Isotropy and Anisotropy

Photons were last in equilibrium with the rest of the Universe at decoupling, which occurred at a temperature of several thousand K, a time of a few hundred thousand years after the big bang, and a redshift of z ~ 1000 (recall: redshift is a measure of time). Since decoupling is the latest time that density fluctuations could have been imprinted on the photons that redshifted to become the microwave background, any structure in that background must date from this era. The adjacent diagram illustrates.
Last Scattering Surface
Because of finite lightspeed, the observable part of the Universe is restricted to the sphere marked as infinite redshift. Decoupling corresponds to a large but not infinite redshift of about 1000. This sphere, which is also marked, is sometimes called the last scattering surface, because it marks the smallest redshift at which the photons could have interacted with (scattered from) matter. With our assumptions about the Hubble constant, the last scattering surface is a sphere of radius about 9000 Mpc. Thus, structure in the microwave background would correspond approximately to this redshift. As noted in the diagram, the last scattering surface also separates a region of the Universe that is relatively transparent inside it from one that is rather opaque outside it (because before decoupling photons could only travel a short distance in matter before being scattered but after decoupling they could travel large distances without interaction).

There is nothing "special" about the last scattering surface. It merely marks the distance at which photons emitted at the time of decoupling would only now be reaching us. Decoupling happened all over the Universe during the same period of time, so any other observer situated at another place in the Universe will also have a last-scattering sphere like the one shown above for Earth, but centered on their location.

Smoothness and Possible Problems
Once the dipole component due to the motion of the Earth is subtracted from the microwave background distribution, it is extremely isotropic (it looks almost the same in any direction). The highly isotropic nature of the CMB indicates that in its very early stages the Universe was almost completely uniform. This raises two problems for the big bang theory.

1. First, when we look at the microwave background coming from widely separated parts of the sky it can be shown that these regions are too separated to have been able to communicate with each other even with signals traveling at light velocity since the birth of the Universe. How then did they "know" to have almost exactly the same temperature? This is called the horizon problem.
2. Second, the present Universe is homogeneous and isotropic, but only on very large scales. For scales the size of superclusters and smaller the luminous matter in the universe is quite lumpy, as we have already seen. If the very early Universe were completely uniform, how could such structure have formed?

Let us put aside the first problem for the moment (we shall return to it shortly when we discuss cosmic inflation; see the right frame) and concentrate on the second. We expect the very early Universe to have been highly uniform because of the strong coupling between photons and matter, which tends to smooth out fluctuations in density. However, if we are to understand the later growth of structure, there must be at least some deviation from uniformity in the microwave background.