Kinematics of the outer halo of M87 as mapped by planetary nebulae
Longobardi, A., Arnaboldi, M., Gerhard, O., Pulsoni, C., Söldner-Rembold, I., 2018, A&A, 620, A111
Multiple models describing the mass and size growth in galaxies with little to no star formation have been put forward by the community. One of these models, that is well supported by observations, is called the two phase formation scenario. In this scenario, the inner component of a passive galaxy is formed at high redshifts on short times scales, while the outer component is formed through a slower accretion process. In the nearby universe, galaxies in clusters are surrounded by a third component called intra-cluster light (ICL). This background component is composed of stars not gravitationally bound to any one galaxy in the cluster. One question facing astronomers is how to disentangle these components for a given galaxy. One relatively successful method is to use surface brightness profiles which analytically describe how the surface brightness of a galaxy changes as a function of radius. An accreted component is often identified using this method as a change in slope of the surface brightness profile at large radii.
Do halo stellar populations change gradually and smoothy with radius into the ICL or are the halo populations and the ICL distinct and dynamically separate populations? Using planetary nebulae (PN) as discrete tracers for the stellar kinematics, we are able to systematically study the physical properties of early type galaxies out to large radii where the surface brightness is too low for absorption line spectroscopy. We apply this analysis to the the giant elliptical galaxy M87 near the dynamical centre of the Virgo cluster, using a total of 298 PN (253 halo-PN and 45 ICL-PN) out to an average radius of 135 kpc to identify the galaxy’s subcomponents: ICL, smooth halo, and crown accretion event. Specifically, gaussian mixture modelling is used to statistically separate the components and is further supported by the difference in PN luminosity functions (LF); the intra-cluster PNLF shows a “dip” that is not apparent in the halo PN luminosity function supporting the the halo/ICL kinematic classification of the PN.
We find that the line-of-sight velocity distribution (LOSVD) of the intra-cluster stars is strongly non-Gaussian with asymmetric wings and a newly identified peak at 900 kms-1. This indicates that the ICL stars around M87 are not yet in dynamical equilibrium and that the Virgo cluster is still being built up. The surface density profile of the ICPN tagged via stellar kinematics decreases as a power law with a logarithmic slope of −α_ICL = −0.79 ± 0.15. This is shallower than the halo PN profile which aligns with surface brightness profile of M87 (see figure). The halo-PN therefore show a steeper surface density profile as compared to the full surface density profile (halo+ICL). The stars tagged as part of the halo show slow rotation (∼< 25 kms−1) consistent with the slow rotator regime. The corresponding velocity dispersion profile increases from 270 kms−1 at a radii 2 − 10 kpc to 300 ± 50 kms−1 at average radii of 50 − 70 kpc. The velocity dispersion profile however decreases steeply to 100 kms−1 at an average radius of 135 kpc. Upon comparison between dynamical models and hydrostatic analysis of the X-ray emitting gas, we conclude that the surface density and velocity dispersion profiles of the halo PN are approximately compatible with being in dynamical equilibrium. Therefore to be consistent, the stellar halo orbits must dramatically change from an isotropic distribution below a radius of 60 kpc to a strongly radially anisotropic configuration at larger radii. This is in agreement with the scenario where the outer halo was formed by accretion of satellites. The differing PNLFs and spatial distributions support the conclusion that the halo and ICL are distinct components.
Three dynamically distinct stellar populations in the halo of M49
Hartke, J., Arnaboldi, M., Gerhard, O., Agnello, A., Longobardi, A., Coccato, L., et al. 2018, A&A, 616, A123
M49 (NGC 4472) is the dominant galaxy in subcluster B of the Virgo Cluster, and a benchmark for studying the build-up of the extended halos of brightest group galaxies in the outskirts of galaxy clusters. We investigated the kinematics in the outer halo of M49, using Planetary Nebulae (PNe) as kinematic tracers. We combined kinematics from the extended Planetary Nebula Spectrograph(PN.S) early-type galaxy survey (ePN.S) with the deepest and most extended photometry of PNe in the halo of M49 to date (Hartke et al. 2017) obtained with SuprimeCam@Subaru.
We studied the position-velocity-plane for bright and faint PN populations out to 95 kpc radius and employed Gaussian mixture models to identify stellar populations with distinct kinematics and accretion histories. This analysis resulted in the detection of three kinematically distinct stellar populations: stellar-kinematic substructure associated with the interaction of M49 with the dwarf irregular galaxy VCC 1249 and two kinematically distinct PN populations associated with the main M49 halo and the extended intragroup light (IGL). The overall luminosity profile and velocity dispersion at 80 kpc are consistent with a flat circular velocity curve extrapolated from X-ray observations. The dispersion of the PNe associated with the IGL joins onto that of the satellite galaxies in subcluster B at 100 kpc radius.
This is the first time that the transition from halo to IGL is observed based on the velocities of individual stars. Therefore the halo of M49 has undergone an extended accretion history within its parent group potential. The blue colours of the IGL component are consistent with a population of stars formed in low-mass galaxies at redshift 0.5 that has since evolved passively.
The build-up of the cD halo of M87 - evidence for accretion in the last Gyr
Longobardi A., Arnaboldi M., Gerhard O., Mihos J. C., 2015a, A&A, 579, L3
We present kinematic and photometric evidence for an accretion event in the halo of the cD galaxy M87 in the last Gyr. Using velocities for ~300 planetary nebulas (PNs) in the M87 halo, we identify a chevron-like substructure in the PN phase-space. We implement a probabilistic Gaussian mixture model to identify the PNs that belong to the chevron. From analysis of deep V-band images of M87, we find that the region with the highest density of PNs associated to the chevron, is a crown-shaped substructure in the optical light. We assign a total of N_(PN,sub)=54 to the substructure, which extends over ~50 kpc along the major axis where we also observe radial variations of the ellipticity profile and a colour gradient. The substructure has highest surface brightness in a 20kpc x 60kpc region around 70 kpc in radius. In this region, it causes an increase in surface brightness by >60%. The accretion event is consistent with a progenitor galaxy with a V-band luminosity of L=2.8\pm1.0 x 10^9 L_(sun,V), a colour of (B-V)=0.76\pm0.05, and a stellar mass of M=6.4\pm2.3 x 10^9 M_sun. The accretion of this progenitor galaxy has caused an important modification of the outer halo of M87 in the last Gyr. By itself it is strong evidence that the galaxy's cD halo is growing through the accretion of smaller galaxies as predicted by hierarchical galaxy evolution models.
The outer regions of the giant Virgo galaxy M87 II. Kinematic separation of stellar halo and intracluster light
Longobardi A., Arnaboldi M., Gerhard O., Hanuschik R., 2015b, A&A, 579, A135
We present a spectroscopic study of 287 Planetary Nebulas (PNs) in a total area of ~0.4 deg^2 around the BCG M87 in Virgo A. With these data we can distinguish the stellar halo from the co-spatial intracluster light (ICL). PNs were identified from their narrow and symmetric redshifted lambda 5007\4959 Angstrom [OIII] emission lines, and the absence of significant continuum. We implement a robust technique to measure the halo velocity dispersion from the projected phase-space to identify PNs associated with the M87 halo and ICL. The velocity distribution of the spectroscopically confirmed PNs is bimodal, containing a narrow component centred on the systemic velocity of the BCG and an off-centred broader component, that we identify as halo and ICL, respectively. Halo and ICPN have different spatial distributions: the halo PNs follow the galaxy's light, whereas the ICPNs are characterised by a shallower power-law profile. The composite PN number density profile shows the superposition of different PN populations associated with the M87 halo and the ICL, characterised by different PN alpha-parameters, the ICL contributing ~3 times more PNs per unit light. Down to m_5007=28.8, the M87 halo PN luminosity function (PNLF) has a steeper slope towards faint magnitudes than the IC PNLF, and both are steeper than the standard PNLF for the M31 bulge. Moreover, the IC PNLF has a dip at ~1-1.5 mag fainter than the bright cutoff, reminiscent of the PNLFs of systems with extended star formation history. The M87 halo and the Virgo ICL are dynamically distinct components with different density profiles and velocity distribution. The different alpha values and PNLF shapes of the halo and ICL indicate distinct parent stellar populations, consistent with the existence of a gradient towards bluer colours at large radii. These results reflect the hierarchical build-up of the Virgo cluster.
The planetary nebula population in the halo of M87
Longobardi A., Arnaboldi M., Gerhard O., Coccato L., Okamura S., Freeman K. C., 2013, A&A, 558, 42
Over the past years we conducted an observational campain with the Planetary Nebulae Spectrograph (Douglas et al. 2002) aimed to measure the radial velocities of PNe in the halos of ETGs (see Figure 1). In our first official data release (Coccato et al. 2009) we combined absorption line data and PNe radial velocity measurements in 16 ETGs. Our analysis showed that: i) PNe are good tracers of the mean stellar population kinematics, as their kinematics and number density agrees with the stellar absorption line kinematics and surface brightness; ii) outer halos have more complex radial profiles of the lR parameter (a proxy for the angular momentum, Emsellem et al. 2007) than observed within 1 Re. Interestingly, in the halo, some fast rotators have declining lR radial profiles, almost reaching the slow rotator regime, while some slow rotators have slowly increasing lR profiles, which reach the fast rotator regime (see Figure 2); iii) the velocity dispersion profiles fall into two groups, with part of the galaxies characterized by slowly decreasing profiles and the remainder having steeply falling profiles; iv) the halo kinematics are correlated with other galaxy properties, such as luminosity, shape, total stellar mass, V/s, and number of PNe per unit luminosity, with a clear distinction between fast and slow rotators.
Distinct core and halo stellar population in the bright coma cluster galaxy NGC 4889
Coccato, L., Gerhard, O., Arnaboldi, M. 2010, MNRAS, 407, L26
We construct radial profiles of line strength indices along the major axis of NGC 4889 by combining literature data for the central regions and new deep spectroscopic data for the halo regions. We then derive age, metallicity and alpha-enhancement radial profiles and their gradients using Single Stellar Population models by Thomas et al. (2003). This represents the most spatially extended dataset with both stellar kinematics and line strength indices for a brightest cluster galaxy.
We observe a different population content and gradient between the central regions of the galaxy (R<18 kpc="" and="" the="" outer="" halo="" r="">18 kpc). The inner ~18 kpc (~1.2 Re) of NGC 4889 are characterized by a strong [Z/H] gradient and a nearly constant values of [alpha/Fe]. The outer regions (18 kpc < R < 60 kpc) are characterized by a constant metallicity content strong negative gradient in the abundance ratio and older ages.
These data indicate that the central parts of NGC 4889 and its halo have undergone different formation mechanisms. Data in the center indicate a short star formation timescale, where the stars formed outside-in, reminiscent of a quasi-monolithic dissipative collapse. On the contrary, the data in the halo suggest that it was accreted from shredded satellite galaxies, as suggested also by numerical simulations, over the central galaxy that was already in place.
Our measurements are also consistent with recent results on the size evolution of bright ETGs with redshift, i.e. at high redshifts ETGs are smaller and more compact than ETGs of similar mass at z = 0. Their effective radius evolves as Re ~ (1+z)-1.3 (van Dokkum et al. 2010). Scaling the present Re of NGC 4889 with this relation would predict Re = 6.2 kpc, at z = 1, which is consistent with the half light radius measured if considering the central regions of the galaxy only, on the assumption that outer regions of NGC 4889 were accreted later, at z < 1. Our finding for NGC 4889 suggests that we may have found local stellar population signatures of the observed ETG size evolution.