Predictions in eternal inflation

Sergei Winitzki

DOI: https://doi.org/10.1007/978-3-540-74353-8_5

2007-01-03

Abstract:In generic models of cosmological inflation, quantum fluctuations strongly influence the spacetime metric and produce infinitely many regions where the end of inflation (reheating) is delayed until arbitrarily late times. The geometry of the resulting spacetime is highly inhomogeneous on scales of many Hubble sizes. The recently developed string-theoretic picture of the "landscape" presents a similar structure, where an infinite number of de Sitter, flat, and anti-de Sitter universes are nucleated via quantum tunneling. Since observers on the Earth have no information about their location within the eternally inflating universe, the main question in this context is to obtain statistical predictions for quantities observed at a random location. I describe the problems arising within this statistical framework, such as the need for a volume cutoff and the dependence of cutoff schemes on time slicing and on the initial conditions. After reviewing different approaches and mathematical techniques developed in the past two decades for studying these issues, I discuss the existing proposals for extracting predictions and give examples of their applications.

General Relativity and Quantum Cosmology,Astrophysics,High Energy Physics - Theory

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to statistically predict cosmological quantities observed at random locations in the context of eternal inflation. Due to eternal inflation, there are an infinite number of regions in the universe with different reheating times. Therefore, the location of an observer on Earth cannot provide information about its specific location in the eternally inflating universe. This makes it necessary to have a statistical framework to predict physical quantities observed at random locations.

Specifically, the paper discusses the following key issues:

1. **The problem of statistical prediction**: How to obtain statistical predictions in an eternally inflating universe, especially in the absence of information about the observer's specific location.
2. **The need for volume cutoff**: In order to make effective statistical predictions, a volume cutoff needs to be introduced to avoid problems caused by infinite volume.
3. **Dependence on cutoff schemes**: The dependence of different volume cutoff schemes on time slices and initial conditions, and how these dependencies affect prediction results.
4. **Review of existing methods**: The paper reviews different methods and techniques developed in the past two decades for studying these problems.
5. **Suggestions for extracting predictions**: Discusses existing methods for extracting predictions and gives examples of the application of these methods.

### Mathematical and physical background

Some of the key formulas involved in the paper include:

- **Hubble parameter**:
\[
H(\phi)=\sqrt{\frac{8\pi}{3}V(\phi)}
\]
where \(V(\phi)\) is the potential energy function and \(\phi\) is the scalar field (i.e., the inflaton field).

- **Growth of quantum fluctuations**:
\[
\langle\hat{\phi}^2(t + \Delta t)\rangle-\langle\hat{\phi}^2(t)\rangle=\frac{H^3}{4\pi^2}\Delta t
\]
This formula describes the linear growth of scalar - field quantum fluctuations over a time interval \(\Delta t\).

- **Langevin equation**:
\[
\dot{\phi}(t, x)=v(\phi)+N(t, x)
\]
where \(v(\phi)=-\frac{1}{4\pi}\frac{dH}{d\phi}\) is the deterministic drift term and \(N(t, x)\) is the noise term, which describes the influence of quantum fluctuations.

- **Noise correlation function**:
\[
\langle N(t, x)N(\tilde{t},\tilde{x})\rangle = C(t,\tilde{t},|x - \tilde{x}|;\phi)
\]
In particular, the noise correlation function at the same point is:
\[
C(t, t, 0;\phi)=\frac{H^2(\phi)}{4\pi^2}
\]

- **Diffusion coefficient**:
\[
D(\phi)=\frac{H^3(\phi)}{8\pi^2}
\]

- **Langevin equation in finite - difference form**:
\[
\phi(t+\Delta t)-\phi(t)=v(\phi)\Delta t+\sqrt{2D(\phi)\Delta t}\xi(t)
\]
where \(\xi(t)\) is a normalized random variable representing white noise.

### Conclusion

By reviewing and analyzing different methods, the paper aims to provide a more solid foundation for statistical prediction in an eternally inflating universe. By understanding these statistical frameworks and methods, observational data can be better explained and predicted, thus providing support for the development of cosmological theory.