1. Structural Parameters for confirmed Globular Clusters in NGC 5128 Matias Gomez Universidad de Concepcion - Chile Collaboration with: K. Woodley, W. E. Harris (McMaster), D.Geisler (Concepcion), G.L.H. Harris (Waterloo) “The Globular Clusters - Dwarf Galaxies Connection” Michigan, Aug. 2007

2. Why Structural Parameters? (SPs) 1) Properties of Globular Clusters (GCs) are correlated with properties of their host galaxies (Brodie & Strader 2006) fundamental key to understand galaxy assembly and evolution 2) GCs seem to occupy a well-defined region in the parameter space (  0 ¸c¸  ) (Djorgovski 1995; McLaughlin 2000; McLaughlin & van der Marel 2005) 3) GC formation scenarios! Need to understand the physical conditions that produce outliers high ellipticity of LMC GCs (Geisler & Hodge 1980), N5128 (Rejkuba et al. 2007) Population of bright/extended/compact/etc GCs in some environments and not in others

3. 5) Sizes seem to indicate that a large variety of objects could have a common origin: GCs/UCDs/FFs/DSCs (but not dSph, Gilmore et al 2007 ) 6) distances for free! (Kundu & Whitmore 2001; Jordan et al. 2005) 4) Effective (half-light) radius remains constant along the entire GC life (Spitzer & Thuan 1972; Spitzer 1987; Meylan & Heggie 1997; Aarseth & Heggie 1998) learn from GC formation by using sizes (Jordan 2004; Brodie & Strader 2006)

4. A puzzling difference: blue clusters are, on average, 20% larger than red ones (Kundu & Whitmore 1998; Larsen et al 2001; Kundu & Whimore 2001) What causes this? 1) Mass segregation + metallicity dependence of stellar lifetime under the assumption of metallicity-independent half-mass radii (Jordan 2004; Jordan et al. 2005) 2) Projection effects (Larsen & Brodie 2003; Spitler et al. 2006) little change predicted in relative sizes with R gc size differences should be largest at small R gc and disappear at large R gc Need to study sizes at large R gc

5. 2) In nearly all GC Systems red clusters are more centrally concentrated than blue ones within any given radius, red (metal-rich) GCs will tend to lie at smaller R gc If GCs follow the same correlation found in the MW, red cluster will appear smaller on average than blue ones. r e increases with galactocentric distance 1) In our Galaxy: (van den Bergh et al. 1991) r e ~ √R gc

6. Why NGC 5128? 2) N ~ 2000 (Harris et al. 2007) (reasonable sample in contrast with MW) 4) 3 mag brighter than Fornax & Virgo, 2 mag brighter than Leo 3) Since merger product (cepheid distance! Ferrarese et al. 2007) potentially rich GC Age-Distribution-Function, spanning many Gyr ideal target for addressing many questions 1) d = 3.8 Mpc, nearest giant elliptical (albeit merger product). GCs can be resolved ELT science case

7. However, nearby is good and bad... 1) GCS very spread out (individual GCs found out to R gc = 40', Peng et al. 2004b ) 2) Low galactic latitude (b~19°) and contamination is a serious problem (Halley and NGC5128 - ©A. Gomez, CTIO) a) blue compact, faint background galaxies b) a rich population of foreground objects

8. Plus... NGC 5128 has its reputation!

9. Previous work: (just a few examples) - HST structural parameters (Holland et al. 1999, Harris et al. 2002, Harris et al. 2007 ACS) - Peng et al. 2004, 4m CTIO, BVI - Harris et al. 2004, 4m CTIO, CMT1 (most extensive study) Needed: a clean sample of GCs from the 10 5 sources in the field (2%) (Dirsch et al. 2003) reduced FOV less than ideal seeing

10. 1.2° Image taken from DSS (North up, East to the left) 25 fields observed at Las Campanas Observatory (Chile) with the Magellan 6.5m and the IMACS camera (0.11“/pix). Seeing between 0.4" and 0.7". Short and long exposures in B,R. This work: IMACS Observations

11. We were able to derive structural parameters for ~370 GCs (no contamination, homogeneous sample, large FOV, very good seeing) Peng et al. (2004) Woodley et al. (2005) Beasley et al. (2007) Harris (2007, ACS imaging) combining with existing Washington photometry from Harris et al. (2004) colours, magnitudes and metallicities Confirmed GCs taken from: + new Magellan/LDSS-2 data + Convolution techniques (ishape, Larsen 1999, 2001)

12. Some results: 1) good 1:1 comparison with entirely independent HST/ACS sizes (different approach too) 2) Red/blue subpopulations: usual behaviour (but note higher dispersion for metal-rich GCs)

13. relative sizes of blue and red cluster depends on R gc (flattening of red GCs at ~1.5 times r eff of NGC 5128)

14. Dividing into inner/outer metal-poor/metal-rich subpopulations: size difference in agreement with usually observed trend at small R gc No difference at large R gc (sample extending up to ~8 times r eff of NGC 5128) Results in favour of projection effects as the origin of size difference (Spitler et al. for NGC4594) Note also the lack of large blue clusters at large R gc !

15. Conclusions: 1) 0.5” high resolution imaging across a 1.2°x1.2° field Coming attractions: 1) GMOS ages and metallicities of a large sample of GCs (PhD thesis of K. Woodley) 2) VIMOS: homogeneous PA coverage (e.g. along minor axis) 2) sizes and SPs for ~370 confirmed GCs out to R gc = 40' (~45 kpc) 3) size difference between metal-rich and metal-poor disappears at large R gc Projection effects are enough to explain difference

16. thanks !!

17. and... South !!