Next Contents Previous


With all the observational progress in the CMB and the tying down of cosmological parameters, what can we anticipate for the future? Of course there will be a steady improvement in the precision and confidence with which we can determine the appropriate cosmological model and its parameters. We can anticipate that the evolution from one year to four years of WMAP data will bring improvements from the increased statistical accuracy and from the more detailed treatment of calibration and systematic effects. Ground-based experiments operating at the smaller angular scales will also improve over the next few years, providing significantly tighter constraints on the damping tail. In addition, the next CMB satellite mission, Planck, is scheduled for launch in 2007, and there are even more ambitious projects currently being discussed.

Despite the increasing improvement in the results, it is also true that the addition of the latest experiments has not significantly changed the cosmological model (apart from a suggestion of higher reionization redshift perhaps). It is therefore appropriate to ask: what should we expect to come from Planck and from other more grandiose future experiments, including the proposed Inflation Probe or CMBPol? Planck certainly has the the advantage of high sensitivity and a full sky survey. A detailed measurement of the third acoustic peak provides a good determination of the matter density; this can only be done by measurements which are accurate relative to the first two peaks (which themselves constrained the curvature and the baryon density). A detailed measurement of the damping tail region will also significantly improve the determination of n and any running of the slope. Planck should also be capable of measuring CellEE quite well, providing both a strong check on the Standard Model and extra constraints that will improve parameter estimation.

A set of cosmological parameters are now known to roughly 10% accuracy, and that may seem sufficient for many people. However, we should certainly demand more of measurements which describe the entire observable Universe! Hence a lot of activity in the coming years will continue to focus on determining those parameters with increasing precision. This necessarily includes testing for consistency among different predictions of the Standard Model, and searching for signals which might require additional physics.

A second area of focus will be the smaller scale anisotropies and `secondary effects.' There is a great deal of information about structure formation at z << 1000 encoded in the CMB sky. This may involve higher-order statistics as well as spectral signatures. Such investigations can also provide constraints on the Dark Energy equation of state, for example. Planck, as well as experiments aimed at the highest ells, should be able to make a lot of progress in this arena.

A third direction is increasingly sensitive searches for specific signatures of physics at the highest energies. The most promising of these may be the primordial gravitational wave signals in CellBB, which could be a probe of the ~ 1016 GeV energy range. Whether the amplitude of the effect coming from inflation will be detectable is unclear, but the prize makes the effort worthwhile.

Anisotropies in the CMB have proven to be the premier probe of cosmology and the early Universe. Theoretically the CMB involves well-understood physics in the linear regime, and is under very good calculational control. A substantial and improving set of observational data now exists. Systematics appear to be well understood and not a limiting factor. And so for the next few years we can expect an increasing amount of cosmological information to be gleaned from CMB anisotropies, with the prospect also of some genuine surprises.


We would like to thank the numerous colleagues who helped in compiling and updating this review, in particular John Kovac, Keith Olive, John Peacock, Clem Pryke, Paul Richards and Martin White. We are also greatful for the diligence of the RPP staff. This work was partially supported by the Canadian NSERC and the US DOE.

Next Contents Previous