It was stated in the early 1960s, shortly before the discovery of the CMB, that there were only facts in cosmology (by Peter Scheuer, see [Longair 1993]). In a similar spirit, we have argued that there are perhaps 4 facts and 4 half facts currently known from CMB anisotropies.
It has been recognized for some time that these anisotropies may answer some of our most fundamental questions about the Universe. The current CMB data already indicate that gravitational instability, in a mostly cold dark matter dominated universe, amplified initially small adiabatic fluctuations into the large-scale structure that we see today. There is the potential to show what inflationary-like process happened in the early Universe. And ultimately, the precise shape of the angular power spectrum holds the key to determining many of the fundamental cosmological parameters, either directly or in combination with other measurements.
However, while it is interesting to track progress in this field and to speculate on what it all means, it seems clear that theorists have had long enough to manoeuvre that the present data no longer strongly constrain any popular cosmological model. With the coming of long duration balloon flights, the imminent launch of the MAP satellite, and the commissioning of three new CMB interferometers, we expect that to change. The BOOMERANG team has already had a successful long duration balloon flight, and the analysis of that data set is eagerly awaited. Similar flights will undoubtedly follow, along with other large data sets from new ground-based experiments. The race is on, since MAP is scheduled for late 2000. A little later, sometime around 2006, will see the launch of the Planck Surveyor. Planck should supply us with essentially cosmic-variance limited information on all the angular scales relevant to primary anisotropies, over the full range of relevant frequencies. Figure 2 is an estimate of how well the power spectrum might be constrained after MAP and after Planck. With the proliferation of high precision data future `A' and `B' lists will be correspondingly longer and more detailed. Our attempt at prognostication is represented in our list `C':
FIG.2: The future of CMB anisotropies as possibly detected by MAP and by Planck, representing the potential state of knowledge roughly 5 and 10 years after the present.
Item 2 will be difficult, but we have no doubt that it will happen. MAP may yield some information, how much is difficult to estimate without more insight into the foregrounds. Currently planned ground-based experiments may also give detections. And Planck should provide polarization measurements over a reasonable range of scales. However, a full investigation of CMB polarization (and certainly the `curl' or B-mode component produced by tensors) may have to await an experiment even beyond Planck.
Item 3 potentially involves information from both 1 and 2. Ultimately we will learn something about high energy physics through understanding the way in which fluctuations were laid down in the early Universe, whether this involves discriminating any tensor component, measuring a changing spectral index, non-Gaussian signatures, or something else. Since the relevant energies are so far beyond what is achievable in particle accelerators, it is likely that cosmological phenomena will be the only way of constraining such models for quite some time. In addition to the `initial conditions', the evolution of the fluctuations will provide us with information on the properties of the dark matter in the Universe which may tie in directly to particle physics theories at the electroweak scale.
Item 4 includes a whole suite of potentially measurable effects, which can be thought of as processing the primary anisotropies. Examples include gravitational lensing, non-linear potential growth, Sunyaev-Zel'dovich effects, details of the reionization process, and extragalactic sources. There is a grey area between what is considered cosmic signal and what is considered a `foreground'. But whatever you call it, there is little doubt that data from the Planck mission, for example, are likely to be mined for many years for the additional astrophysical information they contain.
We expect rapid experimental progress in the next few years, and we trust that theoretical effort will be similarly feverish ([Bond 1996]). As a result, there will no doubt be more physical processes uncovered which affect CMB anisotropies. At present the CDM-dominated inflationary paradigm looks like it's in pretty good shape. Our `C' list may end up being quite inaccurate, and we can even imagine trouble for some entries in our `B' list. However, the spectral information from the CMB, together with the `A' list, provides a very solid foundation for the physics which generates the anisotropies. Therefore we are confident that whatever proves to be the ultimate such list, a thorough investigation of CMB anisotropies will hold the key to learning about the background space-time and formation of large-scale structure in the Universe.
DS is supported by the Natural Sciences and Engineering Research Council of Canada. MW is supported by the NSF.