There has been a steady improvement in the quality of CMB data that has led to the development of the present-day cosmological model. Probably the most robust constraints currently available come from the combination of the WMAP first year data  with smaller scale results from the CBI  and ACBAR  experiments. We plot these power spectrum estimates in Fig. 2. Other recent experiments, such as ARCHEOPS , BOOMERANG , DASI , MAXIMA  and VSA  also give powerful constraints, which are quite consistent with what we describe below. There have been some comparisons among data-sets , which indicate very good agreement, both in maps and in derived power spectra (up to systematic uncertainties in the overall calibration for some experiments). This makes it clear that systematic effects are largely under control. However, a fully self-consistent joint analysis of all the current data sets has not been attempted, one of the reasons being that it requires a careful treatment of the overlapping sky coverage.
Fig. 2 shows band-powers from the first year WMAP data , together with CBI and ACBAR data at higher . The points are in very good agreement with a `CDM' type model, as described in the previous section, with several of the peaks and troughs quite apparent. For details of how these estimates were arrived at, the strength of any correlations between band-powers and other information required to properly interpret them, turn to the original papers [5, 24, 25].
Figure 2. Band-power estimates from the WMAP, CBI, and ACBAR experiments. The WMAP data are the points, while squares are CBI and crosses ACBAR. We have shown only CBI and ACBAR data relevant for > 500, and both experiments also probe to higher than shown. This plot represents only a fraction of experimental results, with several other data-sets being of similar quality. The multipole axis here is linear, so the Sachs-Wolfe plateau is hard to see. The acoustic peaks and damping region are very clearly observed, with no need for a theoretical curve to guide the eye.