Kinematics and Angular Momentum in Early Type Galaxy Halos
TThe General Assembly of Galaxy Halos: Structure, Origin and Evo-lutionProceedings IAU Symposium No. 317, 2015A. Bragaglia, M. Arnaboldi, M. Rejkuba, O. Gerhard c (cid:13) Kinematics and Angular Momentum inEarly Type Galaxy Halos
Jean P. Brodie , Aaron Romanowsky , , and the SLUGGS team UC Observatories, University of California1156 High St, Santa Cruz, CA 95064, USAemail: [email protected] San Jos´e State University, San Jose, CAemail: [email protected] http://sluggs.ucolick.org Abstract.
We use the kinematics of discrete tracers, primarily globular clusters (GCs) andplanetary nebulae (PNe), along with measurements of the integrated starlight to explore the as-sembly histories of early type galaxies. Data for GCs and stars are taken from the SLUGGS widefield, 2-dimensional, chemo-dynamical survey (Brodie et al. 2014). Data for PNe are from thePN.S survey (see contributions by Gerhard and by Arnaboldi, this volume). We find widespreadevidence for 2-phase galaxy assembly and intriguing constraints on hierarchical merging undera lambda CDM cosmology.
Keywords. galaxies: elliptical and lenticular, galaxies: star clusters, galaxies: formation, galax-ies: abundances, galaxies: kinematics and dynamics
1. Introduction
It is now generally agreed that galaxies form in two phases. A widely accepted scenarioinvolves an early phase, occurring at a redshift of 2 or earlier, that produces a relativelycompact nugget that grows over time by continually accreting lower mass satellites in dryminor mergers (e.g., Oser et al. 2010, Naab et al. 2014). Alternatively, large star formingdisks may evolve passively until quenched by process that relate to their densities orvelocity dispersions, perhaps also increasing somewhat in size via dry minor mergers(van Dokkum et al. 2015). Given that more than 90% of the total mass and angularmomentum of a galaxy lie beyond one effective radius (R e ), it stands to reason thattesting models for the assembly of galaxy halos will require observations out to largegalactocentric radius.The SAGES Legacy Unifying Globular Clusters and GalaxieS (SLUGGS) survey (Brodieet al. 2014) uses SUBARU/SuprimeCam imaging and Keck/DEIMOS spectroscopy togenerate 2-dimensional metallicity and kinematic data out to ∼ e for galaxy starlightand out to ∼
10 R e for globular clusters (GCs) in 25 nearby early type galaxies. Here wereport initial results from the SLUGGS survey, whose observational component is near-ing completion. We also include some results from the Planetary Nebula SpectrographGalaxy Survey (PN.S), which uses planetary nebulae to explore kinematics and dynamicsin 33 nearby galaxies (Arnaboldi et al., 2016, in preparation).Underpinning the use of GCs to unravel the formation histories of galaxies is thefact that GC formation accompanies all the major star forming events in a galaxy’shistory. Typically containing 10 to 10 stars, GCs are bright enough to allow integratedspectroscopy out to distances in excess of 50 Mpc. The vast majority of GCs are as old1 a r X i v : . [ a s t r o - ph . GA ] D ec Jean P. Brodie, Aaron Romanowsky and the SLUGGS team
Figure 1.
Complementary ETG surveys. Many other surveys are ongoing that target early typegalaxies. While SLUGGS and PN.S provide excellent velocity resolution and wide field coverage,other surveys have much larger galaxy samples. as we can measure them ( >
10 Gyr) and are bright beacons that were “along for the ride”during all the mergers and acquisitions that have built the galaxies we see today.GC systems typically divide into two subpopulations, a blue, metal-poor populationthat appears to trace the build up of galaxy halos, and a red, metal-rich, population thatis linked to bulge development. The subpopulations are distinct not only in metallicity(e.g., Brodie et al. 2012), but are also kinematically distinct (Pota et al. 2013). Like redGCs, planetary nebulae appear to be closely linked to the starlight in ETGs (Coccato etal. 2009).Many surveys are underway that are targeting ETGs. Figure 1 shows a figure of meritfor these surveys that is an update of the figure in Brodie et al. (2014). SLUGGS andPN.S were designed to offer wide field coverage with very high velocity resolution, but thegalaxy sample is relatively small ( ∼
2. Angular Momentum
Figure 2 shows specific angular momentum, λ R versus radial extent for SLUGGS starsand PNe. The definitions of λ R is different in the two panels. The SLUGGS version uses alocal definition defined in successive annuli. The PN.S version is cumulative (as in Atlas D analyses (Emsellem et al. 2011). We definite a local version to preserve evidence of radialtransitions. Evident from both tracers is the trend for galaxies that were defined as slow arly Type Galaxy Halos Figure 2.
Specific Angular Momentum. Left panel: local specific angular momentum versusradial extent in units of effective radius for ETG stars from the SLUGGS survey. Right panel:cumulative specific angular momentum versus radial extent for ETGs from the P.NS survey(courtesy of L.Coccato and M.Arnaboldi). Both surveys reveal the same trends. Central slowrotators (red) remain slow. Central fast rotators (blue) may rise, plateau or fall with increasingradius.
Figure 3.
The radial distribution of V rms for galaxies in the PN.S survey (courtesy ofL.Coccato and M.Arnaboldi). rotators based on observations in their central regions, to remain slow with increasingradius. Centrally defined fast rotators may continue to rise, plateau or decline withincreasing radius. A similar result was obtained form the VIRUS-P survey of 33 massivegalaxies (Raskutti et al. 2014).
3. Velocity Dispersion
Figure 3 is a plot of root mean square velocity (V rms , a proxy for velocity dispersion)against radial extent from PN.S observations of planetary nebulae. The PN.S team find a Jean P. Brodie, Aaron Romanowsky and the SLUGGS team
Figure 4.
The radial distribution of V rms of GCs for a subsample of the SLUGGS galaxies. dichotomy between galaxies displaying nearly flat and steeply declining profiles. Possibleexplanations for such an effect include a dark matter dichotomy or anisotropy projectioneffects. See the contribution by Napolitano in this volume for further discussion of thispoint. In Figure 4 we show the rms velocity as a function of radius for GCs from theSLUGGS survey. Although we do not see a dichotomy in our data, not all of the SLUGGSgalaxies have yet been included.
4. Mass and Dark Matter
Pota et al. (2015) carried out multi population dynamical modeling of NGC 1407using the spherical Jeans equation and employing stars, metal-rich and metal-poor GCsas three independent tracers of the dark matter distribution. Using a Bayesian MCMCanalysis, we determined that different anisotropies are needed to fit the profiles and thatthe metal-poor GCs have tangential anisotropy. This result for blue (metal-poor) GCsis inconsistent with expectations from hierarchical merging. Kurtosis measurements fora larger number of SLUGGS galaxies (Pota et al. 2013) reveal that the majority of blueGCs are on tangential orbits, while there is a mix of radial and tangential orbits for red(metal-rich) GCs and for PNe.Cappellari et al. (2015) used Jeans axisymmetric models (JAM) on a combination ofATLAS D and SLUGGS data for 14 galaxies classified as fast rotators based on theircentral ( < e kinematics). The JAM models allow spatially varying anisotropy andquite general profiles for the dark matter; no restriction on slope is imposed. This simpleaxisymmetric model fits the data for all 14 galaxies remarkably well and yields a powerlaw density profile with exponent 2.19 ± e > r > e ; see Figure 6). The scatter among the 14 galaxies is only 0.14. Sincethe a power law density profile is not a generic prediction of lambda CDM cosmology,this results offers tight constraints on the cosmological models. It has long been known, arly Type Galaxy Halos Figure 5.
The best fit to generalized NFW profiles for metal rich and metal poor GCs (toppanel) and stars (bottom panel). Based on figures in Pota et al. (2015). The solid lines are the1 σ boundaries of the fits. e.g., from observations of the gas, that the rotation curves of spiral galaxies flatten atlarge radius, reflecting the interplay between dark matter and baryons. This is the firstindication that the same flattening occurs in early type galaxies revealing a surprising“dark matter conspiracy” across markedly different galaxy types.
5. Velocity Position Phase Space
Simulations of satellite infall (e.g. Bullock & Johnston 2005) show that satellite galaxyaccretion can set up a temporary set of nested chevrons in the velocity-position phasespace of the accreted material, due to repeated passage near to the center of the moremassive (accreting) galaxy. Direct evidence of this effect was reported by Romanowskyet al. (2011), based on high precision radial velocities of more than 500 GCs associatedwith M87, the massive elliptical galaxy at the center of the Virgo cluster (Figure 7). Weinferred that M87 had acquired an L ∗ galaxy, bringing in ∼ ∼
300 PNe (Longobardi et al. 2015). Jean P. Brodie, Aaron Romanowsky and the SLUGGS team
Figure 6.
Measured total density profiles from Cappellari et al. (2015). Solid lines are theindividual galaxy measurements which display very little scatter about the isothermal relationwith exponent 2.19 ± Figure 7.
Velocity-position phase space for GCs around M87 reveals chevron structures thatare characteristic of recent massive accretion events. The figure is adapted from Romanowskyet al. (2011). The image in the bottom right hand corner of the plot is from Mihos et al. (2005).
6. Summary
Wide field surveys of early type galaxies (such as SLUGGS and PN.S) are revealingwidespread evidence in favor of the two-phase paradigm of galaxy assembly. In particu-lar, we find a wide range of stellar and PNe radial profiles. Inner slow rotators remainslow; inner fast rotators can rise, plateau, or fall with increasing radius from the center arly Type Galaxy Halos D datafor galaxy centers and SLUGGS data for the outer regions (to typically 4R e ) reveals aremarkable dark matter/baryon conspiracy to produce power law density profiles, withexponent 2 . ± .
04, with very little galaxy-to-galaxy scatter ( <
7. Acknowledgements
This work was supported by NSF grant AST-1211995.
References
Brodie, J. P., Romanowsky, A. J., Strader, J., et al., 2014,
ApJ , 796, 52Brodie, J. P., Usher, C., Conroy, C., et al., 2012,
ApJ , 759, L33Bullock, J. S., & Johnston, K. V., 2005,
ApJ , 635, 931Cappellari, M., Romanowsky, A. J., Brodie, J. P., et al. 2015,
ApJ , 804, L21Coccato, L., Gerhard, O., Arnaboldi, M., et al., 2009,
MNRAS , 394, 1249Emsellem, E., Cappellari, M., Krajnovi´c, D., et al., 2011,
MNRAS , 414, 888Longobardi, A., Arnaboldi, M., Gerhard, O., & Mihos, J. C., 2015,
A&A , 579, L3Mihos, J. C., Harding, P., Feldmeier, J., & Morrison, H., 2005,
ApJ , 631, L41Naab, T., Oser, L., Emsellem, E., et al., 2014,
MNRAS arXiv
ApJ , 725, 2312Pota, V., Romanowsky, A. J., Brodie, J. P., et al., 2015,
MNRAS , 450, 3345Pota, V., Forbes, D. A., Romanowsky, A. J., Brodie, J. P., et al. 2013,
MNRAS , 428, 389Raskutti, S., Greene, J. E., & Murphy, J. D., 2014,
ApJ , 786, 23Romanowsky, A. J., Strader, J., Brodie, J. P., et al., 2012,
ApJ , 748, 29van Dokkum, P. G., Nelson, E. J., Franx, M., et al., 2015,, 748, 29van Dokkum, P. G., Nelson, E. J., Franx, M., et al., 2015,