Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.arcetri.astro.it/science/k20/papers/ve03.ps Äàòà èçìåíåíèÿ: Tue Feb 17 13:47:56 2004 Äàòà èíäåêñèðîâàíèÿ: Fri Feb 28 06:57:21 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: astro-ph

arXiv:astro­ph/0401101
v2
9
Jan
2004
The near-infrared view of galaxy evolution
Andrea Cimatti
INAF - Osservatorio Astrosico di Arcetri, Largo E. Fermi 5, I-50125, Firenze, Italy
Abstract. Near-infrared surveys provide one of the best opportunities to investigate
the cosmic evolution of galaxies and their mass assembly. We brie y review the main
results obtained so far with the K20 and other recent near-IR surveys on the redshift
distribution, the evolution of the luminosity function and luminosity density, the nature
of old and dusty EROs, the evolution of the galaxy stellar mass function, the properties
of the galaxies in the \redshift desert" and the nature of luminous starbursts at z  2.
1 Introduction
Despite the detection of objects up to z  6:5 and the impressive success of
the CDM scenario to account for the properties of the cosmic microwave back-
ground, one of the main and still controversial questions remains how and when
galaxies assembled their mass as a function of the cosmic time. The hierarchical
scenario predicts that galaxies built up their present-day mass through a pro-
gressive assembly of smaller sub-systems driven by the merging of dark matter
halos.
With ground-based observations, one of the most solid approaches to address
the problems of galaxy formation and evolution is to study samples of eld
faint galaxies selected in the near-infrared, particularly in the K-band (2.2m)
[3,23]. Firstly, since the rest-frame near-IR luminosity is a good tracer of the
galaxy stellar mass [5], K-band surveys allow to select galaxies according to
their mass up to z  1:5 ( rest  0:9 1:0m). At higher redshifts, the K-
band starts to sample the rest-frame optical and UV regions, and space-based
observations at  obs > 2m are needed to cover the rest-frame near-infrared (e.g.
SIRTF, ASTRO-F). Secondly, the similarity of the spectral shapes of dierent
galaxy types in the rest-frame near-IR makes the K-band selection free from
strong biases against or in favor of particular classes of galaxies. In contrast, the
selection of high-z galaxies in the observed optical bands is more sensitive to
the star formation activity than to the stellar mass because it samples the rest-
frame UV light and makes optical samples biased against old passive or weakly
star-forming galaxies. Last but not least, near-infrared surveys are less aected
by dust extinction than optical surveys.
Motivated by the above considerations, several near-infrared surveys have
been undertaken during the last decade [12,10,6,18,24,17,15,19] (Fig. 1). Spec-
troscopic surveys are particularly relevant thanks to their capability not only to
derive the redshifts, but also to unveil the nature and the spectral properties

2 Andrea Cimatti
Cimatti et al. 2002/2003 (K20, VLT)
92%
Cowie et al. 1996 (HDS, Keck)
70%
Cohen et al. 1999 (CFGRS, Keck)
84%
Stern
et al.
SPICES
(Keck)
?%
Drory et al. 2003 (90% zphot, MUNICS)
10% Firth et al. 2002 (zphot,LCIR)
H­band
HDFS/FIRES
Franx et al.
Giallongo et al.
SDF
Maihara et al.
HDFN
H­band
Dickinson et al.
GOODS/CDFS (ESO public)
GDDS (Abraham et al., Gemini)
?%
ELAIS­S1 (Cimatti et al., ESO/NTT)
NOAO Deep Wide­Field (Dey/Jannuzi)
ODTS (INT 2.5m)
Fig. 1. Some imaging and spectroscopic near-infrared surveys. The redshift complete-
ness is shown for spectroscopic surveys.
of the targeted galaxies. Since near-IR multi-object spectrographs are not fully
available on 8-10m-class telescopes, most spectroscopy of K-selected galaxies has
been done in the optical. The most crucial probes of massive galaxy evolution
are galaxies with the very red colors expected in case of passive evolution at
z > 1 (e.g. R K s > 5, usually called Extremely Red Objects, EROs). However,
for the typical limiting uxes of near-IR surveys (K s < 20 21), they have very
faint optical magnitudes (R  25 26) already close to the spectroscopic limits
of 8-10m-class telescopes. In addition, for z > 1 their main spectral features
(e.g. the 4000  A break and H&K absorptions) fall in the very red part of the
observed spectra, where the strong OH sky lines and the CCD fringing and low
quantum eôciency make spectroscopy even more demanding.

Galaxy Evolution in Near-IR Surveys 3
2 The K20 survey
The K20 survey is a project aimed at investigating galaxy evolution through deep
ESO VLT spectroscopy of a sample of 546 objects with K s < 20 selected from
two elds covering a total area of 52 arcmin 2 . The two elds are completely inde-
pendent, and one of them is a sub-area of the Chandra Deep Field South. Most
spectroscopy was done in the optical with FORS1 and FORS2 and designed
to reach the highest possible signal-to-noise ratio in the red. ISAAC near-IR
spectroscopy was also done for a small fraction of the sample. The multi-band
(UBV RIzJK s ) imaging available for both elds was used to derive high-quality
photometric redshifts \trained" with the spectroscopic redshifts. The nal sam-
ple covers a redshift range of 0 < z < 2:5, and the current redshift completeness
is 92% (spectroscopic) and 100% (spectroscopic + photometric redshifts).
The K20 survey represents a signicant improvement with respect to other
surveys for faint K-selected galaxies thanks to its high spectroscopic redshift
completeness (the highest to date) extended to very faint red objects, the larger
sample, the coverage of two elds and the availability of optimized photometric
redshifts. We note here that the distribution of the sample over two independent
sky elds is a signicant advantage in reducing the eld-to-eld variation eects.
For more details on the K20 survey see [6] and http://www.arcetri.astro.it/k20/.
3 Established results: galaxy evolution to z  1
Overall, the K20 and other recent near-IR surveys show that galaxies are char-
acterized by little evolution to z  1, so that the observed properties can be
mimicked by a pure luminosity evolution (PLE)-like scenario. This is in con-
trast with the current CDM hierarchical merging models where the assembly
of galaxies occurs later than what is actually observed. In particular:
 For K s < 20, N (z) has a median redshift of z med  0:8 and a high-z tail
extended beyond z  2. Current hierarchical models do not match the observed
redshift distribution [8,18] (Fig.2), but an improved treatment of the merging
processes is providing encouraging results [29,25].
 The rest-frame K s -band luminosity function shows a mild luminosity evolu-
tion up to z  1, with a brightening of about 0.5 magnitudes. Signicant density
evolution is ruled out up to z  1 [11,26,17]. The high-luminosity tail at z  1
is dominated by red early-type galaxies [26]. Current hierarchical models fail
in reproducing the shape and the evolution of the LF as they over-predict the
number of sub-L  galaxies, under-predict that of luminous galaxies, and predict
a strong density evolution.
 The rest-frame K s -band luminosity density evolves slowly to z  1 with
Ks (z) / (1 + z) 0:37 , and much slower than the UV luminosity density [26].
 The properties of \old" EROs with K s < 20 at z  1 (morphology, spectra,
luminosities, ages, stellar masses, clustering) imply the existence of a substantial
population of old (a few Gyr), passively evolving and fully assembled massive
spheroids which requires that major episodes of massive galaxy formation oc-
curred at least at z form  2 [7,9]. Their number density at z  1 is consistent

4 Andrea Cimatti
Cole et al. 2000
Menci et al. 2002
Somerville et al. 2001
Pozzetti et al. 2002
Menci et al.
Ks<19.8
Fig. 2. The K20 survey N(z) compared with the predictions of three hierarchical merg-
ing model (dotted and dashed lines) and a pure luminosity evolution (PLE) model
(solid line). The models refer to total magnitudes Ks < 20 and are not corrected for
the average under-estimate of 0.2 magnitudes due to the photometric selection eects
evaluated in the K20 sample (see [6]). Including this bias in the models results in im-
proving the agreement with the PLE models and increasing the discrepancy with the
hierarchical models, as shown in the inserted plot where the model of Menci et al.
(2002) for Ks < 19:8 is compared to the observed z > 1 tail of N(z) (see [8]).
(within 2) with that of local luminous E/S0 galaxies. Current hierarchical mod-
els severely under-predict the number of \old" EROs [7,9].
 A numerous population of dusty star-forming EROs with disk-like and ir-
regular morphologies emerges at 0:7 < z < 1:7 from near-IR surveys [7,28,9].
These objects are often too faint to be detected in submm surveys due to their
inferred far-IR luminosities < 10 12 L and represent an ensemble of galaxies
important to complement other star-forming systems selected with dierent te-
chiniques. They are also expected to be important contributors to the cosmic

Galaxy Evolution in Near-IR Surveys 5
star-formation density [7,9,4,28]. Also in this case, current hierarchical models
under-estimate the number of \dusty" EROs.
 \Old" EROs seem to have much stronger clustering than \dusty" EROs,
with a comoving r 0 similar to that of present-day luminous ellipticals [13].
 The number density of massive galaxies and the cosmic stellar mass density

 show a slow decrease from z  0 to z  1 qualitatively consistent with a
hierarchical scenario, but much slower than what the current models predict
[20,2,16,11,15,19].
4 Beyond z  1: galaxies in the \redshift desert"
At higher redshifts, the picture becomes more controversial: at z > 1 the evolu-
tion of the near-IR luminosity function seems to depart from a PLE-like pattern
[26], it is still not clear if the number density of old spheroids drops at z > 1:3
[1,27], and unexpected populations of massive star-forming galaxies are being
found in the range of 1:5 < z < 3 thanks to near-IR surveys [21,31,14].
Moreover, the galaxy stellar mass function
and
 (z) display a fast evolution
at z > 1 [15,19], and the K20 survey results suggest a rapid increase of the stellar
mass density from 20 +20
10 % of the local value at 2 < z < 3 to about 100% at
z  1 [20].
These results suggest that 1:5 < z < 2:5 may be the critical cosmic epoch
during which most star formation activity and galaxy mass assembly took place.
However, probing the 1:5 < z < 2:5 range is diôcult due to the lack of strong
spectral features and emission lines redshifted in the observed optical and to the
strong OH emission sky lines severely aecting the spectra beyond 8000  A.
Due to the limited number of spectroscopically identied galaxies, this redshift
range has been traditionally called \redshift desert".
Thanks to the very deep red-optimized optical spectroscopy and to near-IR
spectroscopy, the K20 survey allowed us to populate the \redshift desert" and
to unveil the nature of K-selected galaxies in this critical redshift range. By
combining deep HST+ACS imaging with K20 spectroscopy, we found that a
variety of galaxy types are present at 1:5 < z < 2:5: spheroids at z  1:5, disk-
and spiral-like systems, and many irregular merging-like objects.
Particularly relevant is a sub-sample of luminous galaxies at 1:7 < z < 2:3
spectroscopically identied in the K20 survey [14]. Their inferred star formation
rates of  100 500 M yr 1 , stellar masses up to 10 11 M derived from the
tting of their multi-color spectral energy distributions, and their merging-like
optical morphologies becoming more compact in the near-IR, suggest that these
galaxies may be massive galaxies caught during their major episode of mass
assembly and star-formation, possibly being the progenitors of the present-day
massive spheroidal systems. Current semi-analytical models under-predict by a
factor of  30 the number density of such galaxies [14,29]. Other near-IR surveys
are independently nding candidate massive systems at z  2 3 [21,31].
In order to draw a coherent picture of galaxy evolution at 1:5 < z < 2:5,
it will be crucial to combine K-selected samples (e.g. K20, FIRES [21], GDDS

6 Andrea Cimatti
[22]) with UV- and optically-selected samples more biased toward star-forming
objects with little dust extinction such as those resulting from the Lyman-break
selection at z > 1:4 [30].
5 Acknowledgments
The K20 survey team includes: T. Broadhurst (HUJ), S. Cristiani (INAF-Trieste),
S. D'Odorico (ESO), E. Daddi (ESO), A. Fontana (INAF-Roma), E. Giallongo
(INAF-Roma), R. Gilmozzi (ESO), N. Menci (INAF-Roma), M. Mignoli (INAF-
Bologna), F. Poli (University of Rome), L. Pozzetti (INAF-Bologna), A. Renzini
(ESO), P. Saracco (INAF-Brera), and G. Zamorani (INAF-Bologna).
References
1. Benitez N. et al. 1999, ApJ, 515, L65
2. Brinchmann J., Ellis R.S. 2000, ApJ, 536, L77
3. Broadhurst, T., Ellis, R.S., & Grazebrook, K. 1992, Nature, 355, 55
4. Brusa M. et al. 2002, ApJ, 581, L89
5. Gavazzi G., Pierini D. & Boselli A. 1996, A&A, 312, 397
6. Cimatti A., Mignoli M., Daddi E. et al. 2002c, A&A, 392, 395
7. Cimatti A., Daddi E., Mignoli M. et al. 2002a, A&A, 381, L68
8. Cimatti A., Pozzetti L., Mignoli M., et al. 2002b, A&A, 391, L1
9. Cimatti A., Daddi E., Cassata P. et al. 2003, A&A, in press, astro-ph/0310742
10. Cohen J.G., Blandford R., Hogg D.W. et al. 1999, ApJ, 512, 30
11. Cohen J.G. 2002, ApJ, 567, 672
12. Cowie, L.L., Songaila A., Hu E.M. & Cohen J.G. 1996, AJ,112,839
13. Daddi E., Cimatti A., Broadhurst T. et al. 2002, A&A, 384, L1
14. Daddi E., Cimatti A., Renzini A. et al. 2004, ApJL, in press, astro-ph/0308456
15. Dickinson M., Papovich C., Ferguson H.C., Budavari T. 2003, ApJ, 587, 25
16. Drory N., Bender R., Snigula J. et al. 2001, ApJ, 562, L111
17. Drory N., Bender R., Feulner G. et al. 2003, ApJ, 595, 698
18. Firth A.E., Somerville R.S., McMahon R.G. et al. 2002, MNRAS, 332, 617
19. Fontana A., Donnarumma, I. Vanzella E. 2003, ApJ, 594, L9
20. Fontana A. et al. 2003, A&A, submitted
21. Franx M., Labbe I., Rudnick G. et al. 2003, ApJ, 587, L79
22. Glazebrook K. et al. 2003, astro-ph/0311045
23. Kaumann G.,Charlot S., White S.D.M. 1996, MNRAS, 283, L117
24. Labbe I., Franx M., Rudnick G. et al. 2003, AJ, 125, 1107
25. Menci N. et al. 2004, ApJ, in press, astro-ph/0311496
26. Pozzetti L. et al. 2003, A&A, 402, 837
27. Rodighiero G., Franceschini A., Fasano G. 2001, MNRAS, 324, 491
28. Smail I. et al. 2002, ApJ, 581, 844
29. Somerville R.S. et al. 2004, ApJL, in press, astro-ph/0309067
30. Steidel C.S. et al., this volume
31. van Dokkum P.G., Forster Schreiber N.M., Franx M. et al. 2003, ApJ, 587, L83