Astronomy Letters, Vol. 28, No. 1, 2002, pp. 1–11. Translated from Pis’ma v Astronomicheski˘ı Zhurnal, Vol. 28, No. 1, 2002, pp. 3–13. Original Russian Text Copyright c 2002 by Reshetnikov, Vasil’ev. The Observed Structure of Extremely Distant Galaxies V. P. Reshetnikov * and A. A. Vasil’ev Astronomical Institute, St. Petersburg State University, Bibliotechnaya pl. 2, Petrodvorets, 198904 Russia Received July 18, 2001 Abstract—We have identified 22 galaxies with photometric redshifts zph = 5–7in the northern and southern Hubble Space Telescope deep fields. An analysis of the images of these objects shows that they are asymmetric and very compact (∼1 kpc) structures with high surface brightness and absolute m magnitudes of MB ≈−20 . The average spectral energy distribution for these galaxies agrees with the distributions for galaxies with active star formation. The star formation rate in galaxies with zph =5–7 −1 was estimated from their luminosity at λ = 1500 Atobe˚ ∼30 M yr . The spatial density of these objects is close to the current spatial density of bright galaxies. All the above properties of the distant galaxies considered are very similar to those of the so-called Lyman break galaxies (LBGs) with z ∼ 3–4. The similarity between the objects considered and LBGs suggests that at z ∼ 6, we observe the progenitors of present-day galaxies that form during mergers of protogalactic objects and that undergo intense starbursts. c 2002 MAIK “Nauka/Interperiodica”. Key words: theoretical and observational cosmology; galaxies, groups and clusters of galaxies, intergalactic gas INTRODUCTION Fields) and the technique capable of identifying them (photometric redshifts). The Hubble Space Telescope (HST) and major ground-based telescopes allowed several galaxies Hubble Deep Fields (northern and southern— (not quasars or radio galaxies) at large “cosmolog- HDF-N and HDF-S) are currently the deepest ical” redshifts to be identified and studied in recent “punctures” made into the Universe (Ferguson et al. years (Stern and Spinrad 1999). For instance, objects 2000). In these two small areas (each with a size with redshifts z =5.60 (Weymann et al. 1998) and of ∼5.3 arcmin2) several thousand objects have been m z =5.34 (Spinrad et al. 1998) have been discovered identified down to a limiting magnitude of V ∼ 30 . in the northern Hubble Deep Field. Other galaxies Each Deep Field pixel has an angular size of only with z =5.34 (Dey et al. 1998), z =5.74 (Hu et al. 0.04, which allows objects with sizes ∼1 kpc to be 1999), and z =5.63 (Dawson et al. 2001) are also resolved at almost all z.Originalfield images were known. The importance given to searches for such obtained in F300W, F450W, F606W, and F814W extremely distant and, therefore, very young objects broadband filters (numbers in filter names indicate the is quite understandable, because they serve as crucial central wavelength in nanometers). These filters are tests for modern models of galaxy formation and also often referred to as U300, B450, V606,andI814 because allow the processes in the relatively early Universe to of their relative proximity to the corresponding stan- be studied. dard filters. Later, HST and ground-based infrared The principal aim of our work is to attempt to (JHK) images of the deep fields were taken. These analyze the observed structure of the most distant additional data were obtained mostly with lower objects identified in deep fields in the spectral region angular resolution and lower limiting magnitude. at the boundary between the optical and the infrared. We decided to analyze objects located in two fields and having redshifts ranging from 5 to 7. The up- per limit is due to the fact that at z>7 the part of GALAXY SAMPLE the galactic spectrum with λ0 > 1216 A(˚ λ0 is the The goal of this work unambiguously defines both wavelength in the restframe of the galaxy) falls beyond the region to be searched for the objects we are in- the passband of the I814 filter (the longest wavelength terested in (the northern and southern Hubble Deep filter in which the original deep-field images are ob- tained). At λ0 < 1216 A,˚ the observed flux from dis- * E-mail: Öe×h@a×ØÖÓº×ÔbÙºÖÙ tant objects decreases abruptly due to absorption by 1063-7737/02/2801-0001$22.00 c 2002 MAIK “Nauka/Interperiodica” 2 RESHETNIKOV, VASIL’EV Table 1. Galaxies in HDF-N Table 2. Galaxies in HDF-S N Name I814(AB) zsp zph Type N Name I814(AB) zph Type 3a 3-951.1 26.20 5.34 5.25 Irr 0036 SB-WF-0393-1295 29.86 6.75 Scd 3b 3-951.2 26.95 5.34 5.25 Irr 0089 SB-WF-0648-0506 28.53 5.49 Scd 213 4-473.0 27.17 5.60 5.95 Irr 0277 SB-WF-1510-1270 29.60 6.63 Sbc 328 4-200.0 26.19 – 5.20 Irr 0400 SB-WF-1947-1654 28.23 6.52 SB1 343 3-543.0 27.78 – 5.95 Irr 0470 SB-WF-2127-1818 29.77 6.85 SB1 0543 SB-WF-2328-3923 27.55 5.12 SB2 367 3-489.0 26.74 – 5.18 Irr 0611 SB-WF-2434-2255 28.28 5.20 Irr 576 3-228.0 26.93 – 5.58 Irr 0696 SB-WF-2611-1523 28.73 5.26 SB1 658 4-799.0 27.30 – 5.33 Irr 0881 SB-WF-3032-2976 29.19 6.01 Irr 714 3-44.0 27.20 – 5.29 Irr 0886 SB-WF-3049-1323 28.07 5.56 SB2 0932 SB-WF-3144-0926 29.10 6.94 SB1 intergalactic HI clouds (Madau 1995). We therefore cannot expect the images of more distant objects to 1079 SB-WF-3473-1780 28.93 5.22 SB2 be seen in the I814 band. The lower limit (z>5) is set 1097 SB-WF-3516-0827 29.25 5.72 Irr arbitrarily. that mutually independent data by Fernandez-Soto Recall that the age-redshift dependence is strongly et al. (1999) and Fontana et al. (2000) yield z that nonlinear and galaxies at z ∼ 5–7di ffer little in ph agree fairly well with each other. Thus, the z values age from those at z ∼ 3–4 (of which ∼103 are ph for HDF-N objects with z =5–7identi fied using the already known). Let us illustrate this point using data of Fontana et al. (2000) differ from the z the Einstein–deSitter model as an example. In this ph model, age t is related to z in the following simple determinations of Fernandez-Soto et al. (1999) by +0.24 ± 0.39 (eight galaxies). 2 − 3 −1 way: t = (1 + z) 2 H ,whereH is the Hubble 3 0 0 The subsequent work was based on the origi- nal I -band frames of the two fields available from constant. It thus follows that for H0 =50km/s/Mpc 814 the present-day (z =0) age of the Universe is equal the site of the Space Telescope Science Institute (STScI). To identify real galaxies, we analyzed sur- to t0 =13.0 Gyr. Galaxies with z =3are observed at t =1.6 Gyr, and those with z =4,att =1.2 Gyr. face brightness distributions for all objects with zph = 5–7. An analysis of the profiles of several dozen stars Redshifts z =5and z =7correspond to the age of the Universe of 0.9 and 0.6 Gyr, respectively. The in these fields showed that they had FWHM =0.14. galaxies at z ∼ 3.5 are therefore “older” than those Note that the width of the stellar profiles depends only observed at z ∼ 6 by no more than 109 yr. Most of the slightly on apparent magnitude: FWHM =0.146 ± time of cosmic evolution falls within z ≤ 1. 0.018 and FWHM =0.139 ± 0.026 for stars with m m To identify objects with z =5–7, we used pub- I814 > 22 and fainter stars with I814 > 24 , respec- lished photometric redshifts zph of objects in HDF-N tively. We left in the final list only the resolvable ob- and HDF-S adopted from the catalogs of Fernandez- jects whose brightness distribution is more extended Soto et al. (1999) and Fontana et al. (2000) (see that those of stars (FWHM > 0.14). the corresponding WWW addresses in the original Tables 1 and 2 list candidate distant galaxies in papers). The zph estimates in the corresponding pa- HDF-N and HDF-S, respectively. The first column pers are based on seven-color (V300 to K) photometry. of Table 1 gives the number of the object according The accuracy of zph is difficult to assess, because to the catalog of Fernandez-Soto et al. (1999); the only several galaxies are now known to have spectro- second column, its name according to Williams et al. scopic z exceeding 5 (see Introduction). However, in (1996); the third column, the apparent I814 magni- HDF-N, two such such objects were classified as tude in the AB system of Oke (1974); the fourth galaxies with z>5 based just on photometric red- column, the spectroscopic redshift (Spinrad et al. shifts, and their zph were subsequently corroborated 1998; Weymann et al. 1998); the fifth column, the through spectroscopic observations (Spinrad et al. photometric redshift zph according to Fontana et al. 1998; Weymann et al. 1998). For these two ob- (2000) and the last column, the spectral type of the jects, the difference between the two redshifts z − zph galaxy (i.e., the spectral type that best describes the amounts to −0.13 ± 0.22 (see Table 1).