1. 1 Paleontology of Galaxies Recovering Star Formation & Chemical Enrichment Histories from galaxy spectra Roberto Cid Fernandes UFSC – Florianópolis -Brasil

2. 2 STARLIGHT & its many applications Spectral fits & star-formation histories Asari et al 2006 CF et al 2005 CF et al 2004

3. 3 STARLIGHT & its many applications UCBD galaxies Corbin et al 2006

4. 4 STARLIGHT & its many applications Emission line work (from residual spectra) Stasinska et al 2006

5. 5 STARLIGHT & its many applications HE0450-2958 – The “homeless” QSO CaII Triplet velocity dispersions Vega 2004, Garcia- Rissman et al 2005 Merritt et al 2006

6. 6 etc... Someone al 2007, 2008,...

7. 7 Flori-where?

8. 8 The team(s) Techniques:L Sodré, J Gomes, Merlin, Harry Potter,... Seyfert 2s:Q Gu, J Melnick, E & R Terlevich, D Kunth LLAGN:R González Delgado, E Pérez, H Schmitt, L Martins, T Storchi-Bergmann SDSS: A Mateus, L Sodré, G Stasinska, J Gomes, N Asari, W Schoenell, L Vega, J P Papaqui Ca-Triplet:L Vega, N Asari, A Garcia-Rissman,... WR-galaxies: J Leão, C Leitherer UCBD-galaxies:M Corbin, W Vacca,... The homeless QSO:D Merrit, T Storchi-Bergmann, D Axon,...

9. 9 The talk 1- STARLIGHT: Spectral fits using evolutionary synthesis models How? What it can(’t) do for you Miscelaneous Applications:2 - Seyfert 2s 3 - Low Luminosity AGN 4 - CaT-synthesis 5 - UCBDGs 6 - SDSS:Fits for 582471 galaxies! Mass, age, Z *,  *,... Correlations, correlations & correlations... Star-Formation & Chemical enrichment Histories...

10. 10 The Conclusions 1 STARLIGHT: It works!! Recovers M *,,,  *, A V,... And pretty decent SF & chemical histories too! 2 Seyfert 2s: Heterogeneous Star-Formation history Fits uncover reflected AGN! (BLR + continuum) 3 LLAGN: LINERS are all OLD & boring Transition Objects = OLD or “YOUNG” Young-TOs are dusty 10 8 –10 9 yr post-starbursts... 6 SDSS: Beautiful results, but too much info to digest!!!

11. 11 Spectral synthesis of integrated stellar populations: “...a subject with bad reputation. Too much has been claimed, and too few have been persuaded.” (Searle, 1986)

12. 12 1 – STARLIGHT fits: Examples

13. 13 1 – STARLIGHT: The recipe  ’s (+ gas + dust +...) ≈  SSP’s x 10 -0.4 A( )  x 1 + x 2 + x 3 +...

14. 14 1 – STARLIGHT: The ingredients 150 SSPs from Bruzual & Charlot (2003) 25 ages: t = 1 Mr – 18 Gyr 6 metallicites: 0.005 – 2.5 Z o 1 Reddening law (Cardelli/Calzetti/LMC/SMC) 1 Gaussian LOSVD: G(v *,  * ) 2 or 3 smart collaborators/students lots of CPUs... (~ 150 for ~ 6 months!)

15. 15 1 – STARLIGHT: The ingredients

16. 16 1 – STARLIGHT: The parameters x = x(t,Z) = light FRACTION at  from population t,Z = “population vector” = (x 1,x 2,...x N ) = Star Formation History Other parameters: A V = dust (only 1 so far...) v * = “redshift”  * = velocity dispersion

17. 17 1 – STARLIGHT: The nuts & bolts + Movies! 08, 32, 39 Data: O ±  ; = 3400 – 8900 Ǻ – observed spectrum Model: M (x,A V,v *,  * ) – model spectrum =  SSPs Problem: How to estimate x 1...x N, A V, v * e  * ? ~ 150 + 3 = 153 parameters!  Method: Markov Chain Monte Carlo Metropolis-Hastings + Simulated-annealing Likelyhood guided “quasi-random” walk Prob(x,A V,v *,  * | Data) ~ exp –{  2 / 2}  2 =  {O – M (x,A V,v *,  * )} 2  2

18. 18 Simulations: output ~ input

19. 19 Simulations: but... “AZD” 

20. 20 1 – STARLIGHT: The future + Public version by 2006.9999 (code + SDSS fits) + 2006.9:  -enhanced SSPs – essential for Ellipticals! (Coelho et al, …)... Not so easy... + 2005.5: Nebular continuum.... Done + 2005.9: More than 1 extinction... Done, but...  + 3001.2: Smarter/faster method to explore huge parameter spaces (adaptative-MCMC, Genetic-MCMC, …)

21. 21 CF, Gu, Melnick, Terlevich 2, Kunth, Rodrigues Lacerda, Joguet 2004, MNRAS 2 – The SF-History of Sey 2 nuclei 79 galaxies 65 Sey 2s ~ 200 pc Base = BC03 + FC Strong FC in this Sey 1

22. 22 Result #1: SFH of Sey 2s is very heterogenous! Continuous SF Instantaneous Bursts Power-Law + E-gal “Broad Line Sey 2s” or “Sey 3s” ?!

23. 23 Result #2: Weak (scattered) BLR & FC! Starlight-subtracted spectra reveal weak broad H  in 10/65 Seyfert 2s! Most of these have Hidden BLRs revealed in polarized spectra (Tran 95, 01) Spectral synthesis finds a strong FC component (~ 20%) in these “Broad Line Seyfert 2s” Conclusion: Scattered light!

24. 24 Long-slit spec. of ~ 50 LLAGN Spatial Resolution ~ 1” ~ 100 pc Spectral synthesis of 521 extractions Base of template galaxies representing Y, I & O populations St pop, extinction & brightness profiles 3 – LINERs & TOs CF + Gonzalez Delgado et al 2004, 2005 Gonzalez Delgado + CF et al 2004

25. 25 x(r) A V (r) 3 – Dust in Young TOs: A V (r)

26. 26 2 + 3 = Evolution of (stars in) AGN? = To be continued....

27. 27 4 – Calcium Triplet work A Garcia Rissman (LNA), L Vega (Córdoba/UFSC), N Vale Asari (UFSC), CF, H Schmit (NRL), R González Delgado (IAA), T Storchi-Bergmann (UFRGS) Garcia-Rissman et al 2005, MNRAS + Vega et al 2006 (in prep) + Barbosa et al 2006 (IFU => v-field) Different application of the same code:  *

28. 28 4 – Calcium Triplet work Garcia-Rissman et al 2005 Fits with a base of observed stars (kinematical templates) plus a fake continuum

29. 29 5 – Ultra Compact Blue Dwarf galaxies Corbin et al 2006

30. 30 5 – Ultra Compact Blue Dwarf galaxies Corbin et al 2006 Fits with: - 25 ages - Z 0 /50 SSPs - SMC ext curve - Nebular Cont.

31. 31 5 – Ultra Compact Blue Dwarf galaxies Corbin et al 2006 age Light dominated by young stars But most mass is in old generations NOT primaeval!

32. 32 6 – Synthesis of 582k SDSS galaxies N Asari, J Gomes, W Schoenell, J P Papaqui (UFSC) A Mateus (IAG), L Sodré (IAG) & G Stasinska (Meudon) The SEAGal Collaboration: Semi-Empirical Analysis of Galaxies

33. 33 6 – Synthesis of 350k SDSS galaxies A journey through the fascinating (but scary) world of Mega Data Bases

34. 34 Products of the synthesis x = x(t,Z) : Star-formation & chemical histories M * : stellar Mass t * & Z * : mean stellar age & stellar metallicity  * : velocity dispersion A V : extinction Observed – Model F( ) = pure emission spectrum Emission line fluxes Nebular metallicity, extinction,... 6 – SDSS: 1 st results (CF et al 05, MNRAS)

35. 35 Stellar versus nebular extinction A V (Balmer) = 0.3 + 1.8 A V (Stellar) Normal Emission Line Galaxies (= non-AGN)

36. 36 Stellar age x stellar age indicator

37. 37 Stellar age x nebular age indicator

38. 38  :  * < 70 km/s = 0! Stellar Mass x velocity dispersion

39. 39 Stellar X nebular metallicity! Normal Star Forming Galaxies (= non-AGN) Very interesting prospects for chemical evolution studies... Confirmed, but NOT cited, by a recent paper in MNRAS 

40. 40 Stellar age x stellar mass More massive galaxies are older. “Downsizing” “Anti-Hierarchical” scenario too old too faint...

41. 41  -enhancement in massive gals.

42. 42 Going one step further: SFHs Idea: Dissect the SFH = SFR(t) along the left wing of the Seagull (normal SF galaxies) angle  ~ Z(gas)

43. 43 Going one step further Mass X  ~ Z(gas)Z(stars) X  ~ Z(gas)

44. 44 Going one step further Small, ~ HII galaxies Dominated by ~ 1 burst Big, ~ Starburst nuclei Mixture of bursts mean age

45. 45

46. 46 6 – SDSS: bottomline Too much information! Parameter But great results!! Can’t be too wrong! brigad !! ?

47. 47

48. 48 A nagging question: Which base? Results change if one allows for Z < 0.2 Z 0 populations in the base! 1/5 < Z < 2.5 1/200 < Z < 2.5

49. 49 HE0450-2958 Merritt et al 2006 The “homeless” QSO – not homeless! ULIRG companion galaxy 100 Myr burst ~ interaction age But not homeless afterall... Just a NLSy1 with a wrongly calculated Black Hole mass! QSO star Magain et al 2005

50. 50 5 – Ultra Compact Blue Dwarf galaxies Corbin et al 2006

51. 51 5 – Ultra Compact Blue Dwarf galaxies Corbin et al 2006 Fits with: - 25 ages - Z 0 /50 SSPs - SMC ext curve - Nebular Cont.

52. 52 Coarse (but robust) description of the st pop INTERMEDIATE 10 8-9 yr OLD 10 10 yr YOUNG (+ FC) 10 6–7 yr Population vector: Reduce x from N * ~ 45 => 3 components x = (x Y,x I,x O ) x Y + x I + x O = 1 plane CF et al 01, 03, 04, GD et al 04

53. 53 Result #1: SFH of Sey 2s is very heterogenous! Population Vector: x = flux-fraction x = ( Y+FC, I, O ) Y/FC = 5 Myr + -1.5 FC I = 100 Myr – 1 Gyr O = 1 – 10 Gyr Old pop + burst

54. 54 Result #1: SFH of Sey 2s is very heterogenous! ~ 40% Starburst + Seyfert 2 composites “pure/boring” Seyfert 2

55. 55 W K X [OI]/H  3 – LINERs & Transition Objects OldYoung Young TOs Old TOs Old LINERs Young LINERs AGN CF et al 04, 05 GD et al 04, 06...

56. 56 Long-slit spec. of ~ 50 LLAGN Spatial Resolution ~ 1” ~ 100 pc Spectral synthesis of 521 extractions Base of template galaxies representing Y, I & O populations St pop, extinction & brightness profiles 3 – LINERs & TOs

57. 57 W K (r) 3 – Radial Gradients in LLAGN

58. 58 x(r) A V (r) 3 – Dust in LLAGN: A V (r)

59. 59 Young-TOs have much more dust than Old-TOs or Old-LINERs Young-TOs Old-TOs & LINERs 3 – Dust in LLAGN: A V (r)

60. 60 Young-TOs have much more dust than Old-TOs or Old-LINERs Central 10 8-9 yr population is compact R ≤ 100 pc M * ~ 10 7 M o Young-TOs Old-TOs & LINERs 3 – Dust & St pops in LLAGN

61. 61 2 + 3 = Evolution of (stars in) AGN? = To be continued....

62. 62 Flori-where? You are here...

63. 63 The Star Formation History of Galaxies (as revealed by their spectra) Roberto Cid Fernandes UFSC – Florianópolis -Brasil