46 S. Alexander, N. Yunes / <strong>Physics</strong> <strong>Reports</strong> 480 (2009) 1–55I ≡ 〈 (a 1 ) 2 + (a 2 ) 2〉 , Q ≡ 〈 (a 1 ) 2 − (a 2 ) 2〉 , (290)U ≡ 〈2a 1 a 2 cos δ〉 , V ≡ 〈2a 1 a 2 sin δ〉 , (291)where δ ≡ ɛ 2 − ɛ 1 and the brackets signify a time average over a time long compared to ω −1 . The I parameter measures theintensity of the radiation, while the parameters Q , U, and V each carry information about the polarization of the radiation.Unpolarized radiation (so-called natural light) is described by Q = U = V = 0. The linear polarization of the radiation isencoded in Q and U, while the parameter V is a measure of elliptical polarization with the special case of circular polarizationocurring when a 1 = a 2 and δ = ±π/2. From here on we will simply refer to V as the measure of circular polarization, whichis technically correct if Q = 0.While I and V are coordinate independent, Q and U depend on the orientation of the coordinate system used on the planeorthogonal to the light’s direction of propagation. Under a rotation of the coordinate system by an angle φ, the parametersQ and U transform according toQ ′ = Q cos(2φ) + U sin(2φ), and U ′ = −Q sin(2φ) + U cos(2φ),while the angle defined byΦ = 1 ( ) U2 arctan ,Qgoes to Φ − φ following a rotation by the angle φ. Therefore, Q and U only define an orientation of the coordinate systemand not a particular direction in the plane: after a rotation by π they are left unchanged.Physically, such transformations are simply a manifestation of the oscillatory behavior of the electric field, which indicatethat Q and U are part of a second-rank symmetric trace-free tensor P ij , i.e. a spin-2 field in the plane orthogonal to thedirection of propagation. Such a tensor can be represented as( )P 0P ij = , (292)0 −Pin an orthonormal eigenbasis, where P = (Q 2 + U 2 ) 1/2 is ususally called the magnitude of linear polarization. For example,in two-dimensional, spherical polar coordinates (θ, φ), the metric is Ω ij = diag(1, sin 2 θ) and the polarization tensor is()Q (ˆn) −U(ˆn) sin θP ij (ˆn) =−U(ˆn) sin θ −Q (ˆn) sin 2 , (293)θwhere we recall that (i, j) run over the angular sector only.The temperature pattern on the CMB can be expanded in a complete orthonormal set of spherical harmonics:whereT(ˆn)T 0= 1 +a T (lm) = 1 T 0∫∞∑l=1 m=−ll∑a T (lm) Y (lm)(ˆn) (294)dˆn T(ˆn)Y(lm) ∗ (ˆn) (295)are the coefficients of the spherical harmonic decomposition of the temperature/polarization map and T 0 is the mean CMBtemperature. Likewise, we can also expand the polarization tensor in terms of a complete set of orthonormal basis functionsfor symmetric trace-free 2 × 2 tensors on the 2-sphere,P ij (ˆn)T 0=∞∑l=2 m=−ll∑ [aG(lm)Y G (ˆn) + (lm)ij aC (lm) Y C (ˆn)] (lm)ij, (296)where the expansion coefficients are given bya G (lm) = 1 ∫dˆnP ij (ˆn)Y G ij ∗(lm)T (ˆn), aC (lm) = 1 ∫0 T 0anddˆn P ij (ˆn)Y C ij ∗(lm)(ˆn). (297)The basis functions Y G (ˆn) (lm)ij and Y C (lm)ij(ˆn) are given in terms of covariant derivatives of spherical harmonics by(Y G = (lm)ij N l Y (lm):ij − 1 )2 Ω kijY (lm):k , (298)Y C = N l( )(lm)ij Y(lm):ik ɛ k j + Y (lm):jk ɛ k i , (299)2
S. Alexander, N. Yunes / <strong>Physics</strong> <strong>Reports</strong> 480 (2009) 1–55 47where ɛ ij is the completely antisymmetric tensor on the 2-sphere, a colon in an index list stands for covariant differentiationon the 2-sphere, and N 2 l≡ 2(l − 2)!/(l + 2)! is a normalization factor. Since the Y (lm) ’s provide a complete basis for scalarfunctions on the sphere, the Y G (lm)ij and Y C (lm)ijtensors provide a complete basis for gradient-type (G) and curl-type (C) STFtensors, respectively. This G/C decomposition is also known as the scalar/pseudo-scalar decomposition, which is similar tothe tensor spherical harmonic decomposition of Section 5.Integration by parts transforms Eqs. (297) into integrals over scalar spherical harmonics and derivatives of thepolarization tensor:a G (lm) = N lT 0∫a C (lm) = N lT 0∫dˆn Y ∗ (lm) (ˆn) P ij :ij (ˆn), (300)dˆn Y ∗ (lm) (ˆn) P ij :ik (ˆn)ɛ k j , (301)where the second equation uses the fact that ɛ ij :k = 0. Given that T and P ij are real, all of the multipoles must obey thereality condition a X (lm) ∗ = (−1)m a X (l,−m) , where X = {T, G, C}. The spherical harmonics Y (lm) and Y G (lm)ij have parity (−1)l , butthe tensor harmonics Y C (lm)ij have parity (−1)l+1 .The two-point statistics of the temperature/polarization map is then given viaC XX′l≡ 〈a X (lm) (aX′ (lm) )∗ 〉, (302)where the averaging is over all 2l + 1 values of m and over many realizations of the sky. This two-point statistic is thuscompletely specified by the six (TT , GG, CC, TG, TC, and GC) sets of multipole moments. If the temperature/polarizationdistribution is parity invariant, then C TCland C GClmust vanish due to the symmetry properties of the G/C tensor sphericalharmonics under parity transformations.Parity conservation, however, is a theoretical bias. Lue, Wang and Kamionkowski [134] provided the first time physicalscenario where C TCl= C GCl≠ 0 due to parity violation in the GW power spectrum of the CMB 12 . This physical scenarioconsisted of GWs sourced by the CS interaction term in Eq. (3) with (α, β) = (1, 0) and ϑ = f (φ) some polynomial functionof the inflaton field φ. As we have discussed, CS <strong>modified</strong> gravity leads to amplitude birefringence in GW propagation, whichin turn leads to an excess of left- over right-cicularly polarized GWs that lead to a non-vanishing C TCl[134].In order to understand this, consider GWs during the inflationary epoch. These waves stretch and become classical atwavelengths on the order of λ ∼ 1/µ, where µ ∼ 1/f ′ is some CS energy scale, until they eventually freeze as they becomecomparable to the Hubble radius. When the waves exit this radius, the fraction of the accumulated discrepancy betweenleft- and right-polarized GWs can be estimated through the indexɛ ∼ (M p /µ)(H/M P ) 3 ( ˙φ/H 2 ) 2 , (303)where H is the Hubble scale and f ′′ ∼ 1/µ 2 . The factor H 2 / ˙φ ∼ 10 −5 is associated with the amplitude of scalar densityperturbations, while H/M P < 10 −6 is related to the amplitude of tensor perturbations [134].Since long wavelength GWs produce temperature anisotropies of curl type, an excess of left- over right-polarized GWsproduces a nonzero C TCl. This is because the multipole coefficients a T,C(lm)will be non-vanishing (right) for circularly polarizedGW{a T (lm) = (δm,2 + δ m,−2 )A T l(k) even l (+),−i(δ m,2 − δ m,−2 )A T (k) l odd l (×), (304){a C (lm) = (δm,2 + δ m,−2 )A C l(k) even l (×),−i(δ m,2 − δ m,−2 )A C (k) l odd l ( + ), (305)where A T,Clare temperature brightness functions and (+, ×) stand for plus or cross, linear GW polarizations (see eg. [173]).Likewise the multipole coefficients for the gradient component of the CMB polarization are similar, with the replacementof A T,Clfor polarization brightness functions. For a left, circularly polarized GW, the sign of the even-l moments is reversed.The above equations allow one to understand why parity is not violated with linearly polarized GWs. For example, let usassume that only + modes are present, then C TClby construction. However, a right or left, circularly polarized GW possessesboth + and x modes, and thus the cross-correction is non-vanishing:C TCl= 2(2l + 1) −1 A T l (k)AC l(k). (306)Lue, Wang and Kamionkowski [134] conclude that the parameter ɛ in Eq. (303) could in principle be measured by apost-Planck experiment with a sensitivity of 35 µK, a result that was later confirmed by the more detailed study of Saito,et. al. [165]. Such results have aroused interest in the polarization detection community, pushing them to improve their12 Lue et al. noticed that the CS term violates both parity and time reflections. Thus, since gravity is insensitive to charge, CPT is conserved.
- Page 1 and 2: Physics Reports 480 (2009) 1-55Cont
- Page 3 and 4: S. Alexander, N. Yunes / Physics Re
- Page 5 and 6: S. Alexander, N. Yunes / Physics Re
- Page 7 and 8: S. Alexander, N. Yunes / Physics Re
- Page 9 and 10: S. Alexander, N. Yunes / Physics Re
- Page 11 and 12: S. Alexander, N. Yunes / Physics Re
- Page 13 and 14: S. Alexander, N. Yunes / Physics Re
- Page 15 and 16: S. Alexander, N. Yunes / Physics Re
- Page 17 and 18: S. Alexander, N. Yunes / Physics Re
- Page 19 and 20: S. Alexander, N. Yunes / Physics Re
- Page 21 and 22: S. Alexander, N. Yunes / Physics Re
- Page 23 and 24: S. Alexander, N. Yunes / Physics Re
- Page 25 and 26: S. Alexander, N. Yunes / Physics Re
- Page 27 and 28: S. Alexander, N. Yunes / Physics Re
- Page 29 and 30: S. Alexander, N. Yunes / Physics Re
- Page 31 and 32: S. Alexander, N. Yunes / Physics Re
- Page 33 and 34: S. Alexander, N. Yunes / Physics Re
- Page 35 and 36: S. Alexander, N. Yunes / Physics Re
- Page 37 and 38: S. Alexander, N. Yunes / Physics Re
- Page 39 and 40: S. Alexander, N. Yunes / Physics Re
- Page 41 and 42: S. Alexander, N. Yunes / Physics Re
- Page 43 and 44: S. Alexander, N. Yunes / Physics Re
- Page 45: S. Alexander, N. Yunes / Physics Re
- Page 49 and 50: S. Alexander, N. Yunes / Physics Re
- Page 51 and 52: S. Alexander, N. Yunes / Physics Re
- Page 53 and 54: S. Alexander, N. Yunes / Physics Re
- Page 55: S. Alexander, N. Yunes / Physics Re