Relic gravitational waves: Latest revisions and preparations for new data

The forthcoming release of data from the Planck mission, and possibly from the next round of Wilkinson Microwave Anisotropy Probe (WMAP) observations, make it necessary to revise the evaluations of relic gravitational waves in the existing data and, at the same time, to refine the assumptions and data analysis techniques in preparation for the arrival of new data. We reconsider with the help of the commonly used CosmoMC numerical package the previously found indications of relic gravitational waves in the 7-year (WMAP7) data. The CosmoMC approach reduces the confidence of these indications from approximately $2\sigma$ level to approximately $1\sigma$ level, but the indications do not disappear altogether. We critically analyze the assumptions that are currently used in the cosmic microwave background (CMB) data analyses, and outline the strategy that should help avoid the oversight of relic gravitational waves in the future CMB data. In particular, it is important to keep away from the unwarranted assumptions about density perturbations. The prospects of confident detection of relic gravitational waves by the Planck satellite have worsened, but they are still good. It appears that more effort will be required in order to mitigate the foreground contamination.

Figure 1

One-dimensional likelihood functions for $n_s$ (upper panel) and $r$ (lower panel). In both panels, the olive curves, also marked by (1), denote the results for Case I, the magenta curves (2) denote the results for Case II, the red curves (3) for Case III, and the black curves (4) for Case IV. In both panels, the blue curves, also marked by (5), denote the results for the Case V, i.e. for the case with piecewise power-law spectrum. In the upper panel, the blue solid line (5) shows the likelihood function for $n_s^{(1)}$, whereas the blue dotted line (5) shows the likelihood function for $n_s^{(2)}$. In the lower panel, the blue line (5) shows the likelihood function for $r$.

Figure 2

The maximum likelihood values of $n_s$ and $r$ as functions of ${\ell }_{\max }$.

Figure 3

The marginalized one-dimensional distributions (constraints) of $n_s$ and $r$ as functions of ${\ell }_{\max }$. The vertical bars denote the uncertainties at 68% C.L.

Figure 4

The Planck’s signal-to-noise ratio $S/N$ for expected GW signals considered in the text.