REVIEWS OF MODERN PHYSICS, VOLUME 79, OCTOBER–DECEMBER 2007 Nobel Lecture: Cosmic microwave background radiation anisotropies: Their discovery and utilization* George F. Smoot Lawrence Berkeley National Laboratory, Space Sciences Laboratory, Department of Physics, University of California, Berkeley, California 94720, USA ͑Published 2 November 2007͒ DOI: 10.1103/RevModPhys.79.1349 I. THE COSMIC BACKGROUND RADIATION 3000 K there were insufficient energetic CMB photons to keep hydrogen or helium atoms ionized. Thus the Observations of the cosmic microwave background primeval plasma of charged nuclei, electrons, and pho- ͑CMB͒ temperature anisotropies have revolutionized tons changed into neutral atoms plus background radia- and continue to revolutionize our understanding of the tion. The background radiation could then propagate universe. The observation of the CMB anisotropies an- through space freely, though being stretched by the con- gular power spectrum with its plateau, acoustic peaks, tinuing expansion of the universe, while baryonic matter and high frequency damping tail have established a stan- ͑mostly hydrogen and helium atoms͒ could cluster by dard cosmological model consisting of a flat ͑critical gravitational attraction to form stars, galaxies, and even density͒ geometry, with contents being mainly dark en- larger structures. For these structures to form there must ergy and dark matter and a small amount of ordinary have been primordial perturbations in the early matter matter. In this successful model the dark and ordinary and energy distributions. The primordial fluctuations of matter formed its structure through gravitational insta- matter density that will later form large scale structures bility acting on the quantum fluctuations generated dur- leave imprints in the form of temperature anisotropies ing the very early inflationary epoch. Current and future in the CMB. observations will test this model and determine its key cosmological parameters with spectacular precision and confidence. B. Cosmic background radiation rules As a young undergraduate I heard of Penzias and Wil- A. Introduction son’s ͑1965͒ discovery of the 3 K background radiation and its interpretation by Dicke, Peebles, Roll, and Wilk- In the big bang theory the cosmic microwave back- ingon ͑1965͒, but not until two or three years later did I ground ͑CMB͒ radiation is the relic radiation from the begin to understand the implications and opportunity it hot primeval fireball that began our observable universe afforded. I was a first year graduate student at MIT about 13.7 billion years ago. As such the CMB can be working on a high-energy physics experiment when Joe used as a powerful tool that allows us to measure the Silk, then a graduate student at nearby Harvard, pub- dynamics and geometry of the universe. The CMB was lished a paper ͑Silk, 1967͒ entitled “Fluctuations in the first discovered by Penzias and Wilson at Bell Labora- Primordial Fireball” with the abstract “One of the over- tory in 1964 ͑Penzias and Wilson, 1965͒. They found a whelming difficulties of realistic cosmological models is persistent radiation from every direction which had a the inadequacy of Einstein’s gravitational theory to ex- thermodynamic temperature of about 3.2 K. At that plain the process of galaxy formation.16 A means of time, physicists at Princeton ͑Dicke, 1965; Dicke et al., evading this problem has been to postulate an initial 1965͒ were developing an experiment to measure the spectrum of primordial fluctuations.7 The interpretation relic radiation from the big bang theory. Penzias and of the recently discovered 3 °K microwave background Wilson’s serendipitous discovery of the CMB opened up 8,9 the new era of cosmology, beginning the process of as being of cosmological origin implies that fluctua- transforming it from myth and speculation into a real tions may not condense out of the expanding universe until an epoch when matter and radiation have decou- scientific exploration. According to big bang theory, our 4 universe began in a nearly perfect thermal equilibrium pled, at a temperature TD of the order of 4000 °K. The state with very high temperature. The universe is dy- question may then be posed: would fluctuations in the namic and has been ever expanding and cooling since its primordial fireball survive to an epoch when galaxy for- birth. When the temperature of the universe dropped to mation is possible?” My physics colleagues dismissed this work as specula- tion and not a real scientific enquiry. It seemed to me a *The 2006 Nobel Prize for Physics was shared by John C. field ripe for observations that would be important no Mather and George F. Smoot. This paper is the text of the matter how they came out. Obviously, there were galax- address given in conjunction with the award. ies. Determining if the radiation was cosmic was critical. 0034-6861/2006/79͑4͒/1349͑31͒ 1349 ©2006 The Nobel Foundation 1350 George F. Smoot: Nobel Lecture: Cosmic microwave background … If the 3 K microwave background was cosmic, it must rems of Penrose, Hawking, and Geroch to show that if contain imprints of fluctuations from a very early epoch the CMB was the relic radiation of the big bang, and if it when energies were very high. Silk’s work also made me were observed to be isotropic to a high degree, e.g., a realize the enormously important role of the cosmic part in 100, that one could not avoid having a singularity background radiation in the early universe. Going back in the early universe. The rough argument goes that, if to earlier times when the universe was smaller, one the CMB is cosmological and uniform to a high level, would reach the epoch when the radiation was as bright say one part in X, then one could extrapolate the uni- as the sun. At this epoch the universe was roughly a verse backwards to a time when it was 1/X smaller. If X thousand times smaller than present. This is impres- is sufficiently large, then the energy density in the CBR sively small but one could readily and reasonably ex- ͑microwaves now much more intense and hotter͒ would trapolate back another thousand in size and then the be sufficient to close the universe and cause it to ex- radiation would be a thousand times hotter than the 1 trapolate right back to the singularity. The only premises sun. in the argument were ͑i͒ the CMB was cosmological, ͑ii͒ But in truth, if this was the relic radiation, then the it would be found to be uniform to about a part in 10 000 ͑ pioneering calculations of Gamow and co-workers Al- ͑X=100 in their original optimistic argument but actu- pher, Bethe, and Gamow, 1948; Alpher, Herman, and ally 10 000 in present understanding͒, ͑iii͒ general rela- Gamow, 1948; Alpher and Herman, 1953, 1988, 1990; tivity or a geometric theory of gravity are the correct Alpher et al., 1953, 1967͒ tell us we can comfortably and description, and ͑iv͒ the energy condition that there is reliably look back to the point where the universe was a no substance which has negative energy densities or billion ͑109͒ times smaller. This is the epoch of primor- large negative pressures. Hawking and Ellis provided dial nucleosynthesis when the first nuclei form and their strongly plausible arguments against violation of the en- calculations correctly predicted the ratio of hydrogen to ergy condition. This observation would certainly be a helium and the abundance of a few light elements. At death blow to the numerous popular oscillating universe that epoch the temperature of the radiation was a mil- models and other attempts to make models without a lion times greater ͑and 1024 times brighter͒ than that of primordial singularity. Once again we see theorists pro- the sun. Any object placed in that radiation bath would viding arguments for the cosmic implications that could be nearly instantly vaporized and homogenized. Even be drawn from observations of the CMB—if it were atoms were stripped apart. At such early times the nu- truly cosmological. clei of atoms would be blown apart. The very early uni- One needed to be of two minds about the CMB: ͑i͒ be verse had to exist in a very simple state completely skeptical and test carefully to see that it was not the relic dominated by the cosmic background radiation which ͑ ͒ would tear everything into its simplest constituents and radiation of the big bang and ii assume that it was the spread it uniformly about. relic radiation and had the properties expected and then Also in 1967 Dennis Sciama published a paper look for the small deviations and thus information that it ͑Sciama, 1967͒ pointing out that if this were relic radia- could reveal about the universe. Early on one had to tion from the big bang, one could test Mach’s principle make a lot of assumptions about the CMB in order to and measure the rotation of the universe by the effect use it as a tool to probe the early universe, but as more that rotation would have on the cosmic microwave back- and more observations have been made and care taken, ground. It could rule out Godel’s model of a rotating these assumptions have been tested and probed more universe and its implied time travel supporting Mach’s and more precisely and fully. The history of the obser- principle and keeping us safe from time tourists. Here vations and theoretical developments is rife with this ap- was another fundamental physics and potentially excit- proach. The discovery of the CMB by Penzias and Wil- ing observation that one could make, if the CMB were son was serendipitous. They came upon it without cosmological in origin. having set out to find it or even to explore for some new Not long after ͑submitted October 1967, published thing.