4.3. Observing High Redshifts

We're used to quasars at very high redshifts. But quasars are rare and atypical - we'd really like to know the history of matter in general. One of the most important advances in recent years has been the detection of many hundreds of galaxies at redshifts up to (and even beyond) 5. Absorption due to the hundreds of clouds along the line of sight to quasars probes the history of cosmic gas in exquisite detail, just as a core in the Greenland ice-sheet probes the history of Earth's climate.

Quasar activity reaches a peak at around z = 2.5. The rate of star formation may peak at somewhat smaller redshifts (even though the very first starlight appeared much earlier) But for at least the last half of its history, our universe has been getting dimmer. Gas gets incorporated in galaxies and ``used up'' in stars - galaxies mature, black holes in their centres undergo fewer mergers and are starved of fuel, so AGN activity diminishes.

That, at least, is the scenario that most cosmologists accept. To fill in the details will need better simulations. But, even more, it will need better observations. I don't think there is much hope of ``predicting'' or modelling the huge dynamic range and intricate feedback processes involved in star formation. A decade from now, when the Next Generation Space Telescope (NGST) flies, we may know the main cosmological parameters, and have exact simulations of how the dark matter clusters. But reliable knowledge of how stars form, when the intergalactic gas is reheated, and how bright the first ``pregalaxies'' are will still depend on observations. The aim is get a consistent model that matches not only all we know about galaxies at the present cosmic epoch, but also the increasingly detailed snapshots of what they looked like, and how they were clustered, at all earlier times.

But don't be too gloomy about the messiness of the ``recent'' universe. There are some ``cleaner'' tests. Simulations can reliably predict the present clustering and large-scale distribution of non-dissipative dark matter. This can be observationally probed by weak lensing, large scale streaming, and so forth, and checked for consistency with the CMB fluctuations, which probe the linear precursors of these structures.