5. ADDITIONAL ASPECTS

5.1. Results from numerical simulations

Cosmological simulations not only have been used to investigate the large-scale distribution and physical state of the warm-hot intergalactic medium, they also have been applied to predict statistical properties of high-ion absorption systems that can be readily compared with the UV and X-ray measurements (e.g., Cen et al. 2001; Fang & Bryan 2001; Chen et al.2003; Furlanetto et al. 2005; Tumlinson & Fang 2005; Cen & Fang 2006). Usually, a large number of artificial spectra along random sight-lines through the simulated volume are generated. Sometimes, instrumental properties of existing spectrographs and noise characteristics are modelled, too (e.g., Fangano et al. 2007). The most important quantities derived from artificial spectra that can be compared with observational data are the cumulative and differential number densities (dN / dz) of O VI, O VII, O VIII systems as a function of the absorption equivalent width. An example for this is shown in Fig. 8. Generally, there is a good match between the simulations and observations for the overall shape of the dN / dz distribution (see also Sect. 3.2), but mild discrepancies exist at either low or high equivalent widths, depending on what simulation is used (see, e.g., Tripp et al. 2007). For the interpretation of such discrepancies it is important to keep in mind that the different simulations are based on different physical models for the gas, e.g., some simulations include galaxy feedback models, galactic wind models, non-equilibrium ionisation conditions, etc., others do not. For more information on numerical simulations of the WHIM see Bertone et al. 2008 - Chapter 14, this volume.

Figure 8. The differential number of intervening oxygen high-ion (O VI, O VII,O VIII) absorbers in the WHIM in a cosmological simulation is plotted against the equivalent width of the absorption (for details see Cen & Fang 2006). While for O VII and O VIII no significant observational results are available to be compared with the simulated spectra (see Sect. 4.2), the predicted frequency of O VI absorbers is in good agreement with the observations (Sect. 3.2). Adapted from Cen & Fang (2006).

WHIM simulations also have been used to investigate the frequency and nature of BLAs at low redshift (Richter et al. 2006b). As the simulations suggest, BLAs indeed host a substantial fraction of the baryons at z = 0. From the artificial UV spectra generated from their simulation Richter et al. derive a number of BLAs per unit redshift of (dN / dz)_BLA 38 for H I absorbers with log (N(cm^-2) / b(km s^-1)) 10.7, b 40 km s^-1, and total hydrogen column densities N(H II) 10^20.5 cm^-2. The baryon content of these systems is ~ 25 percent of the total baryon budget in the simulation. These results are roughly in line with the observations if partial photoionisation of BLAs is taken into account (Richter et al. 2006a; Lehner et al. 2007). From the simulation further follows that BLAs predominantly trace shock-heated collisionally ionised WHIM gas at temperatures log T 4.4-6.2. Yet, about 30 percent of the BLAs in the simulation originate in the photoionised Ly forest (log T < 4.3) and their large line widths are determined by non-thermal broadening effects such as unresolved velocity structure and macroscopic turbulence. Fig. 9 shows two examples of the velocity profiles of BLAs generated from simulations presented in Richter et al. (2006b).

Figure 9. Two examples for BLA absorbers from the WHIM in a cosmological simulation are shown. The panels show the logarithmic total hydrogen volume density, gas temperature, neutral hydrogen volume density, and normalised intensity for H I Ly and O VI 1031.9 absorption as a function of the radial velocity along each sightline. From Richter et al. (2006b).

5.2. WHIM absorbers at high redshift

Although this chapter concentrates on the properties of WHIM absorbers at low redshift (as visible in UV and X-ray absorption) a few words about high-ion absorption at high redshifts (z > 2) shall be given at this point. At redshifts z > 2, by far most of the baryons are residing in the photoionised intergalactic medium that gives rise to the Ly forest. At this early epoch of the Universe, baryons situated in galaxies and in warm-hot intergalactic gas created by large-scale structure formation contribute together with only < 15 percent to the total baryon content of the Universe. Despite the relative unimportant role of the WHIM at high z, O VI absorbers are commonly found in optical spectra of high-redshift quasars (e.g., Bergeron et al. 2002; Carswell et al. 2002; Simcoe et al. 2004). The observation of intervening O VI absorbers at high redshift is much easier than in the local Universe, since the absorption features are redshifted into the optical regime and thus are easily accessible with ground-based observatories. However, blending problems with the numerous H I Ly forest lines at high z are much more severe than for low-redshift sightlines. Because of the higher intensity of the metagalactic UV background at high redshift it is expected that many of the O VI systems in the early Universe are photoionised. Collisional ionisation of O VI yet may be important for high-redshift absorbers that originate in galactic winds (see, e.g., Fangano et al. 2007). While for low redshifts the population of O VI absorbers is important for the search of the "mission baryons" that are locked in the WHIM phase in the local Universe, O VI absorbers at high redshift are believed to represent a solution for the problem of the "missing metals" in the early epochs of structure formation. This problem arises from the facts that at high redshift an IGM metallicity of ~ 0.04 is expected from the star-formation activity of Lyman-Break Galaxies (LBGs), while observations of intervening C IV absorption systems suggest an IGM abundance of only ~ 0.001 solar (Songaila 2001; Scannapieco et al. 2006), thus more than one order of magnitude too low. Possibly, most of the missing metals at high z are hidden in highly-ionised hot gaseous halos that surround the star-forming galaxies (e.g., Ferrara et al. 2005) and thus should be detectable only with high ions such as O VI rather than with intermediate ions such as C IV. Using the UVES spectrograph installed on the Very Large Telescope (VLT) Bergeron & Herbert-Fort (2005) have studied the properties of high-redshift O VI absorbers along ten QSO sightlines and have found possible evidence for such a scenario. Additional studies are required to investigate the nature of high-z O VI systems and their relation to galactic structures in more detail. However, from the existing measurements clearly follows that the study of high-ion absorbers at large redshifts is of great importance to our understanding of the formation and evolution of galactic structures at high z and the transport of metals into the IGM.

5.3. Concluding remarks

The analysis of absorption features from high ions of heavy elements and neutral hydrogen currently represents the best method to study baryon content, physical properties, and distribution of the warm-hot intergalactic gas in large-scale filaments at low and high redshift. However, the interpretation of these spectral signatures in terms of WHIM baryon content and origin still is afflicted with rather large systematic uncertainties due to the limited data quality and the often poorly known physical conditions in WHIM absorbers (e.g., ionisation conditions, metal content, etc.). Future instruments in the UV (e.g., COS) and in the X-ray band (e.g., XEUS, Constellation X) hold the prospect of providing large amounts of new data on the WHIM with good signal-to-noise ratios and substantially improved absorber statistics. These missions therefore will be of great importance to improve our understanding of this important intergalactic gas phase.

Acknowledgements. The authors thank ISSI (Bern) for support of the team "Non-virialized X-ray components in clusters of galaxies". SRON is supported fiancially by NWO, the Netherlands Organization for Scientific Research.