Ultra-Compact Dwarf Galaxies: Dark Matter Dominated Systems? or Simple Stellar Clusters?

An Invited Review presented at the 1st California Astronomy Postdoc Symposium, University of California, Santa Cruz, 22nd August 2008.

Conference website: http://www.stsci.edu/~jkalirai/CaliPostdocSymposium/postdoc.html

Transcript

1. 1st California Postdoc Symposium, UC Santa Cruz, August 21 & 22 2008. Arna Karick UC Davis/IGPP, LLNL Friends of UCDs: - Michael Gregg (UC Davis/LLNL) - Mike West (ESO Chile) - Michael Drinkwater, Peter Firth & Katya Evstigneeva (UQ) - Michael Hilker (ESO Garching) - Kenji Bekki (UNSW) - Bryn Jones (Queen Mary, London) - Steve Phillipps (Bristol) Ultra-Compact Dwarf Galaxies Dark Matter Dominated Systems? or Simple Stellar Clusters?
2. The UCD story in 30 minutes or less… Over the past decade, a new class of compact stellar systems have been discovered in galaxy clusters….  First Discoveries  Fornax Cluster Spectroscopic Survey  Follow-up Observations: HST/STIS and VLT/UVES & Keck/ESI  The UCD Rush  A few words on nomenclature  Summary of General Properties  Recent and Ongoing Investigations
3. 1995: West et al. Galaxy clusters contain N IGC ∝ M cl Distribution follows the cluster mass distribution 1996: Minniti et al. Small survey of NGC1399 globular clusters Object near NGC1404 is identified as a very compact dwarf galaxy: V=18.5, (V– I) = I.48 1999: Hilker et al. Imaging of Fornax Cluster galaxies. Two compact objects are spectroscopically confirmed to be cluster members. These are later re-discovered in the FCSS. “..can be explained by a a very bright GC or a compact elliptical like M32. Another explanation may be that these represent the nuclei of dwarf elliptical galaxies.” - SE region (30’ across) of the Fornax Cluster. (at d=20 Mpc, 30’~183 kpc) - NGC1399 (central cluster galaxy): vrad =1425 kms-1 - brightest compact object is near NGC1404 (vrad =1947 kms-1) and has vrad =1591 ± 36 kms-1. Image: CTIO 4m MOSAIC (g’,r’,i’), Karick, Gregg & Drinkwater The First Discoveries
4. Image: ~3x2 degree B,V,I mosaic, taken with the CTIO Curtis Schmidt Telescope (Karick, Gregg & Drinkwater ) Fornax Cluster Spectroscopic Survey 1999 – 2001: FCSS: Drinkwater et al. 2000, Phillips et al. 2001 Galactic stars Fornax all-object spectroscopic survey of ~14,000 resolved and unresolved objects (selected from the APM catalogue) in four Fornax Fields. ~340 Fornax galaxies +5 new cluster members identified within a two degree region, centered on NGC1399. QSOs Galaxies
5. Image: ~3x2 degree B,V,I mosaic, taken with the CTIO Curtis Schmidt Telescope (Karick, Gregg & Drinkwater ) V=1491 kms-1
   
   V=1312 kms-1
   
   V=1591 kms-1
   
   V=1920 kms-1
   
   V=1355 kms-1
   
   Luminosities: −13.5 < M V < −12.3 mag [NGC1399 GCs: −11.5 < M V < −6 mag] Velocities: 1312 kms-1 < v rad <1920 kms-1 Sizes: < 100pc (complete at these luminosities) ~150 kpc Early studies predicted GC populations to extend further out (Forbes et. Al 1997) ~650 spectroscopically confirmed GCs (~450 Dirsch et al. 2004, +200 Schubeth et al.)
6. HST/STIS imaging of UCDs HST/STIS imaging of 5 UCDs and a comparison Fornax Cluster dE,N, FCC303 (Drinkwater et al. 2003) UCD profiles are well fitted by King and de Vaucouleurs R1/4 law profiles. Effective radii range from, r e =10–22 pc. UCD3, the most luminous UCD, required a 60 pc scale length exponential halo. FCC303 (dE,N) has an 8 pc core and a 300 pc halo. Most compact Virgo dwarfs have scale lengths of 160pc. FCC303 UCD1 UCD2 UCD3 UCD4 UCD5 Galactic GCs have typical effective radii of r e = 3–5 pc. Largest M31 GCs have M/L ~ 1– 2 M King Model Mass estimates: UCDs: 1 – 5 x 107 M, FCC303 (nuc): 1.4 x 107 M, most massive GCs: 4 x 106 M → UCDs have M/L ~ 2 – 4 M
7. VLT/UVES & Keck/ESI Spectroscopy VLT/UVES and Keck/ESI Spectroscopy of 5 UCDs and a comparison dE,N, FCC303 (Drinkwater et al. 2003, Hilker et al. 2006 & Evstigneeva et al. 2007) Velocity dispersions range from, σ = 24–37 km s-1 considerably higher than Galactic GCs (~5-10 kms-1). UCDs lie well off the GC: L∝σ1.7 relation in a previously unoccupied region, but on an extrapolation of the elliptical galaxy: L∝σ4 Faber-Jackson relation. Galaxies? (DM) or Stellar Clusters? (no DM?) GALAXY THRESHING (Bekki et al. 2001, 2003) The location of the UCDs and dE,Ns supports the “galaxy threshing” model. A normal dE,N halo, accounting for ~98% of the dE,Ns luminosity is tidally stripped, reducing the total luminosity by a factor of 100 but barely changing the central velocity dispersion. Over time this disruption contributes diffuse stellar material to the ICM. The surviving nuclei are dispersed into intracluster space or added to the envelopes of brighter galaxies where they can masquerade as stars. Evstigneeva et al. 2006
8. The Fornax Cluster 20 Mpc away, z ~ 0.0043 ~340 “likely” members in the FCC 186 morphologically classified as dwarf galaxies (Ferguson 1989) Velocity dispersion: giants 308 ± 30 kms-1 dwarfs 429 ± 41 kms-1 (infalling?) Evidence for cluster substructure: Chandra: Scharf et al. (2004) FLAIR: Drinkwater, Colless & Gregg (2001) 6dF: Karick et al. (2008, sub) (radial) infall region: galaxies are unlikely to be interacting gas rich dwarfs dominate (Waugh, PhD 2006) high SFR in dwarf galaxies (Drinkwater et al. 1999) Central core: hot gas and high densities dEs and Es dominate galaxies likely to interact Top: optical image and SW subcluster from the Flair analysis. Bottom: soft emission (0.3-1.5 keV)
9. The UCD Rush  First simulations of “galaxy threshing” (Bekki et al. 2000, 2001)  Fornax Compact Object Survey: ~50 “compact objects” (Mieske et al. 2002,04) → First metallicity measurements, [Fe/H] avg ~ –0.5 dex  Follow-up 2dF spectroscopy of Fornax: 60 more UCDs (Gregg et al. 2008) → Further investigation of the stripping hypothesis.  Deep u,g,r,i,z imaging of Fornax UCDs (Karick, PhD 2005) → Color analysis and candidate selection for follow-up spectroscopy  Follow-up VLT/FLAMES spectroscopy: 30 more UCDs (Firth et al. 2007) : 150 UCDs/IGCs (Bergond et al. 2007)  Keck/LRIS and Magellan/IMACS spectra of 90 Fornax UCDs,GCs and dE,Ns → Significant vairations in stellar population (Gregg & Karick, in prep)  10 UCDs discovered in Abell 1689 appear to be stripped dwarf galaxies. → two objects have properties similar to M32.  Massive GCs found in CenA but no UCDs (Martini et al. 2004)  6 “Dwarf-Globular Transition Objects” (DGTOs) discovered near M87 (ACSVCS) → highest M/L ~ 6 – 9 M, cannot be explained by SSP (Hasegan et al. 2005)  9 UCDs discovered in Virgo from 2dF observations (Jones et al. 2006)  Survey of 6 nearby galaxy groups - no UCDs (Evstigneeva et al. 2007) ‘99 – ‘01 ‘02 – ‘06
10. The UCD Rush  First simulations of “galaxy threshing” (Bekki et al. 2000, 2001)  Fornax Compact Object Survey: ~50 “compact objects” (Mieske et al. 2002,04) → First metallicity measurements, [Fe/H] avg ~ –0.5 dex  Follow-up 2dF spectroscopy of Fornax: 60 more UCDs (Gregg et al. 2008) → Further investigation of the stripping hypothesis.  Deep u,g,r,i,z imaging of Fornax UCDs (Karick, PhD 2005) → Color analysis and candidate selection for follow-up spectroscopy  Follow-up VLT/FLAMES spectroscopy: 30 more UCDs (Firth et al. 2007) : 150 UCDs/IGCs (Bergond et al. 2007)  Keck/LRIS and Magellan/IMACS spectra of 90 Fornax UCDs,GCs and dE,Ns → Significant vairations in stellar population (Gregg & Karick, in prep)  10 UCDs discovered in Abell 1689 appear to be stripped dwarf galaxies. → two objects have properties similar to M32.  Massive GCs found in CenA but no UCDs (Martini et al. 2004)  6 “Dwarf-Globular Transition Objects” (DGTOs) discovered near M87 (ACSVCS) → highest M/L ~ 6 – 9 M, cannot be explained by SSP (Hasegan et al. 2005)  9 UCDs discovered in Virgo from 2dF observations (Jones et al. 2006)  Survey of 6 nearby galaxy groups - no UCDs (Evstigneeva et al. 2007) ‘99 – ‘01 ‘02 – ‘06
11. The UCD Rush  First simulations of “galaxy threshing” (Bekki et al. 2000, 2001)  Fornax Compact Object Survey: ~50 “compact objects” (Mieske et al. 2002,04) → First metallicity measurements, [Fe/H] est ~ –0.5 dex  Follow-up 2dF spectroscopy of Fornax: 60 more UCDs (Gregg et al. 2008) → Further investigation of the stripping hypothesis.  Deep u,g,r,i,z imaging of Fornax UCDs (Karick, PhD 2005) → Color analysis and candidate selection for follow-up spectroscopy  Follow-up VLT/FLAMES spectroscopy: 30 more UCDs (Firth et al. 2007) : 150 UCDs/IGCs (Bergond et al. 2007)  Keck/LRIS and Magellan/IMACS spectra of 90 Fornax UCDs,GCs and dE,Ns → Significant vairations in stellar population (Gregg & Karick, in prep)  10 UCDs discovered in Abell 1689 appear to be stripped dwarf galaxies. → two objects have properties similar to M32.  Massive GCs found in CenA but no UCDs (Martini et al. 2004)  6 “Dwarf-Globular Transition Objects” (DGTOs) discovered near M87 (ACSVCS) → highest M/L ~ 6 – 9 M, cannot be explained by SSP (Hasegan et al. 2005)  9 UCDs discovered in Virgo from 2dF observations (Jones et al. 2006)  Survey of 6 nearby galaxy groups - no UCDs (Evstigneeva et al. 2007) ‘99 – ‘01 ‘02 – ‘06
12. The UCD Rush  First simulations of “galaxy threshing” (Bekki et al. 2000, 2001)  Fornax Compact Object Survey: ~50 “compact objects” (Mieske et al. 2002,04) → First metallicity measurements, [Fe/H] avg ~ –0.5 dex  Follow-up 2dF spectroscopy of Fornax: 60 more UCDs (Gregg et al. 2008) → Further investigation of the stripping hypothesis.  Deep u,g,r,i,z imaging of Fornax UCDs (Karick, PhD 2005) → Color analysis and candidate selection for follow-up spectroscopy  Follow-up VLT/FLAMES spectroscopy: 30 more UCDs (Firth et al. 2007) : 150 UCDs/IGCs (Bergond et al. 2007)  Keck/LRIS and Magellan/IMACS spectra of 90 Fornax UCDs, GCs and dE,Ns → Significant vairations in stellar population (Gregg & Karick, in prep)  10 UCDs discovered in Abell 1689 appear to be stripped dwarf galaxies. → two objects have properties similar to M32.  Massive GCs found in CenA but no UCDs (Martini et al. 2004)  6 “Dwarf-Globular Transition Objects” (DGTOs) discovered near M87 (ACSVCS) → highest M/L ~ 6 – 9 M, cannot be explained by SSP (Hasegan et al. 2005)  9 UCDs discovered in Virgo from 2dF observations (Jones et al. 2006)  Survey of 6 nearby galaxy groups - no UCDs (Evstigneeva et al. 2007) ‘99 – ‘01 ‘02 – ‘06
13. Image: ~3x2 degree B,V,I mosaic, taken with the CTIO Curtis Schmidt Telescope (Karick, Gregg & Drinkwater ) Gregg et al.2008 60’ Fornax Cluster Fornax
14. 9 objects widely distributed around M87 with luminosities (-13.7 < Mv < -11.5) similar to the Fornax objects. (Jones et al. 2006) 6 objects (+13 candidates) close to M87. Three have M/L ratios of 6-9 M which cannot be explained by SSP models. (Hasegan et al. 2006) 4 candidate IGCs near M87 from HST/ACS imaging. Fainter than Fornax UCDs and metal-poor. (Williams et al. 2006) Another 6 UCDs between M87 and M49 (Firth et al. 2008) Virgo Cluster
15. A few words on nomeclature.. With each new investigation comes a new acronym: …not to be confused with Ultra-faint, Ultra-handsome Dwarf Galaxies  Ultra-Compact Dwarf Galaxies (UCDs)  Ultra-Diffuse Star Clusters (UDSs)  Compact Cluster Objects (CCOs)  Compact Cluster Systems (CCSs)  Old Massive Clusters (OMCs)  Compact Stellar Systems (CSS)  Dwarf-Globular Transition Objects (DGTOs)  “…UCDs or PYFA’s” -- E. Werhner, CEDGSC, Munich July 2008  Massive Compact Objects (MCOs) “… definition of MCOs includes a number of objects that are considered as UCDs in other works..” (Dabringhausen et al 2008)
16. M31 (Andromeda) Globular Clusters M31: - 1012 M roughly 80% mass of the MW - roughly 460 GCs Back seat bandits: Mayall-II/G1 : - massive (10 x 106 M) x2 mass of ωCen - metallicity spread (several generations of stars) core of a tidally stripped dwarf galaxy? - large ellipticity Mayall II is named after Nicholas U. Mayall and O.J. Eggen in 1953. Sidney van den Bergh named it G1 in 1977.
17. Milky Way Globular Clusters derived from the Latin word globulus - a small sphere - roughly 160 globulars (GCs) in the Milky Way - 90% lie within 100,000 light-years (0.5r stellar halo) - single age (old ~13 Gyr), single (low) met populations - (all stars form at the same time) - Very little or no HI gas and dust - (stars + stellar winds drive out gas) - M/L ratios ≈ 1 (unlike galaxies - NO DARK MATTER) - typical effective radii of re = 3-5 pc Back seat bandits: Omega Centauri (wCen) - old (12 Gyr), massive (5 x 106 M) - metallicity spread (several generations of stars) - core of a tidally stripped dwarf galaxy? - contains a 4 x 105 M BH (Gemini observations) M15 : massive and most dense GC (core collapse), probably contains a central black hole.
18. General Properties “UCDs - a mixed bag of objects” Hillker et al. 2006
19. Unusually luminous globular clusters. (Hilker et al. 1999; Mieske et al. 2002) Extremely luminous star clusters formed from young clusters during galaxy mergers. (Fellhauer & Kroupa 2002, Maraston et al. 2004) Remnant nuclei of dE,Ns or low luminosity spirals which have been tidally disrupted or “threshed” during passages close to the central cluster galaxy, NGC1399. (Bekki et al. 2001, 2003, Goerdt et al 2008) Extreme cases of known galaxies. (e.g. dE,Ns with very faint halos) Highly compact galaxies formed in the early Universe in their own DM halos? (Blanchard et al. 1992, Tegmark et al. 1997) -smoothness of GCLF support this scenario. - break in the L vs. R eff relation? collpase physics - is there an upper mass limit for normal GCs? - is there a turnover point on the L ∝ σ 1/4 relation? - UCDs once resided in DM halos which were stripped - IGCs should be more concentrated than dEs - requires low σ, radial orbits and very compact cores - ωCen, M22, G1? - final UCD may contain dark matter but is not DM dominated - no dark matter? Formation Scenarios - would have dark matter
20. …back seat bandits UCD1 has strong Mg, Hb absorption as well as a strong CN feature, similar to NGC1399 -- larger more metal-rich system? Two UCDs/IGCs are X-ray point sources in the Chandra Fornax Survey. L=2 x108 ergs-1 (Scharf et al. 2005) Consistent with a NS LMXRB Preliminary analysis of IMACS spectroscopy by Marla Geha showed variation in line indices. CTIO imaging (Karick et al 2008)and ACS imaging (Puzia &Hilker 2008,in prep) shows some red clusters are extended.
21. Stellar Populations of UCDs Fornax Virgo  Lick/IDS metallicites; Mgb, <Fe>, Hβ gives [α/Fe]. Also CaT measurements  Virgo UCDs are older (>10Gyr) on average than dE,N nuclei and have lower metallicities,[Z/H] of -1.35 to -0.5 dex  Fornax UCDs are more metal-rich, [Fe/H] ~ -0.9 to -0.5 ± 0.5 dex  Similar to GC results in M49 and M87 (Cohen et al. 1998, 2003)  In Virgo the threshing scenario seems less plausible.
22. Hasegan et al. (2005) found similar objects in Virgo - DGTOs Metallicities: –1.5 < [Fe/H] < –0.7 dex Velocity dispersions: 20 — 30 kms-1 and M/L ratios: 6 - 9 M The largest GCs in M31 have M/L ~ 1 — 2 M VLT/UVES Spectroscopy : Four Fornax UCDs + dE,N FCC303 (Hilker et al. 2006) Velocity dispersions of UCDs: σ = 22.9 ± 2.6 — 29.2 ± 1.5 kms-1 Velocity dispersion of FCC303: σ = 27.1 ± 3.1 kms-1 Dynamical masses of UCDs: Mmod = 1.8 — 3.73 x 107 M [UCD3: 9.45 x 107 M] Dynamical M/L ratios: M/L = 3.15 — 4.99 (consitent with SSP M/L) Dynamical Masses and M/L ratios But high M/L ratios can be explained by tidal “heating”. Fellhauer & Kroupa (2006) High M/L ratio’s CAN be reduced by adopting various stellar populations models with differing IMFs
23. Dynamical Masses and M/L ratios An extreme IMF as an explanation for high M/L ratios in UCDs? The CO index as a tracer of bottom heavy IMF (Mieske & Kroupa 2008) Salpeter IMF vs. Kroupa IMF ? (Dabringhausen et al 2008) ….Relaxation time seems to be important
24. UV-metallicity properties of Fornax UCDs/IGCs The FUV is sensitive to the excess UV emission produced by EXTREME horizontal branch stars (EBH). This “UV-excess” is observed in many early-type galaxies, ωCen and some M87 GCs. What are the UV — optical properties of UCDs in Fornax? & Do Fornax UCDs exhibit a UV — Mg2 or [Fe/H] relation? Work with Michael Gregg, Soo-Chang Rey & Marla Geha (preliminary results)
25. GALEX Detections: UV detections: detect sources using SExtractor & compared to GALEX catalogue photometry. g,r,i,z UCD photometry: Used more accurate coordinates from our CTIO g,r,I,z imaging to match to GALEX detections (< 4” radii) 100 catalogued UCDs/IGCs within the GALEX field GALEX NUV (1344-1786Å ) & FUV (1771-2831Å) archival images: 1.2 deg FOV with 1.5” pixels (~4.5” FWHM)
26. 100 UCDs/IGCs/GCs need to be careful!
27. Optical Photometry  accurate ACS/HRC V & I (F606W & F814W) photometry of 12 UCDs (Evstigneeva et al. 2008)  CTIO 4m-Blanco MOSAIC imaging of the core of Fornax g,r, i,z photometry of UCDs/GCs (Karick et al. 2008, PhD thesis)
28. UV — optical properties of UCDs ...preliminary results Fornax UCDs (red points) appear to be similar to the UV-bright clusters of M87 (black points). → He - enrichment? → enhanced mass loss of old stars? → something else? g,r,i,z & V, I converted to HST STMAG GALEX (FUV - V) to HST (FUV-V)STMAG similar UV-optical analysis by Mieske et al. 2008 (in press) Sohn et al. (2006)
29. Sohn et al. (2006) IMACS spectroscopy of UCDs/GCs …preliminary results Medium resolution (1.3Å), high S/N (~50) spectroscopy of: 32 Fornax UCDs and 61 GCs to measure line strength indices. Observations taken by Marla Geha. Mean [Fe/H] derived from various absorption lines (Mg2, Fe etc.) using calibrations from Puzia et al. (2002) In progress.. (Gregg &Karick 2008) -- Flux calibrate data -- can measure the strength of the D4000 break to get age estimates -- determine [α/Fe] ratios. -- compare to UV pop models. (may also affect our Fe/H values)
30. UV — Mg2 properties of UCDs ...preliminary results Sohn et al. (2006) Further investigation.. UV-excess vs. UCD mass + metallicity? UV-Sloan photometry -intermediate age populations (with Soo-Chang Rey)
31. Conclusions There is a lot more work to do! Stay tuned for more exciting results. Acknowledgements Thanks UCSC for organising this conference Some of the work reported here was done under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory in part under Contract W-7405-Eng-48 and in part under Contract DE-AC52-07NA27344.