intracluster planetary nebulae as dynamical probes of the diffuse light in galaxy clusters

M. Arnaboldi ICPNe as probes of diffuse light in clusters July 2nd, 2005Magda Arnaboldi, PNe as Astronomical Tools July 2nd, 2005

Intracluster Planetary Nebulae as Dynamical Probes of the Diffuse Light in Galaxy

Clusters

Magda ArnaboldiINAF, Observatory of Turin

• Observations of Diffuse Light• Intracluster Light in Cosmological

Simulations• ICPNe in the Virgo Cluster: projected

phase-space distribution• Future prospects

M. Arnaboldi ICPNe as probes of diffuse light in clusters July 2nd, 2005

1951. Zwicky claimed the discovery of diffuse light in the regions between galaxies belonging to the Coma cluster.

1971-1977. Follow-up photographic surveys for diffuse light in Coma and rich clusters.

1989-1995. CCD photometry of diffuse light. First accurate measurements in Coma (Berstein et al. 1995). ProblemsProblems

typical surface brightness of the ICL is less than 1% of sky brightness;

it is difficult to disentangle between diffuse light associated with the halo of the cD galaxy at the cluster centre and the diffuse light component

Observations of ICL in clusters

Since 1995, large CCD and mosaic cameras have allowed measurements of ICL in Abell clusters.


Diffuse light measured in z∼0.25 clusters from stacking of SDSS imaging data. SB measurements of the ICL ranges from 27.5 mag �” at 100 kpc to 32 mag �” at 700 kpc in the observed R band (Zibetti et al. 2005).

Measurements of * in nearby clusters

Presence of diffuse light is traced by existence of tails, arcs and/or plumes with typical B = 27.8 mag �”, very narrow (~ 2 kpc ) and extended (~ 100 kpc) in Coma and Centaurus.

(Gregg & West 1998, Threntam & Mobasher 1998, Calcáneo-Roldán et al. 2000)

Diffuse light measured by deep CCD photometry out to large radii in clusters. (Abel 1651: Gonzalez et al. 2000; Abell 1413 & MKW7: Feldmeier et al. 2002, 2004a; HGC 90: White et al. 2003; Gonzalez et al. 2004).


An alternative method for probing intracluster light is through the direct detection and measurements of the stars themselves

Direct detections of IC stars

•Intergalactic SupernovaeIntergalactic Supernovae (Gal-Yam et al. 2003) & NovaeNovae (Neil et al. 2004).•Intergalactic Globular ClustersIntergalactic Globular Clusters (West et al. 1995, Jordàn et al.

2003)•UltraCompact Dwarfs UltraCompact Dwarfs (Drinkwater et al. 2003)

• IIntracluster red giant stars (IRGB)ntracluster red giant stars (IRGB) (Ferguson, Tanvir &

von Hippel, 1998; Durrell et al. 2002). Excess of red number counts in Virgo IC fields with respect to the HDF.

•CCompact isolated HII regionompact isolated HII region (Gerhard, Arnaboldi, Freeman,

Okamura 2002). It will dissolve by internal process in 108 yr. Stars and metals will then be added to the diffuse stellar population nearby.

• IC IC HI HI cloud cloud (Oosterloo & Van Gorkom 2005)

IC stars and gas in Virgo cluster


Merritt (1984): The ICL is removed from galaxies early during the cluster collapse and its distribution is predicted to follow closely that of galaxies. ICL smoothly distributed and ICL smoothly distributed and dynamically olddynamically old

We expect different distribution functions f(x,v) for the ICL depending on the formation mechanism.

Moore et al. (1996): the ICL is produced during galaxy harassment and tidal stirring during late infall. ICL still distributed in tails or plumes, ICL still distributed in tails or plumes, dynamically youngdynamically young..

The ICL in cluster is relevant for the baryonic fraction condensed in stars, star formation efficiency, and the metal enrichment of ICM via IC stars, especially in the cluster centre. It contains a fossil record of galaxy evolution and interactions in the cluster.

Importance of ICL in galaxy Importance of ICL in galaxy clustersclusters


Cosmological Simulations of Cluster Cosmological Simulations of Cluster FormationFormation

High-resolution resimulation of a part of a Universe that collapes into a galaxy cluster. Dark matter subhalos grow, fall into the cluster, may survive or merge into larger halos.

same processes may act on stars in galaxies, producing also ICL

from Springel et al. (2001)


Transvr. Velocities

Radial Velocities

R,Vz

x,Vz

y,Vz

Velocity distributions and projected phase space diagrams for the IC stars

N. Napolitano et al. 2003, ApJ, 594, 172

The two-dimensional phase space diagrams in one cluster show filaments, clusters of particles, and empty regions, all of which indicate a young dynamical age!


Cosmological Hydrodynamic Simulations with Cosmological Hydrodynamic Simulations with Star Formation and FeedbackStar Formation and Feedback

•Studies of ICL in cosmological simulations require a model of star formation from cold gas, including cooling and feedback effects. Recent studies are by Murante+2004, Willman+2004, Sommer-Larsen+2005. Here we use Gadget-2 with the two-phase model of Springel & Hernquist (2003).

• Current hydrodynamic simulations have significantly lower particle number than dark matter only simulations. Thus they cannot resolve small galaxies in clusters, probably causing overestimate of predicted ICL fraction. Recent high-resolution simulations obtain correct half-mass radii for galaxies above few 1010M. The large galaxies contribute a substantial part of the ICL; this part can be studied.

• Galaxies must be identified by a substructure-finding algorithm; here we use SKID (Stadel 2001).


ICL in a Cosmological Tree+SPH Simulation

Gadget-2 (V. Springel)Star FormationCooling (Z=0)2x4803 particlesMass Resolution:DM: 4.6 109 h-1 M

Gas: 6.9 108 h-1 M

Softening: 7.5 h-1 Box Size: 192 h-1 MpcCDM Concordance model (8=0.8)

G. Murante et al. 2004, ApJ, 607, L83

• We identified 117 clusters with M>1014 h-1M

• Stars in clusters were divided in two classes: bound & unbound.

• we evaluated radial density profiles of the two components


• “stacked” 2D star profile used.

• Fit range: from center to radius@1/3 central surface density

• Sersic fit is critically sensitive to the range

=3.66 (total

average), 4.37 (low-luminosity objects), 1.24 (high luminosity

objects) Clear evidence for ICL at large at large radii!radii!

133.3log)(log

/1

ee r

rr

Sersic Sersic fitsfits

ICL more centrally concentrated than galaxy light (see also Zibetti et al. 2005)

GalaxiesICLSersic fit


ICL fractions, star ICL fractions, star agesages

• More massive clusters show greater ICL fractions

• Stars in field are older than stars in galaxies (“slow” tidal effects?).

Murante et al. 2004; see also Sommer-Larsen et al. 2005


High resolution cosmological High resolution cosmological simulationssimulations

Gadget-2 (V. Springel)Star FormationCooling (Z=0)Feedback1.5 106 particles in ClusterMass Resolution:DM: 1.0 108 h-1 MGas: 1.5 107 h-1 M

Softening: 2.1 h-1 Mass 1.6 1014 h-1 M

Virial R: 1.1 h-1 MpcCDM Concordance model (8=0.8, h=0.7)

Borgani et al. (2005)


Distribution of Dark Matter and StarsDistribution of Dark Matter and Starsz=3 z=1 z=0z=3 z=1 z=0

Murante et al. 2005, in prep.


Galaxies and Intracluster StarsGalaxies and Intracluster Starsz=3 z=1 z=0z=3 z=1 z=0


• PNe trace light because the luminosity-specific stellar death rate should be independent of the precise state of the underlying stellar population (Renzini & Buzzoni 1986).

• The [OIII] line emission at 5007Å is the strongest emission from a PN; it allows the identification & the measurement of its radial velocity

•We obtain PN number density distribution and 2D radial velocity fields in regions where the stellar surface brightness is too faint with respect to the night sky!

ICPNe in the Vigo cluster: projected phase-space distribution

ON-HON-[OIII] OFF-(V+R)


ICPNe as light tracersICPNe as light tracers:1. Narrow band imaging surveys with large field mosaic

cameras (WFI@ESOMPI 2.2m tel. + SuprimeCam@SUBARU) 2. Development+tests of selection criteria based on

photometric catalogs from Sextractor (Arnaboldi et al. 2003, ESO

Messenger 112, 37) for the identification of the ICPNe associated with the ICL (Arnaboldi et al. 1996, 2002, 2003; Okamura et al. 2002; Feldmeier et al. 1998, 2002, 2003, 2004; Aguerri et al. 2005, AJ,

129, 2585).

Planetary Nebulae as tracers of cluster evolution

ICPNe as kinematical tracersICPNe as kinematical tracers:1. Follow up studies aiming at measuring spectra from multi-slit

spectrograph (FORS2 @ ESO VLT; Gerhard et al. 2002, ApJ, 580, L121, Arnaboldi et al. 2003, AJ, 125, 514)

2. First spectra from multi-fiber spectrograph for a statistical significant sample of ICPN (FLAMES @ ESO VLT; Arnaboldi et al.

2004, ApJ, 614, L33; ESO PR 24/04)


Mosaic image in [OIII] obtained at the WFI@ESO-MPI 2.2m tel.

Current surveys

Layout of the fields already acquired within this project in the Virgo cluster (Map of Virgo from Binggelli et al. 1987).

Mosaic image in [OIII] obtained with the SuprimeCam @Subaru 8.2m tel.


Results from narrow band surveysThe ICPNe number density distribution in

the Virgo cluster is highly inhomogeneous:• inhomogeneous distribution in single fields,• field-to-field number density fluctuations.

The distribution of ICPNe in the Subaru field is highly inhomogeneous

No ICPN in LPC. Agreement with Kud+2000

HST RGBs. From Aguerri, Gerhard, Arnaboldi & al. 2005, AJ, 129, 2585


Results from narrow band surveys (cont.)

Significant field-to-field variation of the ICL in the Virgo clusterNo clear number density trend with distance from cluster center in M87Mean surface luminosity density of ICL in Virgo core : 2.7 106 L ’�(note the large rms = 2.1 106 L ’ � due to field-to-field variations!) Mean surface brightness of ICL in Virgo core : B=29 mag ”�Mean fraction of light in ICL: <10 % (Aguerri et al. 2005; see also Feldmeier et al. 2004)Fraction of stars in the ICL increases with the density of the environment: <2% in loose groups (Castro-

Rodriguez+2003, Durrel et al. 2004), <10% in Virgo-like (Arnaboldi et al. 2003, Feldmeier et al. 2004,

Aguerri et al. 2005), 20% in rich clusters including cD halos (Gonzalez et al. 2000, Feldmeier et al. 2002, Zibetti et al. 2005).


Results from spectroscopic follow-up….

April 2002 FORS2 @ VLT & DOLORES @ TNG (in the Subaru field only)

March 2003 FLAMES @ VLT for FCJ, Core and Subaru fields

Flames FOV.

FORS2 & Flames of ICPNe spectra show the [OIII] doublet!

ICPN single spectraM.Arnaboldi et al. 2004, ApJ, 614, L33

Ly


ICPNe: they are brighter than the PNLF cut off for M87

M87

PN in the M87 halovmean = 1280 km/s=240 km/s

M86

M84

NGC 4388


-150

797

75020971151

1191

7301025

3049

721

2373

671

226

Observed ICPN radial velocities in the Subaru field

Flames FOVFlames FOV

I34, I35, I36, I38 are over-luminous [OIII] emitters & are bound to M84 – Pop. Effects.


Implications on adopted criteria

• We understand the selection biases that may lead to wrong identification, i.e. continuum objects erroneously classified as emission line, in on-off imaging surveys on mosaic frames.

• The fraction of spectroscopically confirmed targets is now f >80 % (down to the limiting magnitudes), while in Freeman et al. (2000) was 50%!

Summary spectroscopic results!

•Giant galaxies in clusters are very extended - PNe still associated with M87 out to ~70 kpc & very extended halo also around M84

•The velocity histograms show strong field-to field variations.

•Dynamical times at the location of these fields are from 2× 108 yr (FCJ) to 8×108 yr (SUB). Phase mixing to erase field-to-field variations would take few Gyr.


Conclusions

1. Observations indicate that diffuse light is important in understanding cluster evolution, star formation history and the ICM enrichment.

2. Measuring the projected phase space constraints how and when this light originated and ICPNe are the only abundant stellar component of the ICL whose kinematics can be measured.

3. We can then explore the effect of low/dense environment on galaxy evolution with MSIS technique! (see O. Gerhard’s talk, this conference. )

intracluster planetary nebulae as dynamical probes of the diffuse light in galaxy clusters

Documents

coma cluster

cluster centre

feedbackstudies of icl

importance of icl

cluster collapse

discovery of diffuse

large galaxies

small galaxies