Highly-ionized oxygen absorbers in the intergalactic medium

Recent ultraviolet and X-ray observations of intergalactic O VI and O VII absorption systems along lines of sight to bright quasars have opened a new window on to the 'warm-hot intergalactic medium'. These systems appear to provide a significant reservoir for baryons in the local Universe, and comparison to cosmological simulations suggests that their abundance roughly matches theoretical predictions. Here we use analytical arguments to elucidate the physical properties of the absorbers and their role in structure formation. We first show that if the absorbers result from structure-formation shocks, the observed column densities naturally follow from post-shock-cooling models, if we include fast-cooling shocks as well as those that cannot cool within a Hubble time. In this case, the known O VI absorbers should show stronger O VII absorption than expected from collisional-ionization equilibrium (and much more than expected for photoionized systems). We then argue that higher-temperature shocks will be spatially associated with more massive virialized objects even well outside the virial radius. Thus, the different oxygen ions will trace different structures; O VII absorbers are the most common because that ion dominates over a wide temperature range (corresponding to a large range in halo mass). If each dark matter halo is surrounded by a network of shocks with total cross-section a few times the size of the virialized systems, then we can reproduce the observed number densities of absorbers with plausible parameters. A simple comparison with simulations shows that these assumptions are reasonable, although the actual distribution of shocked gas is too complex for analytical models to describe fully. Our models suggest that these absorbers cannot be explained as a single-temperature phase.