**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

*High-Energy Astrophysics: American and Soviet Perspectives/Proceedings from the U.S.-U.S.S.R. Workshop on High-Energy Astrophysics*. Washington, DC: The National Academies Press. doi: 10.17226/1851.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

*High-Energy Astrophysics: American and Soviet Perspectives/Proceedings from the U.S.-U.S.S.R. Workshop on High-Energy Astrophysics*. Washington, DC: The National Academies Press. doi: 10.17226/1851.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

*High-Energy Astrophysics: American and Soviet Perspectives/Proceedings from the U.S.-U.S.S.R. Workshop on High-Energy Astrophysics*. Washington, DC: The National Academies Press. doi: 10.17226/1851.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

**Suggested Citation:**"CHAOTIC INFLATIONARY UNIVERSE AND THE ANISOTROPY OF THE LARGE-SCALE STRUCTURE." National Academy of Sciences. 1991.

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chaotic Inflationary Universe and the Anisotropy of the Large-Scale Structure G.V. CHIBISOV AND YU.V. SHTANOV P.N. Lebedev Physical Institute INTRODUCTION Inflationary universe models began their history from the seminal paper by Guth (1981), and since then they have won great recognition among physicists. The last years were marked by a considerable progress towards understanding the actual picture of inflation. It was realized that the inflationary universe is in fact chaotic (see Goncharov 1987 for a review), that globally it is strongly inhomogeneous, and that the inflation in the universe as a whole is eternal In such a picture the region available to modern observations is just a tiny part of the universe, in which inflation finished about 101° years ago. In spite of the great popularity of the chaotic inflationary universe models, it is usually taken for granted that their specific features (such as strong global inhomogeneity of the universe) can hardly lead to any observable consequences. The argument is that all we see is jUSt a tiny part of the universe, a region about 1028 cm, and the typical scales of considerable inhomogeneities are much greater than this size. In contrast to this opinion we want to show that such observable consequences really can Ernst. In spite of the tremendous spatial size of the inhomogeneities under consideration they are well inside the actual horizon of the modern observer. Hence their observable manifestation does not contradict causality. The problem Is to discover such a manifestation. The phenomenon we are going to discuss is closely connected with the origin of structure (galaxies, clusters, etc.) in the observable region. As it is now well Clown (for a review see Brandenberger 1985), primordial density fluctuations relevant to structure formation could originate from vacuum 68

HIGH-ENERGY ASTROPHYSICS 69 fluctuations at the inflationary stage during the permanent stretching of fluctuations scales. In most of the works on this topic the vacuum Ductua- tions evolution was considered on a spatially homogeneous background. At the same time, as already emphasized above, the inflationary universe is not homogeneous even on the classical level. Indeed, permanently produced fluctuations on scales bigger than the Hubble scale (i.e., with wave numbers k < aH) can be treated as a classical background inhomogeneity of the inflationary universe. Our main idea then is to consider the vacuum fluctuations evolution on the inhomogeneous background. An essential change is the change of the mode functions which determine the vacuum state. We will see that this leads to a distortion of the resulting primordial fluctuations spectrum as compared with that in a homogeneous background model The primordial fluctuations spectrum becomes anisotropic. And this in its turn results in the anisotropy of the observable large-scale structure. The phenomenon can be understood in terms of vacuum polarization. The classical background inhomogeneity polarizes the vacuum state, and this results in the distortion of the primordial fluctuations spectrum. One can see that our phenomenon is very similar to the well-known phenomenon discovered by Sachs and Wolfe (1967~. In this phenomenon the an~sotropy of the microwave background radiation is due to large-scale metric inhomogeneity. In our phenomenon the large-scale metric inhomo- geneity in a similar way influences vacuum fluctuations at the inflationary stage, resulting in the structure anisotropy. The main difference between these effects is that in our case the inhomogeneity on scales much bigger than the size of the observable region are significant. Indeed, vacuum fluctuations spreading with a speed close to the speed of light pass the distance to the true horizon, which at the present tune exceeds by many orders of magnitude the size of the observable region. Using the above-mentioned analogy with the Sachs and Wolfe phe- nomenon it is not difficult to make simple estimates for the value expected of the large-scale structure anisotropy. On the inflationary stage the small- scale vacuum fluctuations spread almost with the speed of light. Then they pass a Hubble distance in a Hubble time. As it can be shown, metric fluctuations on this distance are of the order h2' . 2/M4 (1) where y is the scalar field, the dot denotes its time denvative, and Mp is the P~nck mass (we are using the units in which ~ = c = 1~. The expression (1) also gives the order of the squared anisotropy produced in a Hubble time. 1b obtain an estimate for the squared anisotropy produced during the whole evolution we should integrate (1)

70 2 irk ~2 where ~ is He effective anisotropy degree and AMERICAN AND SOVIET PERSPECTIVES (2) (3) is a timelike parameter the expansion index. The upper integration limit in (23 corresponds to the moment of the Hubble radius crossing (when k = aH) and the lower one-to some initial moment If the scalar field evolution is the so-called slow rolling-down T = lNa _ i dV(~) 3H dy ' then using also the approximate expression valid at the inflationary stage H2 8~ V( ~ we obtain from (2) (4) (5) Vo-Ok 3Mp ' (6, where VO is the initial value of the scalar field potential energy, and Vk is the value at the moment of He Hubble radius crossing. Our simple estimate shows that the anistropy can be an appreciable value as typical values of VO are (Linde 1985) Vo ~ Mp, and Vk ~ VO. FLUCTUATIONS ON THE: INHOMOGENEOUS BACKGROUND (7) 1b obtain more exact expressions we must consider the evolution of the mode functions bake Hat derange He scalar field smaD-s~le fluctuation, on the nonhomogeneous background metric We will take into account only the fluctuations of the scalar type. Then we would have for the classical background metric (see Chibisov and Mukhanov 1983; Mukhanov 1985, 1988) ds2 = aid) t1 + 2~C)dy2-(1 - 2¢C)dX2] (8) where He is the so-called relativistic potential It describes the metric fluctuations on scales bigger than the Hubble scale

HIGH-ENER~ ~TROP~SICS 71 Be = | (2fi)3(q>(- ah + Ark al ), (9) where A) are unperturbed mode functions, and ak, a+ are the usual annihilation~reation operators. In (8) and (9) ~ is the conformal time. For A) we can use the approximate expression A) ~ g2'r97~ k3/2 exp(-iky + it. We must solve the equation for Ark Ok + m (I) Ark = 0, (10) (11) where O is me d'Alambertian on the background with the metric (8), and mom) = d2V(So)/d~2. We find the solution for t<Pk in the form wok = are- )(1 + am-iSk). (12) Here b~(°)k iS the solution on the homogeneous background, and Sk(y, i: is a complex valued function linear in (c. We develop Sk in powers of 1/1` Sk =k(So+ kSi,+. . .), and find the solution for SO and S:: So(77'x) = 2 J Achy x 0(0 q))dy, no (13) (14) rat S1(TJ,£) =i [I\ _ (~)2] J ~c(r1 x n(71 - ?~))(r7 rI)dr/, (15) so _ _ where n = k/k is a unit vector in k-direction. The solution in the shape (123 with (14) and (15) correspond to the choice of the "local" vacuum state for the scalar field. We trace the evolution of Ark iD the shape (12) with the solutions (14~- (15) up to the moment of the Hubble radius crossing. After this moment the evolution of the fluctuation becomes non-adiabatic. Moreover, the self-gravi~ of the scalar field becomes important, and we have to include the metric fluctuations into consideration. It is very convenient to work in terms of the gauge-invariant variable hi, which describes the metric fluctuations. Then In the victim of the Hubble radius crossing time we can

72 AMERICAN AND SOVIET PERSPECTIVES link the mode Ark to the mode ski ~ is easily done by meam of the linear perturbation theory equation Then we have (^ + 42.22~2) ~ 42 2 ~ d (tY Ok = ~(- )(1-iCk(9' X)), (16) (17) where Ck is a complex valued function, linear me and similar to Sk in (12~. Developing Ck in powers of 1/k similar to (13) Ci=k(Co+kCi+. . .), and counting powers of 1/k we obtain from (16) Co _ So, C:-So ~ i(n~V)So + Disc (18) (19) The evolution of Kim ~ after the Hubble radio crms~g is deter- mined by the linear perturbation theory (see Mukhanov 1988~. After this time the spatial shape of irk ~ "frozen" and only in amplitude changes with time. Thus at the moment of the end of the inflationary stage (~7'~ we have Okay], X)-Ok Off ~ ~1-~Ck-~0k' X). Note that A) is the unperturbed solution. PRIMORDIAL FLUCTUATIONS SPECTRUM (20, The expression (20) allows us to calculate the primordial fluctuations spectrum. 1b do this we calculate the correlation function (~(~)~), where (A--q>(pf,x). Developing Ck(:i~ in powers of ~ and then pro ceeding to the integration over new wave numbers k-VC0 ~=0 we will come to the standard shape of the correlation function (~(X)~(~)) = / (2 )3 ~ Hi- ) ~2 (1 + ~k)e3~, (21) with dln(k3 1 (I) 12 irk-- ~dlak - (22)

HIGH-ENERGY ASTROPHYSICS ark * J do (Vn)[~c(?7, x + n(rlk - ?7)) - ~C(rt' X - n(Rk - al))] I x = 0 770 73 It can be shown that the anisotropy which appeared in (21) and which is described by the value ZJk is quadrupole at least at 84%, that is Vk = no~n,~3 [~e,~ + {a;' M]' (23) where Ao<' is a traceless matrix with very weak (logarit}unic) dependence on k. A value which will characterize the degree of the anisotropy will be {,2 = tr(A2). (24) ~ obtain a characteristic value for the anistropy we must average (24) with respect to the random background inhomogeneities ~c. In fact this is av- eraging over the vacuum state of which the inhomogeneities me originated. Using (22) we can obtain an approximate result irk y2 where (25) d In~k3 1~°) {2 . (26) The expression (25) differs from the estimate (2) only by a factor. Hence instead of the estimate (6) we will have <62> ~ Ilk Vo-Vk 3 Me (27) It can be shown by direct calculations that for the values of k typical for the present clustering ~lk/3 ~ 10 - 2 Thus for typical values of VO Even by (7) we obtain (~ 10~1 that is, the allotropy of the order of 10%. CHAOTIC INFLATION (28) (~29) Now let us take into account one of the peculiar features of the chaotic inflationary universe. Namely, that such a universe consists of independent

74 AMERICAN AND SOVIET PERSPECTIVES domains of a physical size [ph = 0~-~)- Hey are independent ~ a sense that physical processes inside any such domains do not depend on the processes inside any other domain (Goncharov et al. 1987~. From this it follows that the integration in (25) must be taken over a history of the domain of which our observable region has originated. The size of this domain at all times during the inidationa~y stage must be taken to be of the order of the Hubble size. Then the history of such a domain becomes random. For the random history of the scalar field evolution ANTI a Fo~er- Planck equation can be obtained (Starobinsky 1986; see also Goncharov et al. 1987 for a review) where ({' ~ = ~ ~-APT 2(BP)~], A ~' B H2 (30) (31) P($o, r) is a -dependent scalar field distn~ution in the domain considered. Now let us denote the solution to (303 with the initial conditions P($o,0) = b(~-~0) by Z(~,TO,T). Then a formal solution for Z(SP,~0,7) can be written in terms of path integral over all trajectories that start at ~ and at A: Z(~, SooT) = /(D~) exp(-St~) where is the "action," and (32) Sty= (ad-A)2B-l ODD) = n B 2 (<p(~))d~p(a) (33) (34) is the Pam integration measure. From the expression (33) it is clear that the most probable path is that which satisfies do -= A, (35) and this Is just the slow rolling-down expression (43. Thus we come to a conclusion that our estimate (27) is the most probable value of (25~. ANISOTROPY OF TEE LARGE-SCAIE STRUCTURE It is not difficult to show that the anisotropy of the primordial fluc- tuations spectrum loads the anisotropy of the large-scale structure. The

HIGH-ENERGY ASTROPHYSICS 75 effect is conserved on scales on which nonlinear processes do not yet play an important role, that is, scales more than 50 Mpc. We can easily cal- culate the two-point correlation function (by) = (~0~), where Aid) = [p(~/(p>. The integrals over k-space must be taken with some small-scale cutoff (Peebles 1980~. This can be done by a simple cut-off function exp (-1~1) where e is a cut-off scale. Then the isotropic part of the correlation function is found to be Q4y4 (y~2 + it and the anisotropic part is where (36) ((x) =-hit ~2' ~(37) h(X) = e4y4 (y~2 + i)3 (38) y = x/¢, and C in (36) and (38) is some t-independent constant. One can use the correlation function calculated to test experimentally the phenomenon discussed In this paper. DISCUSSION In conclusion, it is easy to show that during the greater part of the inflationary stage ache vacuum fluctuations scales of interest are much smaller than the Planck scale. One such scale, an old-fashioned field theory, is likely to be invalid and one is supposed to use, for example, the superstring theory (Green e' al. 1987~. This last however is still in progress, and we don't yet know how to handle it when dealing with such small scales. So we use a field theoretical approach hoping that further investigations based on a more advanced theory will not change our result dramatically. In context of the superstring theory, the phenomenon discussed might be caused by string vacuum polarization due to large-scale inhomogeneity of the inflationary universe. The problem of small scales disappears to some extent if one takes into account that the fluctuation scale is not a gauge invariant variable and thus can be changed by me change of the reference frame. This can be well illustrated in the case of flat space-time: a photon can be of any energy, in particular, of energy exceeding the Planck energy.: 1Ihe authors are grateful to B.L Spokoiny and AA Starobinsly for this remark

76 AMERICAN AND SOVIET PERSPECTIVES From the observational point of view our result means a possibility of the large-scale structure anisotropy together with highly isotropic microwave background radiation. Observational discovery of the phenomenon consid- ered might serge as confirmation to the chaotic inflation scenario and, in fact, throw some light upon physics on very small scales, in the context discussed just above. NOTE ADDED IN PROOF: The extended version of this work is published in the Int. J. Mod. Phys. A52625 (1990~. ACKNOWLEDGEMENTS The authors egress their gratitude to UP. Gnshchok, N~ Gurench, AD. Linde, OF. Mukhanov, B.L. Spokoiny, and NA Starobinsly for valuable discussions. REFERENCES Brandenberger, R.H. 1985. lees. Mod. Phys. 57: 1. Chibisov, G.V., and V.~. Mukhanov. 1983. Preprint FLAN 154. Goncharov, AS., ~D. Iinde, and VF. Mukhanov. 1987. Int. J. Mod. Phys A2: 561. Green, M.B., J.H. Schwarz, and E. Written. 1987. Superseding Theory. Cambridge Universitr Press, Cambridge. Vols 1,3 Guth, A.H. 1981. Phyla Rev. D23: 347. Iinde, ~D. 1985. Phys. Lett. 16213: 281. Mukhanov, V.F. 1985. Pistma Zh. Eksp. Tear. His. 41: 40Z Mukhanov, V.F. 1988. Zb. Eksp. Tear. Fix. 94: 1. Peebles, P.J.E. 1980. I-he Large-Scale Structure of the Universe Princeton University Press, Pnnceton. Sachs, R.K, and ~M. Wolfe. 1967. Ast~phys J. 147: 73. Starobinsk~r, AN 1986. In: de Vega, HJ. and N. Sanchez (eds.~. Field Theory, Quantum Gravity and Strings. Springer, Berlin. Lecture notes in Phys. 246: 107.