The Galaxy Evolution Science Case for a Large Ground-based Telescope

1. The Galaxy Evolution Science Case for a Large Ground-based Telescope Betsy Gillespie December 4, 2002 Grateful acknowledgements to: Arjun Dey’s “Galaxy Formation and Evolution” The team for the Canadian XLT science case In particular: Bev Oke Bob Abraham Ray Carlberg Jean-Pierre Veran Laurent Jolissaint

2. Galaxy Evolution • What are the assembly histories of galaxies? Where and when did the first stars form? Where and when did the first galaxies form? How did they come to be the way they are today? • We must: • Test hierarchical CDM, origin of the luminosity function, the morphology-density relation, the Hubble sequence, the Milky Way • Relate distant galaxies to local fossil evidence

3. The Requirements for Progress • The star formation and chemical enrichment histories of galaxies as a function of time: • Star formation histories (rest-frame optical at least); want both for a Sloan-sized sample and as a function of position in the galaxy • Chemical enrichment (rest-frame UV and optical); want as a function of position in the galaxy • Identifying the intrinsic properties, ultimately the masses, of galaxies at high redshift • Mass measures from internal dynamics • Mass measures from strong lensing

4. The Requirements for Progress • Measuring morphologies and the merger rate as a function of time (to constrain hierarchical models) to z=6: • Evolution of different morphological types; identification of most strongly evolving populations at different redshifts • Pair counts at higher redshift (coupled with an understanding of pair selection effects and a theoretical understanding of merger timescales) • Detecting the first objects in the universe

5. Wide-field case: large samples for redshift/abundance surveys • Need large samples to break into sub-types (500,000 galaxies MINIMUM, nearly size of SDSS) • Want many categories: • Mass • Luminosity • Redshift • Environment • Metallicity • Morphological type (Lilly et al. 1995)

6. Sensitivity is vital for a survey down the luminosity function • Arrows show S/N=3 limits for 10,000 seconds in 0.3-arcsec seeing • 500,000 galaxies in <100 clear nights requires at least 1700 galaxies per exposure • >=16-meter requires deeper imaging than HDF to feed a large survey (JWST!) RAB PHOTOMETRIC REDSHIFT

7. Major advantages with some correction over the wide field (in the optical) • Arrows show S/N=3 limits for 10,000 seconds in 0.5-arcsec seeing If seeing degrades to 0.5 arcsec, sensitivity worse by 0.5 magnitudes for unresolved sources Requires 2.4 x longer exposures RAB PHOTOMETRIC REDSHIFT

8. Near-IR case: for chemical abundances, star formation histories weak absorption Lines in the optical and near-infrared [OII] to z = 6 Ha to z = 3 L/M optical • Few strong lines in optical • between redshifts of • about 1 to 3 • NEED near-IR K H J Plot from Oke & Barton (2000)

9. The importance of the NIR: sensitivity as a function z in Ha and [OII] • At z < 1.5, [OII] in optical and Ha in NIR are comparable even with no dust; no metallicity effects in using Ha star formation rate • Beyond z=1.5, both lines perform well in NIR; for z=2-3, Ha is best 30m [OII] 30m Ha Globular cluster forming in 1 dynamical timescale NGST Ha NGST [OII] Sensitivity to unresolved emission lines, R=3000, T=10,000 sec

10. Internal kinematics: absorption lines Most distant galaxies fall below the contiuum surface brightness limit (from NOAO GSMT Book)

11. Emission lines: “Typical galaxies” at z=1.5 z=0 z=1.5 8m z=1.5 20m z=1.5 30m 8-meter telescopes only detect the center! 20-hour exposure

12. Assumed psf for these simulations Image-quality studies at HIA: • Chris Morbey -- telescope • Laurent Jolissaint and Jean-Pierre Veran --- added atmosphere and AO Bulk property like Strehl ratio likely important, but detailed features of psf not important… K-band PSF

13. Major issue for image quality: how lumpy is star formation? Typical ground-based Resolution for local galaxies 30-meter telescope diffraction limits Diameter (pc) Star clusters in Antennae have <Re> = 4 pc (Whitmore et al. 1999) redshift

14. Higher redshifts: z > 3 • Galaxy morphologies “more challenging” to recognize • Large disks may not be in place, but relative velocities of lumps will provide information about dynamical state and/or total mass (z=3 galaxy from Hubble Deep Field; HST psf ~ 0.1” ~ 770 pc)

15. Distant Galaxy Morphology: details help enable merger rate measurement Simulation by Bob Abraham CFHT 2-hour exp. z=0.8 z=0.6 20-meter 2-hour exp. (“seeing” 2x Diffraction limit) z=0.6 z=0.8

16. Detecting the first objects in the universe • At z=6-10, Lya is at 0.85 < l < 1.4 mm: regime where a 30-meter is much more sensitive than JWST • JWST NIRcam proposal science case: parameters of first objects, with Charlot & Fall (1993) Lya analysis, gives Lya fluxes up to ~10-18.4 erg s-1 cm2 at z=10 Sensitivity to unresolved emission lines, R=3000, t=10,000 seconds

17. Progress with 8-10 meter telescopes • Large redshift surveys to z=3-4 (not too far down LF and most not in NIR) • Most diagnostics will be rest-frame UV (exception is VIRMOS) • Will measure unobscured SFR as a function of redshift • Kinematics of bright or strongly star-forming galaxies to perhaps z=1.5 (plus occasional shear of Lyman break galaxy)

18. Complementarity to JWST, ALMA • Spectroscopic follow-up to JWST imaging surveys • Locally, the huge bursts of star formation are dust-enshrounded (e.g., ultraluminous infrared galaxies) • argues for complementary imaging and spectroscopy at longer wavelengths

19. Summary of Requirements • Sensitivity. • For spectroscopic surveys of huge samples • For internal dynamics with no systematic problems • To detect the “first” objects in the universe • Near-IR capabilities. • For accurate chemical abundances directly comparable to what we know at low redshift • For Ha star formation rates to z=3 • For [OII](3727) star formation rates beyond z=5. • Wide field. • Large survey (~106 galaxies) of luminosity function as a function of galaxy type • Good image quality. • For better sensitivity • If star formation in the universe is lumpy on small scales • For high-redshift morphologies: the diffraction limit of a 30-meter telescope is nearly 5 times better than a 6.5-meter JWST

20. Major unresolved issues and required work • What will the image quality be? Over what field and fraction of the sky? Is some wide-field correction feasible? • What is the expected distribution of emission-line (and UV) flux from high-redshift galaxies? Is it lumpy? Is it likely to remain resolved or unresolved? (Both spatially and in velocity width?) • If this question cannot be answered, what is the best strategy to adopt? • Kinematic simulations “from scratch” for adjustable parameters. • Worth pursing issue of comparison with NGST at 2.5 to 4 microns.

21. REFERENCE SLIDES

22. Scalings: Magnitude Limits

23. Scalings: Exposure Times Not just a matter of patience! Many studies require large samples of objects.

24. Lookback Time Lambda cosmology: ~9 Gyr to z=1.5 (only 4 Gyr from z=1.5 to z=6) Time H0=70 km/s/Mpc W0=0.3 WL=0.7 z

25. “Average” Galaxies at Intermediate Redshift: z=1 z=0 z=1 8m z=1 20m z=1 30m 10-hour exposure

26. “Average” Galaxies at Intermediate Redshift: z=1.5 z=0 z=1.5 8m z=1.5 20m z=1.5 30m 10-hour exposure

27. Longer Exposures still only see the center! z=0 z=1.5 8m z=1.5 20m z=1.5 30m 20-hour exposure