ls in CDM framework Log M l,tot Plateau Spread distribution
gradients Observed trends of the M/L in early type galaxies M/L “gradients” for systems with PNe kinematics
gradients Observed trends of the M/L in early type galaxies M/L “gradients” from PNe kinematics+ systems with extended long-slit spectroscopy (out to 2 R e )
PNe have been extensively used in the recent past to constrain the mass distribution in early-type galaxies up to unprecedented distances from the galaxy center Early works were based on poor statistical sample (the reliability of which has been discussed in Napolitano et al. 2001) year GALAXYn.PNe R out /R e M/LM/L * 1986M N (6) N N N N N N Recently the improved observation techniques are allowing much larger samples 2001N N > N > N > N
Observed trends of the M/L in early type galaxies Inner stellar kinematics (long-slit spectroscopy) PNe kinematics
B / * Observed trends of the M/L in early type galaxies Inner stellar kinematics (long-slit spectroscopy) PNe kinematics
Observed trends of the M/L in early type galaxies Sample of 21 ETGs In order to improve statistics we have added a sample of ETGs with extended long-slit spectroscopy data B / *
B/* B/* B/* B/* / * Observed trends of the M/L in early type galaxies Sample of 21 ETGs In order to improve statistics we have added a sample of ETGs with extended long-slit spectroscopy data
MASS Predicted M/Ls in CDM framework M l,tot = 1.1x10 11 M sol NGC 3379
MASS Predicted M/Ls in CDM framework M l,tot = 1.4x10 11 M sol NGC 5128
Predicted M/Ls in CDM framework M l,tot = 5x10 11 M sol MASS NGC 3379
MASS Predicted M/Ls in CDM framework M l,tot = 1.1x10 11 M sol
MASS Predicted M/Ls in CDM framework M l,tot = 1.4x10 11 M sol
Predicted M/Ls in CDM framework M l,tot = 5x10 11 M sol MASS
Predicted M/Ls in CDM framework NGC 5128 M l,tot = 1.4x10 11 M sol * = 3.5
Predicted M/Ls in CDM framework NGC 1399 M l,tot = 5x10 11 M sol * = 9
Predicted M/Ls in CDM framework NGC 3379 M l,tot = 1.5x10 11 M sol * = 7
Stellar population versus dynamical M/L
Predicted M/Ls in CDM framework
Napolitano et al Equilibrium? Low concentration c=5±2 22 sf =0.1 sf =0.2 sf =0.6
Predicted M/Ls in CDM framework Star mass following an Hernquist (1990) profile NFW dark halos with concentrations by Bullock et al. (2001) Luminous mass and dark mass are 1-parameter family models M DM /M lum is a 2-parameter quantity: total stellar mass, M *, and virial mass, M vir (or concentration c vir ).
Predicted M/Ls in CDM framework Luminous mass and dark mass are 1-parameter family models M DM /M lum is a 2-parameter quantity: total stellar mass, M *, and virial mass, M vir (or concentration c vir ). Hernquist (1990) NFW97 profile Shen et al. (2003) Bullock et al. (2001) Total stellar massvirial mass
Predicted M/Ls in CDM framework M * and M vir are related to the efficiency which the baryons cool in stars with, assuming that the original baryon fraction is the same for all the dark halos and close to the cosmic value from CBM measurements (Bennett et al. 2003) where Star formation efficiency
Predicted M/Ls in CDM framework NGC 5128 M l,tot = 1.4x10 11 M sol * = 3.5
Predicted M/Ls in CDM framework NGC 1399 M l,tot = 5x10 11 M sol * = 9
Predicted M/Ls in CDM framework NGC 3379 M l,tot = 1.5x10 11 M sol * = 7