Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stecf.org/goods/spectroscopy/CDFS_Mastercat/newstanway04.pdf Äàòà èçìåíåíèÿ: Thu Sep 16 11:25:44 2004 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 01:21:21 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http www.stsci.edu hst udf

The Astrophysical Journal, 604:L13­L16, 2004 March 20
2004. The American Astronomical Society. All rights reser ved. Printed in U.S.A.

THREE Lya EMITTERS AT z 6: EARLY GMOS/GEMINI DATA FROM THE GLARE PROJECT Elizabeth R. Stanway,1 Karl Glazebrook,2 Andrew J. Bunker,1 Roberto G. Abraham,3 Isobel Hook,4 James Rhoads,5 Patrick J. McCarthy,6 Brian Boyle,7 Matthew Colless,8 David Crampton,9 Warrick Couch,10 Inger JÜrgensen,11 Sangeeta Malhotra,5 Rick Murowinski,9 Kathy Roth,11 Sandra Savaglio,2 and Zlatan Tsvetanov2
Received 2003 December 17; accepted 2004 February 10; published 2004 February 23

ABSTRACT We report spectroscopic detection of three z 6 Lya-emitting galaxies, in the vicinity of the Hubble Ultra Deep Field, from the early data of the Gemini Lyman Alpha at Reionisation Era (GLARE) project. Two objects, GLARE 3001 (z p 5.79) and GLARE 3011 (z p 5.94), are new detections and are fainter in z (zAB p 26.37 and 27.15) than any Lyman break galaxy previously detected in Lya. A third object, GLARE 1042 (z p 5.83), has previously been detected in line emission from the ground; we report here a new spectroscopic continuum detection. Gemini/GMOS-South spectra of these objects, obtained using nod and shuffle, are presented together with a discussion of their photometric properties. All three objects were selected for spectroscopy via the i-drop Lyman break technique, the two new detections from the GOODS version 1.0 imaging data. The red i z colors and high equivalent widths of these objects suggest a high-confidence z 1 5 Lya identification of the emission lines. This brings the total number of known z 1 5 galaxies within 9 of the Hubble Ultra Deep Field to four, of which three are at the same redshift (z p 5.8 within 2000 km s 1), suggesting the existence of a large-scale structure at this redshift. Subject headings: galaxies: evolution -- galaxies: formation -- galaxies: high-redshift -- galaxies: starburst
1. INTRODUCTION

The work of detecting and analyzing objects at ver y high redshift is a challenging but rapidly advancing field. In recent months, increasing numbers of galaxies have been found at redshifts of z 6, as a result of both narrowband selection (e.g., Rhoads et al. 2003; Taniguchi et al. 2003; Hu et al. 2003) and the use of Lyman break methods (e.g., Dickinson et al. 2004; Stanway, Bunker, & McMahon 2003a, hereafter SBM03). However, until now, spectroscopic confirmation of ver y high redshift candidates has only been possible for objects at the bright end of the galaxy luminosity function. Spectroscopic sur veys for fainter objects are essential for shaping our understanding of the universe at these redshifts, and, with modern instrumentation and new techniques for faintobject spectroscopy, such sur veys are now possible. The Gemini Lyman Alpha at Reionisation Era (GLARE) sur vey,
1 Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge, CB3 0HA, UK; ers@ast.cam.ac.uk, bunker@ast.cam.ac.uk. 2 Department of Physics and Astronomy, Johns Hopkins University, 3400 North Charles Street, Baltimore, MD 21218-2686; kgb@pha.jhu.edu, savaglio@tarkus.pha.jhu.edu, zlatan@pha.jhu.edu. 3 Department of Astronomy and Astrophysics, University of Toronto, 60 St. George Street, Toronto, ON M5S 3H8, Canada; abraham@astro.utoronto.ca. 4 Department of Astrophysics, Nuclear and Astrophysics Laborator y, Oxford University, Keble Road, Oxford OX1 3RH, UK; imh@astro.ox.ac.uk. 5 Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218; rhoads@stsci.edu, san@stsci.edu. 6 Obser vatories of the Carnegie Institute of Washington, 813 Santa Barbara Street, Pasadena, CA 91101; pmc2@ociw.edu. 7 Australia Telescope National Facility, P.O. Box 76, Epping, NSW 1710, Australia; brian.boyle@csiro.au. 8 Anglo-Australian Obser vator y, P.O. Box 296, Epping, NSW 1710, Australia; colless@aao.gov.au. 9 Herzberg Institute of Astrophysics, National Research Council, 5071 West Saanich Road, Victoria, BC V9E 2E7; david.crampton@nrc.ca, richard.murowinski@hia.nrc.ca. 10 School of Physics, University of New South Wales, Kensington, Sydney, NSW 2052, Australia; wjc@edwin.phys.unsw.edu.au. 11 Gemini Obser vator y, 670 North A'ohoku Place, Hilo, HI 96720-2700; ijorgensen@gemini.edu, kroth@gemini.edu.

described in this Letter, aims to obtain extremely deep spectra of ver y faint high-redshift star-forming galaxies, selected via i z color from the Hubble Space Telescope (HST ) Advanced Camera for Sur veys (ACS) Ultra Deep Field (UDF; Beckwith, Somer ville, & Stiavelli 2003) and Great Obser vatories Origins Deep Sur vey (GOODS; Giavalisco et al. 2004) data, using the Gemini Multi-Object Spectrograph (GMOS) on Gemini South. In this Letter, we present spectra and fluxes12 of three Lya emitters, two of which are new, identified in the early observation data of the GLARE project.
2. GMOS OBSERVATIONS

The spectra described in this Letter were obtained as part of the Gemini South program GS-2003B-Q-7 using the GMOS (Hook et al. 2003) and targeting sources in the region of the Hubble UDF (Beckwith et al. 2003). The goal of program GS2003B-Q-7 was to obtain a total of 100 hr of exposure on this UDF mask. This Letter concerns the first 7.5 hr of data. Ob° ser vations were made with the R150 grating at a 9000 A central ° wavelength and with a custom-made 7800 A long-pass filter ° ("RG780") giving a spectral range from 7800 A up to the CCD ° cutoff (about 10,000 A) and a spectroscopic resolution of ° ° 15 A (4 pixels FWHM at 3.5 A pixel 1) with 0 . 7 wide slits and a resolving power of l /Dl FWHM 550. Since the seeing disk (!0 . 5 FWHM) was smaller than the slit width (0. 7), the true resolution is somewhat better for a source that does not fill the slit. The data were taken using the "nod and shuffle" (N&S) obser ving mode (Cuillandre et al. 1994; Glazebrook & BlandHawthorn 2001; Abraham et al. 2004, hereafter GDDS1). Our N&S setup and obser vational scheme follow that in GDDS1, except that the slits were 2 . 5 long with a 1 . 5 nod. Data reduction also follows GDDS1 (Appendices B and C). Fifteen
12 We adopt the following cosmology: a flat universe with QL p 0.7, QM p 0.3, and H0 p 70 h70 km s 1 Mpc 1. All magnitudes are quoted in the AB system (Oke & Gunn 1983).

L13


L14

STANWAY ET AL.
TABLE 1 Summary of Spectroscopic and Photometric Properties Object 1042 ...... 3001 ...... 3011 ...... z 5.83 5.79 5.94 Peak ° (A) 8309.1 8252.7 8434.0 Flux 1 (ergs cm 2 s 1) 0.97 # 10 0.79 # 10 1.0 # 10
17 17 17

Vol. 604

Flux 2 (ergs cm 2 s 1) 0.91 # 10 0.76 # 10 1.0 # 10
17 17 17

FWHM ° (A) 16.3 20.5 24.5 25.48 26.37 27.15

z 0.03 0.06 0.12

iz 1.48 0.09 1.66 0.20 11.68 (3 j)

Rh (z ) (arcsec) 0.09 0.14 0.13

(h

2 70

SFRUV M, yr 1) 15 6.0 1.7

(12 j) (12 j) (25 j)

Notes.--The wavelengths are measured in power points, and flux 2 by fitting a Gaussian ° of 16.5 A). All objects are undetected at 3 j calculated using the UV flux­SFR relation of

air. The error on the fluxes due to slit losses and photon noise is 20%. Flux 1 is taken between zero line profile to the data. The FWHM of this fit is shown in the sixth column (with a spectral resolution in v (v 1 29.4 ) and also resolved in the z band (with a stellar half-light radius Rh p 0 .05 ). SFRs are Madau et al. (1998), having removed line contamination (and assuming b p 1.1).

1800 s frames were obtained in queue obser ving during 2003 November (19th, 24th, 28th, and 30th) during conditions of 0 . 4­0 . 5 seeing and high transparency. Relative flux calibration was performed using obser vations of a standard star, and the absolute flux calibration was done by normalizing the spectra of the mask alignment stars to their z -band photometric fluxes. Sky lines were used for wavelength calibration. One- and two-dimensional sky-subtracted spectra were inspected for the presence of emission lines. An advantage of the N&S data reduction technique is that lines appear in a positive-negative dipole pattern in the two-dimensional images, making them easy to distinguish from residual CCD defects and other nonastrophysical effects.
3. HST/ACS PHOTOMETRY AND CANDIDATE SELECTION

Spectroscopic candidates for this program were selected using the i-drop Lyman break technique (see SBM03 or Dickinson et al. 2004). This photometric selection method has been used by a number of authors to identify z 1 5 galaxies (e.g., Bremer et al. 2004 and Lehnert & Bremer 2003 at z 5, Dickinson et al. 2004 and SBM03 at z 6). In each of these sur veys, the Lyman break spectral feature is detected by means of a color-cut criterion. At z 6, the Lyman break passes through the z filter, and hence a cut is usually placed in the range

1.3 ! i z ! 1.5, with the exact value affecting the sur vey redshift range and low-redshift contamination. Within the UDF itself, 19 candidates were drawn from the catalog of the reddest (i z ) objects, prereleased by the UDF team13 in order to facilitate ground-based spectroscopic followup and satisfy (i z )AB 1 1.5. The 5 . 5 GMOS field is bigger than the UDF, so in the outlying region we selected objects from the GOODS-South field. The GOODS version 1.0 data release14 comprises co-added imaging from 5 "epochs" of obser vations, reaching 3 j magnitude limits v lim p 29.44, i lim p 28.83, and z lim p 28.52. We selected candidates using (i z )AB 1 1.3 mag and zAB ! 27.5 mag and identified 18 objects in the slit-mask area of which 10 were placed on the mask. The faint end of this selection reaches the magnitude limit in the i band. The rest of the mask area was used for a blank sky sur vey. Of the objects presented in this Letter, object 1042 was identified as a candidate in both catalogs, while objects 3001 and 3011 lie outside the UDF and were selected from the GOODS version 1.0 data. Table 1 presents the broadband photometric properties of these objects. Magnitudes are measured in a 0 . 3 diameter aperture, and an aperture correction of 0.32 mag (measured from point sources on the images) is applied to obtain a total magnitude for these compact sources. Postage stamp images in the v, i , and z bands are shown in Figure 1. Since object 1042 lies within the UDF, more accurate photometr y and morphological information should soon be available for this object. All three objects are spatially resolved with half-light radii in the range 0 . 09­0 . 14 (cf. stellar value of 0 . 05), well detected in the z band, faintly detected (or in the case of object 3011 undetected at 3 j) in i , and undetected at 3 j in the v band.
4. RESULTS

Single isolated emission lines have been identified for each s of the three objects. GLARE 1042 (R.A.p 3h32m40.0, decl. p 27 48 15 . 0 [J2000]) is an i-drop initially selected by SBM03 and already confirmed as a z p 5.83 Lya emitter (Stanway et al. 2003b; Dickinson et al. 2004). GLARE 3001 (R.A. p s s 3h32m46.0, decl. p 27 49 29 . 7) and 3011 (R.A. p 3h32m43.2, decl. p 27 45 17 . 6) are new. Their properties are given in Table 1. All three lines are almost unresolved given our dispersion, although 1042 is known to be highly asymmetric from published spectra (Stanway et al. 2003b; Dickinson et al. 2004). We also detect the continuum break in the spectrum of GLARE 1042, although at a low signal-to-noise ratio (flux ° density 0.6 # 10 19 ergs cm 2 s 1 A 1). This is consistent with the level of continuum reported in the low-resolution ACS
Fig. 1.--GOODS-South imaging of GLARE 1042, GLARE 3001, and GLARE 3011 in v (F606W), i (F775W), and z (F850LP). The panels are 3 to a side (17 h701 kpc at z p 5.8). North is up.
13 14

See http://www.stsci.edu/hst/udf. Available at ftp://archive.stsci.edu/pub/hlsp/goods/v1.


No. 1, 2004

GLARE LINE EMITTERS AT z 6

L15

slitless spectrum of Dickinson et al. (2004) and also with the broadband colors of this object. A good continuum spectrum of this object should be obtained if GLARE does achieve a 100 hr total exposure. The spectra of the candidates are shown in Figure 2.
5. INTERPRETATION

5.1. The Redshifts of GLARE 1042, 3001, and 3011 The most plausible line identifications given the red i z selection are Lya at z p 5.83, z p 5.79, and z p 5.94 for 1042, 3001, and 3011, respectively. Hb l4861.3 or [O iii] ll5006.8, 4958.9 at z 0.7 or Ha l6562.8 at z 0.25 are ruled out be the absence of nearby lines and because galaxies at these redshifts do not have strong i z continuum breaks. An unresolved [O ii] ll3726.1, 3728.9 doublet at redshifts z 1.2 is a possibility for 3001 and 3011 (1042 is already known to be strongly asymmetric, thus ruling it out; Stanway et al. 2003b). The best evidence against this is the red i z color. The reddest possible color from an early-type spectral energy distribution at z p 1.2 is i z p 1.2 (SBM03), and our two new objects are redder than this at 95% confidence. An [O ii] identification would also imply rest-frame equivalent ° widths of 50­250 A (see § 5.2), which is higher than is normally seen in z 1 [O ii] emitters (Hammer et al. 1997). 5.2. Equivalent Widths The emission-line equivalent widths can be calculated from their z -band photometr y and line fluxes. However, a subtlety arises as the emission lines for these objects contribute significantly to their continuum flux in both the z and i filters; the calculated equivalent widths are sensitive to the detailed shape of the filter transmission cur ve. Accounting for these effects, Ly ° we calculate equivalent widths Wresta p [20, 30, 100] 10 A for [1042, 3001, 3011] respectively, assuming that there is negligible flux shortward of the Lya line due to absorption by the Lya forest. From stellar synthesis models of star-forming regions (e.g., Charlot & Fall 1993), the theoretical Lya equivalent width for a young region of active star formation is Ly ° Wresta 100 ­ 200 A. However, obser ved Lya emission from star-forming galaxies is generally weaker in continuum-selected samples at lower redshifts, which typically have Wrest p ° 5­ 30 A (e.g., Steidel et al. 1996), or even in absorption. Our z 6 sample has a comparable selection, but the equivalent width of GLARE 3011 is notably larger. We note that z 6 narrowband-selected samples also produce higher equivalent ° widths with a median of 200 A (Malhotra & Rhoads 2002; Hu et al. 2003). 5.3. Active Galactic Nucleus or Starburst? Careful inspection of the spectra reveals no other lines. The only other significant line in our range is the high-ionization rest-UV line N v l1240, which is usually prominent in active ° galactic nuclei (AGNs). Our flux limits at l rest p 12 40 A are ° f ! [2.3, 2.3, 1.8] # 10 18 ergs cm 2 s 1 (3 j), for a 15 A FWHM (unresolved) line. This gives lower limits of f (Lya)/f (N v) 1 [4.3, 3.5, 5.6] (3 j). Typical line ratios from composite QSO spectra are f (Lya)/f (N v) p 4.0 (Osterbrock 1989). Although our constraints are admittedly weak, the nondetection of N v l1240 favors the interpretation that the Lya arises from the Lyman continuum flux produced by OB stars rather than the harder UV spectrum of a QSO.

Fig. 2.--Unbinned GMOS-South spectra of GLARE 1042, GLARE 3001, and GLARE 3011 around the Lya candidate line emission. The continuum level is indicated on 1042

None of the objects in this Letter correspond to X-ray sources in the Chandra Deep Field­South (CDF-S) 1 Ms catalog that covers this region (Giacconi et al. 2002). This allows us to place a limit on their X-ray fluxes in the 0.5­ 2 keV (soft) and 2 2­8 keV (hard) bands at 5.5 # 10 17 and 4.5 # 10 16 h 70 ergs cm 2 s 1, respectively. The hard X-ray flux limits rule out a type II QSO like CDF-S 202 (Norman et al. 2002) but not CXO 52 (Stern et al. 2002). The soft X-ray limits would permit low-luminosity Seyfert galaxies and QSOs (Malhotra et al. 2003), but the line emission we see is not broad. Given that an AGN interpretation for the obser ved emission lines is unlikely, the rest-frame UV spectra of these objects is likely to be dominated by star formation. The relation between ° the flux density in the rest-frame UV around 1500 A and the star formation rate (SFR, in units of solar mass per year) is given by L UV p 8 # 10 27 SF R ergs s 1 Hz 1 (Madau, Pozzetti, & Dickinson 1998) for a Salpeter (1955) stellar initial mass function (IMF) with 0.1 M, ! M ! 125 M,. The rest-frame UV at z 6 is redshifted into the z photometric band. The inferred SFR for the objects in this Letter, given their z magnitudes, line contamination, and identified redshifts, are shown in Table 1.


L16

STANWAY ET AL.

Vol. 604

5.4. Overdensity at z p 5.8? The redshift of the new Lya emitter GLARE 3001 is ver y similar to that of Lya emitter SBM03 No. 3 reported by Bunker et al. (2003) and is also close to that of the reconfirmed z p 5.83 object (variously called GLARE 1042, SBM03 No. 1, and CDFS J033240.0 274815 in this Letter, Stanway et al. 2003b, and Dickinson et al. 2004, respectively). The i-drop selection procedure used to identify candidate objects for spectroscopy is in principle sensitive to objects in the redshift range 5.6 ! z ! 7.0, although the sharp falloff in the transmission of the z filter effectively restricts detection of objects to a smaller redshift range that varies with absolute magnitude (5.8 ! z ! 6.5 for an Lzp3 Lyman break galaxy; see SBM03). The line centers of three of the four Lya emitters detected in the GOODS-South field are separated by 2000 km s 1 (only 5% of the sur vey redshift range). The maximum angular separation of the three objects (GLARE 1042, 3001, SBM03 No. 3) is less than 11 . 5, corresponding to a projected 1 separation of only 4.0 h 70 Mpc. This suggests the interesting possibility of an overdensity of Lya emitters at z p 5.8 in this field, perhaps similar to that obser ved at z p 3.1 in SSA 22 by Steidel et al. (2000). To quantify the statistical likelihood of this happening by chance, we undertook a Monte Carlo simulation. A population uniformly distributed in the redshift range 5.6 ! z ! 7.0 was considered, and the effects of the GMOS throughput (including the RG780 filter) and increased noise due to sky lines on spectroscopic detection, and intergalactic medium absorption leading to incomplete z filter coverage on Lyman break photometric selection, were modeled. The effects of an apparent magnitude limit on the photometric selection were also calculated, using the measured luminosity function for Lyman break galaxies at z p 3 (Steidel et al. 1999) as an approximation for the currently unknown z p 6 luminosity function. Although each of these effects bias detection probability toward the lower end of the accessible redshift range, the probability of selecting three or more out of four galaxies within 2000 km s 1 of z p 5.80 was still small (2.1%). This suggests that we have statistical evidence of a redshift spike at z p 5.8 in this field, which augurs well for the future UDF obser vations. 5.5. Comparison with Previous Work The spectroscopic properties of the three galaxies presented in this work are similar to those of other confirmed high-redshift

Lyman break galaxies. Lehnert & Bremer (2003) report Lya luminosities of 10 42­10 43 ergs s 1 for six z p 5 line emitters in a sample of 12 Lyman break candidates. The fluxes for our z p 6 Lya emitters fall comfortably into this range of luminosities and probe to a comparable depth. The emission-line equivalent widths of the GLARE candidates are also comparable to those of other spectroscopically confirmed Lyman Ly ° break galaxies; e.g., at z p 5.8, Wresta p 20 A (Bunker et al. Ly ° 2003), and at z p 6.17, Wresta p 50 A (Cuby et al. 2003).
6. CONCLUSIONS

In this Letter, we have presented photometr y and spectra for three objects with extreme i z colors and line emission that may be Lya at z 5.8, obser ved as part of the GLARE project. Our main conclusions can be summarized as follows: 1. The GLARE project has detected three ver y high redshift objects (z p 5.83, 5.79, and 5.94) in the first 7.5 hr of integration time on Gemini/GMOS-South. To the best of our knowledge, the two new Lya emitters are fainter in z than any previous Lyman break­selected objects with a spectroscopic redshift. 2. The i z color selection can successfully identify objects lying at these redshifts and magnitudes, and the faint end of the galaxy luminosity function at z p 6 is now within the reach of 8 m telescopes. 3. We have evidence of an overdensity of z p 5.8 objects in a narrow redshift spike in the GOODS-South field; further obser vations (e.g., narrowband imaging) would be invaluable in confirming the existence of such a structure.

The work done in this Letter was based on obser vations obtained at the Gemini Obser vator y, which is operated by AURA, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: NSF (US), PPARC (UK), NRC (Canada), CONICYT (Chile), ARC (Australia), CNPq (Brazil), and CONICET (Argentina). K. G. and S. S. acknowledge generous funding from the David and Lucille Packard Foundation. The work done in this Letter was also based on obser vations made with the NASA/ESA Hubble Space Telescope, obtained at the STScI, which is operated by AURA, Inc., under NASA contract NAS5-26555.

REFERENCES Abraham, R. G., et al. 2004, AJ, in press (GDDS1) (astro-ph/0402436) Beckwith, S. V. W., Somerville, R., & Stiavelli, M. 2003, STScI Newsl., 20(4), 1 Bunker, A. J., Stanway, E. R., Ellis, R. S., McMahon, R. G., & McCarthy, P. J. 2003, MNRAS, 342, L47 Bremer, M. N., Lehnert, M. D., Waddington, I., Hardcastle, M. J., Boyce, P. J., & Phillipps, S. 2004, MNRAS, 347, L7 Charlot, S., & Fall, S. M. 1993, ApJ, 415, 580 Cuby, J.-G., Le Fevre, O., McCracken, H., Cuillandre, J.-C., Magnier, E., & ` Meneux, B. 2003, A&A, 405, L19 Cuillandre, J. C., et al. 1994, A&A, 281, 603 Dickinson, M., et al. 2004, ApJ, 600, L99 Giacconi, R., et al. 2002, ApJS, 139, 369 Giavalisco, M., et al. 2004, ApJ, 600, L93 Glazebrook, K., & Bland-Hawthorn, J. 2001, PASP, 113, 197 Hammer, F., et al. 1997, ApJ, 481, 49 Hook, I., et al. 2003, Proc. SPIE, 4841, 1645 Hu, E. M., Cowie, L. L., Capak, P., McMahon, R. G., Hayashino, T., & Komiyawa, Y. 2003, preprint (astro-ph/0311528) Lehnert, M. D., & Bremer, M. N. 2003, ApJ, 593, 630 Madau, P., Pozzetti, L., & Dickinson, M. 1998, ApJ, 498, 106 Malhotra, S., & Rhoads, J. E. 2002, ApJ, 565, L71 Malhotra, S., Wang, J. X., Rhoads, J. E., Heckman, T. M., & Norman, C. A. 2003, ApJ, 585, L25 Norman, C., et al. 2002, ApJ, 571, 218 Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713 Osterbrock, D. E. 1989, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei (Mill Valley: University Science Books) Rhoads, J. E., et al. 2003, AJ, 125, 1006 Salpeter, E. E. 1955, ApJ, 121, 161 Stanway, E. R., Bunker, A. J., & McMahon, R. G. 2003a, MNRAS, 342, 439 (SBM03) Stanway, E. R., Bunker, A. J., McMahon, R. G., Ellis, R. S., Treu, T., & McCarthy, P. J. 2003b, preprint (astro-ph/0308124) Steidel, C. C., Adelberger, K. L., Giavalisco, M., Dickinson, M. E., & Pettini, M. 1999, ApJ, 519, 1 Steidel, C. C., Adelberger, K. L., Shapley, A. E., Pettini, M., Dickinson, M., & Giavalisco, M. 2000, ApJ, 532, 170 Steidel, C. C., Giavalisco, M., Pettini, M., Dickinson, M. E., & Adelberger, K. L. 1996, ApJ, 462, L17 Stern, D., et al. 2002, ApJ, 568, 71 Taniguchi ,Y., Shioya, Y., Ajiki, M., Fujita, S. S., Nagao, T., & Murayama, T. 2003, J. Korean Astron. Soc., 36, 123