Introduction to Cosmology

192 Cosmic Microwave Background

500

D

TEl

/l [

μK

2 ]

• 0.4


• 0.2

0.2

0.4 Planck 143 +^217

0

1000 1500 2000 2500

l

Figure 8.5The temperature–E-polarization cross-power spectrum as a function of angular
scale (topxaxis) and multipole moment (bottomxaxis) [6]. Reproduced from the freely
accessible Planck Legacy Archive with permission of George Efstathiou, Kavli Institute for
Cosmology, University of Cambridge. (See plate section for color version.)

Note the qualitative feature in Figure 8.3 that the TE component is zero at the
temperature–power spectrum maxima, as is expected (the polarization is maximal
at velocity maxima and minimal at temperature maxima, where the velocities are
minimal), and that it exhibits a significant large-angle anti-correlation dip at퓁≈150,
a distinctive signature of superhorizon adiabatic fluctuations. The positions and
amplitudes of the peaks and troughs in the temperature (T) power spectrum Figure 8.3
and the temperature–polarization (TE) cross-power spectrum in Figure 8.5 contain a
wealth of information on cosmological parameters.
In (훺m,훺휆)-space, the CMB data determine훺 0 =훺m+훺휆most precisely, whereas
supernova data (discussed in Section 4.4) determine훺휆−훺mmost precisely. Combin-
ing both sets of data with data on large-scale structures from 2dFGRS [11] (discussed
in Chapter 9) which depend on푛s,훺mℎ,훺bℎand which put limits on훺휈ℎ, one breaks
the CMB power spectrum parameter degeneracies and improves the precision.
It is too complicated to describe the simultaneous fit to all data and the evaluation
of the parameter values here, so we just quote the results published asPlanck+WP
by the Planck Team [6]. This includes information also from the WMAP 9-year polar-
ization low multipole data [12] Note that all errors quoted refer to single-parameter
fits at 68% confidence, marginalized over all the other parameters. If one takes into
account all recent parameter determinations, the weighted mean central values and
the median statistics central values are in fairly good agreement among themselves
and with the Planck values, but there is no absolute criterion to select the true values.
Planck+WP finds that the scalar index is

푛s= 0. 960 ± 0. 0073. (8.47)