Annu. Rev. Astron. Astrophys. 2005. 43:
861-918 Copyright © 2005 by Annual Reviews. All rights reserved |
We end this review with the following question: What have we learned from damped Ly systems that we did not know before? We attempt to answer this question by listing results judged to be robust. These are also summarized in Tables 2 and 3, which describe cosmological and local properties, respectively. Table 3 lists both the medians and the means to show the effects of the upper limits placed on various parameters. Specifically, the means only include positively detected quantities while the medians include upper limits.
Property | z | N a | b | xmed c | d | Min e | Max f |
log10N(HI) | z > 1.6 | 199 | 20.83 | 20.60 | 0.33 | 20.30 | 21.70 |
[M/H] | 0.3 < z < 4.9 | 130 | -1.11 | -1.48 | 0.55 | -2.65 | 0.04 |
[Zn/Fe] | 0.7 < z < 3.3 | 38 | 0.54 | 0.42 | 0.25 | -0.01 | 1.05 |
[/Fe] | 0.8 < z < 4.7 | 70 | 0.42 | 0.38 | 0.18 | 0.03 | 1.00 |
vlow g | 1.7 < z | 95 | 114. | 90. | 83.7 | 16. | 430. |
vhigh h | 1.7 < z | 75 | 209. | 190. | 113.4 | 20. | 528. |
log10 f(H2) | 2.0 < z < 3.4 | 33 | -2.22 | < -5.93 | 0.82 | < -6.98 | -0.64 |
log10 c i | 1.7 < z < 4.2 | 57 | -26.57 | < - 26.93 | 0.49 | < - 27.69 | -25.35 |
G0 j | 1.7 < z < 4.5 | 39 | 9.6 | 5.4 | 6.5 | <0.24 | 23 |
log10 * k | 1.7 < z < 4.2 | 40 | -1.95 | -2.20 | 0.30 | < - 3.55 | -1.55 |
a Number of DLAs in sample
|
Most of the neutral gas in the Universe in the redshift interval 0 < z < 5 is in damped Ly systems. The cosmology and mean intensity of extragalactic radiation are sufficiently well known to justify the assumption of gas neutrality for N(H I) 2 × 1020 cm-2. The close agreement between the mass per unit comoving volume of neutral gas in damped Ly systems and visible matter in current galaxies indicates that damped Ly systems comprise a significant neutral-gas reservoir for star formation at high redshift.
The comoving density of neutral gas, g(z), decliens by a factor of two between z 3.5 and z = 2.3. While the evolution at 0 < z < 2.3 is more uncertain, g(z) at z 3.5 is a factor of three higher than at z = 0. The neutral-gas content of the Universe varies little in the redshift interval z = [1.6, 4.5]. However, g(z) at z = 0 is about three times lower than the average value within z = [1.5, 4.5].
Damped Ly systems are metal poor at all redshifts (see Table 1), but exhibit a metallicity "floor," [M/H] -2.6 (Table 1), indicating a different enrichment history than that of the Ly forest.
The cosmic metallicity doubles every Gyr at z > 2, but the median [M/H] is sub-solar at z < 1.6.
From the large [Zn/Cr] ratios and the increase of the [Zn/Fe] and [Si/Fe] ratios with increasing metallicity we know that damped Ly systems exhibit evidence for depletion by dust and that the dust content is far lower than in the Galaxy.
The presence of a plateau in the [N / ] versus [ / H] plane near [N / ] -0.7 indicates a minimum age of 0.25 Gyr for damped Ly systems, which suggests that they are not transient objects but instead probably have ages comparable to the Hubble time at the absorption epoch.
Ionized gas in damped Ly systems exhibits a different velocity structure from the neutral gas, unlike the agreement between the velocity structures of these two phases in the Galaxy.
H2 and other molecules are rarely present in damped Ly systems. Studies of those systems exhibiting H2 absorption indicate the presence of an FUV radiation field with J 10-19 ergs cm-2 s-1 Hz-1 sr-1, which resembles (a) the interstellar radiation field in the Galaxy and (b) J predicted by the C II* technique.
The frequency distribution of the absorption velocity intervals, v, has a median of 90 km s-1. This property cannot be reproduced by single-disk CDM scenarios proposed so far, and is difficult to reproduce for sightlines passing through dwarf galaxies. Damped Ly systems with large values of v exhibit a systematic absence of low values of [M/H] and high values of N(H I).
We cannot rule out the hypothesis that galaxies identified with damped Ly systems at z < 1.6 are drawn from a cross-section weighted sample of normal galaxies; i.e., an inflated populations of dwarfs is not required.
C II* 1335.7 absorption is detected in about half of randomly selected samples of damped Ly systems. The inferred [C II] 158 µm cooling rates indicate heating rates far in excess of those supplied by FUV background radiation, requiring a local heat source. The evidence accumulated so far suggests that the likely site of C II* absorption is gas in a cold neutral medium (CNM).
Next, we describe critical unsolved problems in damped Ly research.
What is the median mass, Mmed, of the dark-matter halos containing damped Ly systems? This is the critical diagnostic for discriminating among most hierarchical models, in which Mmed < 109 M, from hierarchical models with feedback or passive evolution models, in which Mmed > 1011 M.
Does the damped Ly luminosity function overlap that of Lyman Break Galaxies? Partial overlap is suggested by the luminosities of the few objects detected in emission.
What are the properties of the interstellar gas in damped Ly systems? These are crucial for understanding whether or not the gas in which C II* absorption is detected can support a CNM.
Are stars forming in damped Ly systems when they are detected?
How are the star-formation and accretion histories of damped Ly systems related? The C II* technique indicates that star formation depletes the neutral-gas reservoir of damped Ly systems more rapidly than indicated by the decrease of y with time at z 2.3. Does this require accretion of neutral gas onto damped Ly systems at rates comparable to the star formation rates?
What is the solution to the "missing metals" problem? Evidence for metal-enriched gas ejected from damped Ly systems, or for light mainly emitted from compact "bulge" regions would help in deciding between these hypotheses.
What is the intrinsic nucleosynthetic [Si/Fe] ratio in damped Ly systems? Are the intrinsic abundances of damped Ly systems -enhanced?
What is the cosmic metallicity of low-z damped Ly systems? Is the column-density weighted mean metallicity of low-z damped Ly systems biased by undersampling and by obscuration?
How can we improve numerical simulations of damped Ly system evolution? The next steps involve more accurate modeling of star formation and mechanical feedback.
A major goal of damped Ly system research is to give a clear and decisive answer to the question, "What is a damped Ly system?" Obviously this has not yet been accomplished. Rather, what we have found is that a significant fraction of damped Ly systems are a population of H I layers exhibiting many of the complexities of the ISM of the Galaxy. They clearly play an important role in the formation of galaxies and undoubtedly interact with other structures in the high-redshift Universe through a variety of feedback mechanisms. Observations of damped Ly systems provide an amazingly rich data set that gives information about galaxy formation unavailable by other means. Specifically, observations of damped Ly systems are the only way to study in detail the neutral gas that gave rise to galaxies at high redshifts. We hope that the interplay between new observations and improved theoretical modeling will lead to significant insights into the process of galaxy formation.
Acknowledments
This review was written while one of us (AMW) was on sabbatical leave at the Institute of Astronomy, Cambridge, and AMW wishes to thank the Institute of Astronomy for the hospitality extended to him during his visit and for the award of a Sackler fellowship. AMW is particularly grateful to Max Pettini for many valuable discussions about our favorite mutual topic. AMW and JXP also wish to thank the Kavli Institute of Theoretical Physics, Santa Barbara, for the hospitality extended to them during their attendance at the Galaxy Intergalactic Medium Interactions program. This material is based on work supported by the National Science Foundation under Grant No. AST 03-07824 awarded to AMW and JXP and Grant No. AST-0201667 awarded to EG.