||Cosmology based on CMB experiments entered a
new era early in the 1990s, when NASA's Cosmic Background Explorer (COBE)
flew an experiment conceived by George Smoot, an astrophysicist with
Berkeley Lab's Physics Division and professor of physics at the University
of California at Berkeley. Smoot's COBE experiment measured small
variations in the temperature of the otherwise featureless microwave
background for the first time.
The cosmic microwave background represents the moment, 300,000 years after the hot Big Bang, when temperatures had cooled enough for protons to capture free electrons and form hydrogen atoms. Photons, freed from scattering interactions with electrons, streamed away from the vanished surface of this photon-electron-baryon fluid. As they reach us today, these photons have lost energy, "cooled off," to the equivalent of those emitted by a black body at a temperature less than three degrees above absolute zero.
Before this fluid condensed into atoms and photons, pressure oscillations were moving at the speed of sound through the dense hot blob. Gravity pulled in and radiation pressure pushed out, making the whole spherical mass ring like a bell; at the moment the photons were freed, some regions were contracting, heated by compression, and other, cooler regions were expanding; these motions imprinted temperature differences on the cosmic microwave background and recorded the harmonics of the sound wave.
The sound wave reveals the precise characteristics of its resonating chamber -- which was the entire universe at that moment. Measuring the smallest possible angles between pairs of warmer or cooler points on a fine-scale map of today's microwave sky yields a CMB power spectrum, a graph that reveals such basic parameters as whether the universe is flat or open.
In a flat universe, with familiar geometry, the sound wave would have traveled across a cosmos equivalent in dimension to a single degree of today's sky; thus the oscillations we see today would repeat on a scale of about one degree. But an open universe is one in which space is "negatively curved;" similar oscillations today would appear smaller. Thus by determining the scale at which the oscillations repeat -- closer to one degree, or closer to half a degree -- it is possible to determine whether the universe is flat or open.
Other kinds of information also reside in the pattern of anisotropies. "We owe the stars and galaxies -- all the concentrations of matter essential to our existence -- to minute density perturbations in the primordial universe," says astrophysicist and computer scientist Julian Borrill of the National Energy Research Computing Center (NERSC). "Alternative cosmologies try to account for these perturbations in different ways. Each model generates a specific waveform, starting with a large peak representing the fundamental harmonic. The harmonic in inflation models is followed by resonances, a series of smaller peaks. Models incorporating defects in the early universe probably just 'ramp down.' To choose among models, we need to derive the real curve."
The COBE data established points on the lower slope of the first peak of the waveform; analysis of the BOOMERanG North America data set has roughed out the shape of this peak, whose position favors a flat universe and whose large amplitude is consistent with the idea that some unknown force, in addition to an initial push from the Big Bang, is driving universal expansion -- presumably the cosmological constant. The evidence for the cosmological constant is further strengthened by comparing independent data from distant supernovae.
To extend the curve of CMB power spectrum, the current generation of CMB experiments were designed to measure even smaller anisotropies, over even more sky, at even higher resolution. The COBE satellite detected variations in temperature as small as one part in 100,000 and mapped the sky with an angular resolution of about 10 degrees; the PLANCK satellite, planned to fly in 2007, will detect variations as small as one part in a million with a resolution of 10 arc-minutes -- and over the entire sky.
Apart from the sheer size of the new data sets, separating very weak signals from noise will pose a mounting challenge to computational analysis of the CMB in the years to come. NERSC is determined to meet the challenge, whose reward will be a treasury of priceless information about the structure and evolution of our universe, from its earliest visible moment until the present.