5.2 Structure Formation and Primeval Inhomogeneity

The COBE detection of CMB anisotropy on angular scales of 10° was a major milestone (Smoot et al. 1992), providing the first evidence for the fluctuations that seeded all the structure in the Universe and strong evidence for the gravitational instability picture for structure formation, as the size of the inhomogeneity was sufficient to explain the structure observed today. It also ushered in a powerful new probe of structure formation and dark matter. An early implication of COBE was galvanizing: nonbaryonic dark matter is required to explain the structure seen today. Because baryons are tightly coupled to photons in the Universe and thereby supported against gravitational collapse until after decoupling, larger amplitude density perturbations are required, which in turn lead to larger CMB temperature fluctuations than are observed.

Two key issues are the character and origin of the inhomogeneity and the quantity and composition of matter, discussed above. It is expected that there is a spectrum of fluctuations, described by its Fourier decomposition into plane waves. In addition, there are two generic types of inhomogeneity: curvature perturbations, fluctuations in the local curvature of the Universe which by the equivalence principle affect all components of the energy density alike; and isocurvature perturbations, which as their name indicates are not ingrained in the curvature but arise as pressure perturbations caused by local changes in the equation of state of matter and energy in the Universe.

The two most promising ideas for the fundamental origin of the primeval inhomogeneity are quantum fluctuations which become curvature fluctuations during inflation (Hawking 1982; Starobinskii 1982; Guth and Pi 1982; Bardeen et al. 1983) and topological defects (such as cosmic strings) that are produced during a cosmological phase transition (see e.g., Vilenkin & Shellard 1994). The inflation scenario will be discussed in detail later on. Topological defects produced in a cosmological symmetry-breaking phase transition around 10^-36 sec generate isocurvature fluctuations: the conversion of energy from radiation to defects leads to a pressure perturbation that propagates outward and ultimately leads to a density inhomogeneity. The defect scenario is currently disfavored by measurements of CMB anisotropy (Allen et al. 1997; Pen et al. 1997).

One graphic indicator of the progress being made on the large-scale structure problem is the number of viable models: the flood of data has trimmed the field to one or possibly two models. A few years ago the defect model was a leading contender; and another, more phenomenological model put forth by Peebles was also in the running (Peebles 1987). Peebles' model dispensed with nonbaryonic dark matter, assumed _B = ₀ ~ 0.2, and posited local variations in the distribution of baryons (isocurvature perturbations) of unknown origin. Its demise was CMB anisotropy: it predicted too much anisotropy on small angular scales. The one clearly viable model is cold dark matter plus inflation, which is discussed below. The challenge to theorists is make sure that at least one model remains viable as the quantity and quality of data improve!