Properties of Lyman-α and Gamma-Ray Burst selected Starbursts at high Redshifts

Properties of Lyman-α and Gamma-Ray Burst selected Starbursts at high Redshifts Johan P. U. Fynbo (University of Copenhagen, jfynbo@astro.ku.dk) Motivation Lyman-α selected startbursts at z=3 Gamma-Ray Burst selected starbursts Summary and Open Questions COLLABORATORS: Brian Krog (Århus/DK) Cedric Ledoux (ESO/Chile) Palle Møller (ESO/Garching) Kim Nilsson (Copenhagen/DK) Bjarne Thomsen (Århus/DK) Andy Fruchter (STScI/US) Javier Gorosabel (IAA/Spain) Jens Hjorth (Copenhagen/DK) Páll Jakobsson (Copenhagen/DK) Jesper Sollerman (Stockholm/S) Nial Tanvir (Hertfordshire/UK) Paul Vreeswijk (ESO/Chile) + the GRACE collaboration

Damped Ly-α Absorbers Lyman-break galaxies Lyman-α galaxies High-z Galaxies Relations? Which class is dominating? Dusty starbursts

Luminosity function for Lyman-break galaxies Faint-end slope: Steidel et al.: α -1.6 70% of light emitted below R=25.5 (25%) Gabasch et al. (2004): α -1.0 factor of 5 less light below R=25.5! Important to probe the faint end of the Luminosity Function! Adelberger & Steidel (2000)

Lyman-α selection of starbursts (LEGOs) Selection Bias: Hydrogen recombination line selection, hence we select actively starforming galaxies. Biased against: dusty galaxies (high metallicity) red/dead galaxies galaxies with cold kinematics (Mas-Hesse et al. 2003) Fynbo et al. (A&A and Ph.D. thesis, 2000) See also work by Møller & Warren, Hu & Cowie, Steidel et al., Rhoads et al., Ouchi et al., Bunker et al. and many others

The Building the Bridge Survey for z=3 Lyman-α Emitters 3 VLT/FORS fields (7 x7 ) - Last field completed in March this year 60Å wide filters corresponding to Lyman-α at z=2.85, z=3.15, and z=3.20 20-35 candidates per field 60 confirmed z=3 galaxies 10 arcmin -2 per unit z to 7 10-18 erg s -1 cm -2 (5σ). About 5 times more numerous than R<25.5 Lyman-break galaxies. Fynbo et al. (A&A, 2003)

Multi-object spectroscopy FORS1/2 @ VLT (8-12 LEGOs per mask) N B R

Examples of images and 1-d spectra

Luminosities: Comparison with Lyman-break Galaxies Upper limits

How to get more information on the properties of LEGOs: LEGOs in the GOODS South Field (z=3.20) 5

Morphology fwhm 0.1 arcsec R d 600 pc AGNs? No detectable X-ray emission in 2Ms Chandra Image (also not in stacked image, Nilsson et al., in prep.) 3 arcsec = 24 kpc

Photometry: the majority is bluer and fainter than typical Lyman-Break Galaxies LBGs Papovich et al. (2001) Krog et al. in prep. (preliminary)

Gamma-Ray Bursts: Examples from CGRO/BATSE Duration: msec minutes Relativistic Blastwave (ARA&A reviews: van Paradijs, Kouveliotou & Wijers 2000, Mészáros 2002)

GRB 030329: final proof of connection between Long-duration (>2 sec) GRBs and Supernovae Redshift: z=0.1685 Nearest GRB with measured redshift Afterglow: Very bright X-ray/optical/radio Host: starbursting dwarf galaxy (M -16) Hjorth et al. (Nature, 2003)

GRB selection of starbursts It is now established that GRBs are caused by the deaths of massive stars (a subset of SNIb/c, i.e. Wolf-Rayet star progenitors) GRB host galaxies star formation selected galaxies (see poster 16 by Lise Christensen!) GRBs are generally believed to be unbiased tracers of star formation (as gamma-rays penetrate dust), but this has not been established (see also poster 89 by Nial Tanvir, and recall Neil Trentham s talk) 3 Examples: GRB000926, GRB021004, GRB030323

Damped Ly-α Absorbers Lyman-break galaxies Lyman-α galaxies Which class is dominating? GRBs may give the answer! Dusty starbursts

GRB 000926 Lyα Emission Fynbo et al. (A&A, 2001) Fynbo et al. (A&A, 2002) NOT+HST/WFPC2 z=2.038 Host: R(AB)=24.0 A V 0.2 mag

GRB021004 Fynbo et al. (submitted to ApJ) HST/ACS z = 2.330 Host: V(AB)=24.5 A V 0.3 mag Very blue!

GRB030323 Vreeswijk et al. (A&A, 2003) VLT + HST/ACS z = 3.371 Host: V(AB)=28.0 A V <0.5 mag [Fe/H]=-1.4 [S/H]=-1.2 Highest column density measured from Lyα

Ly-α emission from GRB hosts 971214, R=26.0, z=3.42 Yes 000131, R=27.0, z=4.50? 000301, R=28.5, z=2.04? 000926, R=24.0, z=2.04 Yes 011211, R=25.0, z=2.14 Yes 020124, R>29.5, z=3.20? 021004, R=24.4, z=2.33 Yes 030226, R>26.2, z=1.99? 030323, R=28.0, z=3.37 Yes 030429, R>26.3, z=2.66? Fynbo et al. 2003, A&A, 406, L63 All hosts are sub L* relative to the LBG luminosity function and none are detected with SCUBA (Tanvir et al. 2004). For 5 hosts Ly-α emission is detected. In no case we have positive information that a GRB host has no Ly-α emission. GRB hosts are excluded to be drawn randomly from the Lyman-break galaxy population (25% of LBGs have Ly-α in emission). - Is the majority of the star formation at z>2 located in faint, low Z starbursts? - Preference for low metallicities? See also work by Le Floc h et al. (2003), Berger et al. (2003), Tanvir et al. (2004)

Dark GRBs (no detected afterglow) Around 70% of searches for optical afterglows up to 2001 only produced upper limits - sometimes despite deep searches: Dark bursts Why? 1. Dust obscuration 2. Very high redshift (z>7) 3. Afterglow Luminosity function ( intrinsic causes ) But: In the sample of Hete-2/SXC bursts from 2001-2002, >90% have optical afterglows. We need a well-understood sample! Fynbo et al. (A&A, 2001) See also Lazzati et al. (2002), Berger et al. (2002), De Pasquale et al. (2003), Klose et al. (2003)

The Future: The Swift satellite (Oct 7 2004) (survived Frances) Swift will detect 100 GRBs per year (hopefully) The precision in the localisations will be: BAT: 1-4' XRT: ~5 UVOT: <1 all after ~100sec BAT is 5 times more sensitive than BATSE.

Summary and Open Questions Lyman-α emission selected starbursts are much fainter and bluer (younger, less dust) than typical Lyman-break galaxies in current samples GRB selected starbursts at high redshifts seem to be drawn form the Lyman-α emitting starburst population (need better statistics!) Is most of the star formation at high z located in faint, low metallicity dwarfs? Are GRBs biased tracers of star formation?