Facility (SIRTF), which will provide much more extensive coverageof the spectrum and will cover the whole sky. Space experiments havein fact already provided important results. For example, the intensityof the total infrared radiation from the sky has been accuratelymeasured by the Diffuse Infrared Background Experiment aboard theCOBE satellite. Much of this radiation is from the Milky Way, andidentifying the cosmological component is a difficult task.

In the future, a main goal will be to image distant galaxies at thehighest possible angular resolution, both with the HST and with groundbasedtechniques. The development of adaptive optics technology, whichsharpens the view of ground-based telescopes, and its applicationto distant galaxy research will open a powerful new channel of informationabout the distant universe. The angular size of a typical distantgalaxy is about the same as the size of the blurring caused by lookingthrough Earth's atmosphere; hence almost all of the detail is scrambled.Sharp images of distant galaxies are important because they are atest of evolutionary models. For example, it is becoming increasinglyclear that an important process in the early evolution of galaxiesinvolves galaxies merging together, or at least strongly interactingwith each other. If this idea is correct, astronomers expect to seemore multiple, merging, and disturbed galaxies at high redshifts,when the frequency of this activity was higher. The first deep imagesof the universe coming from HST seem to bear this out. The HST imageon the back cover is typical in showing many disturbed and interactingobjects among the smaller, and thus presumably more distant, objects.

Evolution of Large-Scale Structure Back in Time

Since density fluctuations tend to grow, the amplitude of the densityvariations associated with the large-scale structure must have beensmaller when the universe was younger. The unique capability of largetelescopes to look deep into space corresponds to being able to lookback in time—cosmologists can map the distant universe and see thegalaxy distribution as it was billions of years ago. By comparingdifferent depths in space, cosmologists can in effect “make a movie” of the developing structure. Successive scenes in the movie arefirst galaxy formation, then cluster formation, and finally superclusteringtoday. Present optical and x-ray data hint strongly that clusteringin the universe continues to grow rapidly, but these observationsare still primitive. The evolution of galaxies and large-scale structureis a sensitive probe of alternative models of structure formation,but one that has been little utilized to date. With the completionof giant new optical telescopes (such as the Keck (Figure 5 ), the Gemini, and other large telescopes under construction), aswell as the refurbishment of the HST with a new, advanced camera,progress in this important field should accelerate.

Looking all the way back to a time when the universe was only a quarterof its present age requires maximum light-gathering power since verydistant galaxies must be observed. There are several requirementsfor such studies—the largest possible optical-infrared telescopeswith wide fields of view; spectrographs capable of measuring manygalaxies simultaneously; the largest possible optical CCDs and infraredarray detectors; and deep surveys in other wavelengths, includingradio and x-ray regions.

Supernovae, Quasars, and Absorption Line Systems: Probes for Cosmology

In a closed cosmological model (e.g., q0 ≥ 0.5, Λ = 0), space is positively curved and finite. A two-dimensional analogis the curved and finite surface of a sphere. In an open model, spaceis negatively curved and infinite. A twodimensional analog is a hyperboloid,which is shaped like a saddle. At a given redshift, sources of thesame intrinsic luminosity appear to be larger and brighter in a closeduniverse than in an open universe, because of the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 25
Cosmology: A Research Briefing Facility (SIRTF), which will provide much more extensive coverageof the spectrum and will cover the whole sky. Space experiments havein fact already provided important results. For example, the intensityof the total infrared radiation from the sky has been accuratelymeasured by the Diffuse Infrared Background Experiment aboard theCOBE satellite. Much of this radiation is from the Milky Way, andidentifying the cosmological component is a difficult task. In the future, a main goal will be to image distant galaxies at thehighest possible angular resolution, both with the HST and with groundbasedtechniques. The development of adaptive optics technology, whichsharpens the view of ground-based telescopes, and its applicationto distant galaxy research will open a powerful new channel of informationabout the distant universe. The angular size of a typical distantgalaxy is about the same as the size of the blurring caused by lookingthrough Earth's atmosphere; hence almost all of the detail is scrambled.Sharp images of distant galaxies are important because they are atest of evolutionary models. For example, it is becoming increasinglyclear that an important process in the early evolution of galaxiesinvolves galaxies merging together, or at least strongly interactingwith each other. If this idea is correct, astronomers expect to seemore multiple, merging, and disturbed galaxies at high redshifts,when the frequency of this activity was higher. The first deep imagesof the universe coming from HST seem to bear this out. The HST imageon the back cover is typical in showing many disturbed and interactingobjects among the smaller, and thus presumably more distant, objects. Evolution of Large-Scale Structure Back in Time Since density fluctuations tend to grow, the amplitude of the densityvariations associated with the large-scale structure must have beensmaller when the universe was younger. The unique capability of largetelescopes to look deep into space corresponds to being able to lookback in time—cosmologists can map the distant universe and see thegalaxy distribution as it was billions of years ago. By comparingdifferent depths in space, cosmologists can in effect “make a movie” of the developing structure. Successive scenes in the movie arefirst galaxy formation, then cluster formation, and finally superclusteringtoday. Present optical and x-ray data hint strongly that clusteringin the universe continues to grow rapidly, but these observationsare still primitive. The evolution of galaxies and large-scale structureis a sensitive probe of alternative models of structure formation,but one that has been little utilized to date. With the completionof giant new optical telescopes (such as the Keck (Figure 5 ), the Gemini, and other large telescopes under construction), aswell as the refurbishment of the HST with a new, advanced camera,progress in this important field should accelerate. Looking all the way back to a time when the universe was only a quarterof its present age requires maximum light-gathering power since verydistant galaxies must be observed. There are several requirementsfor such studies—the largest possible optical-infrared telescopeswith wide fields of view; spectrographs capable of measuring manygalaxies simultaneously; the largest possible optical CCDs and infraredarray detectors; and deep surveys in other wavelengths, includingradio and x-ray regions. Supernovae, Quasars, and Absorption Line Systems: Probes for Cosmology In a closed cosmological model (e.g., q0 ≥ 0.5, Λ = 0), space is positively curved and finite. A two-dimensional analogis the curved and finite surface of a sphere. In an open model, spaceis negatively curved and infinite. A twodimensional analog is a hyperboloid,which is shaped like a saddle. At a given redshift, sources of thesame intrinsic luminosity appear to be larger and brighter in a closeduniverse than in an open universe, because of the

OCR for page 25
Cosmology: A Research Briefing Figure 5. The first Keck telescope, atop Mauna Kea, Hawaii. This 10-m-diameteroptical telescope is the first in a new generation and will be joinedby the second Keck telescope. (© 1992 Roger Ressmeyer—Starlight.) focusing effects of the curvature and the more rapid deceleration.Astronomers endeavor to identify and employ classes of bright sourcesof known or calculable luminosity as “standard candles.” By measuring the apparentluminosity at various redshifts of these sources, cosmologists candetermine whether the universe is closed or open. Supernova explosionsprovide sources of this kind because their intrinsic brightness isgoverned by the physics of the explosion, of which there is goodtheoretical understanding. Quasars are star-like objects with large redshifts and inferred luminositiesthat are often hundreds of times those of normal galaxies. They arethought to occur in the nuclei of galaxies, but the conditions requiredfor a galaxy to harbor a quasar are not known. Quasars show strongevolution in the sense that they emitted much more energy at earliercosmic epochs, but why this is so is also unknown. Moreover, theobserved rapid variation in brightness of individual quasars is notunderstood. The task of understanding the nature and origin of quasarsis still at the forefront of cosmological research. Quasars can serve as background lamps against which absorption fromintervening material can be detected spectroscopically. The materialmay be in the form of gas in galaxies (the galaxy itself may or maynot be visible) or in the form of intergalactic clouds of gas thathave never been processed through stars. The technique has exceptionallyhigh sensitivity to small amounts of material, and so it providesa probe of the universe that is independent and complementary tothat provided by visible galaxies. The change in the average numberof absorbers as a function of redshift is an important diagnosticfor understanding the evolution of these objects. Spectroscopic attributesof the absorbers can tell us about their physical properties as wellas the intensity and spectrum of the intergalactic radiation fallingon them. These absorption studies have shown that the absorbing gas occursin lumps, and that there is little neutral hydrogen in a smoothlydistributed intergalactic medium. One explanation of this lack ofneutral hydrogen is that, at some point, the entire universe wasreionized—heated so hot that hydrogen atoms were broken up into theirconstituent protons and electrons. But if the universe was reionized,what were the heating agents and when did the reionization occur?On the other hand, if there was no epoch of reionization, what conditionsyielded such high efficiency in clearing intergalactic space of neutralhydrogen? This field will be advanced with the further identificationof close pairs of quasars to provide nearly coincident lines of sight,as well as the further identification of galaxies likely to be responsiblefor individual absorption

OCR for page 25
Cosmology: A Research Briefing systems. At present these absorption lines provide one of our onlyprobes of nonluminous structure at high redshift. Continued theoreticalas well as observational studies provide our best hope for an accuratepicture of the intergalactic medium and its evolution. Gravitational Lenses What are gravitational lenses, and why are they important? The gravitational lens is a relatively newly discovered phenomenonthat is emerging as an important research tool in cosmology. Lensingcan be produced when light propagating through the universe is deflectedby the gravitational field of a massive object positioned near itspath. Gravitational lensing effects are similar to those producedwhen a glass lens deflects the path of light rays in a camera, butthe deflections are too small to be observed in a terrestrial laboratory.However, in 1979 the discovery of a double quasar, which was actuallytwin images of the same quasar, provided the first convincing demonstrationthat gravitational lensing produces observable effects in the cosmos. Gravitational lensing has been observed in several forms. There arenow approximately 20 known cases of strong lenses in which two ormore images of a background source are produced by a foreground gravitationallens. This striking phenomenon is the most easily recognized, andearly work in gravitational lensing concentrated on the study ofstrong lensing. Figure 6 shows an Einstein cross, a lens that produces four multiple imagesof a distant quasar with a central object. Another manifestationis weak lensing, in which background sources, though not multiplyimaged, are visibly distorted by the presence of the interveninggravitational field (Figure 7). Changes in the gravitational field, produced for example by relativemotion of the source and the lens, have been observed through thechanges in the magnification of the source that produce variationsin the source brightness. The multiple images found in strong lensesare Figure 6. The image of an Einstein cross produced by a gravitational lens.A distant quasar is precisely aligned behind the foreground galaxywhose gravitational field deflects the light from the quasar intofour distinct images. (Courtesy of the Space Telscope Science Institute.) associated with different propagation times from source to observer,and the resulting time delay between the arrival times of signalsfrom each image has also been observed. In the best studied case,the measured time delay from one image to the other is approximately1.5 years. Finally, the presence of a population of massive objects,such as galaxies, screening a population of background sources, suchas quasars, produces lensing effects detectable through statisticalanalysis of the ellipticities of the lensed images. Gravitational lenses provide a unique opportunity to infer the propertiesof the space-time in which they are embedded, the mass distributionof the lens, and the detailed properties of the background source.These opportunities are being realized as instruments improve inangular resolution and sensitivity, as our data-handling capabilitiesgrow, and as increased computational capacity can be applied to theoreticalanalyses of lenses.

OCR for page 25
Cosmology: A Research Briefing Figure 7. Gravitational lensing of distant background galaxies by the potentialwell of a foreground cluster of galaxies. The distinctive arcs area result of strong lensing in which the images formed by the lensencircle the optical axis of the system. (Courtesy of Space TelescopeScience Institute.) Measuring cosmological parameters with gravitational lenses Because the light rays in a gravitational lens system propagate acrosslarge distances, their interpretation is subject to assumptions aboutthe cosmological model. Observations of gravitational lenses thereforeprovide a way to measure or constrain the cosmological parametersH0, q0, and Λ described above. The Hubble constant is most directly inferred througha measurement of the time delay in the arrival of signals from themultiple images of a strong gravitational lens. The success of thistechnique is predicated on a precise understanding of the distributionof the matter in the lens, which must be reconstructed purely fromthe properties of the images. The statistics of gravitational lensesprovide a powerful technique for the determination of cosmologicalparameters. For example, the frequency of occurrence of gravitationallenses depends strongly on the geometry of the universe. A high-redshiftquasar is twice as likely to be gravitationally lensed in a q0 = 0 universe as in a q0 = 0.5 universe. Similarly, a high-redshift quasar is about 15 timesmore likely to be strongly lensed in a flat universe if Λ = 1 thanif Λ = 0. Current limits appear to preclude models in which the Λ term dominates the universe; improved constraints will be possiblewith improved imaging using the newly completed Very Long BaselineArray (VLBA), the HST, and ground-based optical telescopes equippedwith adaptive optics systems. Gravitational lensing, along with the spectroscopic absorption studiesof intergalactic clouds, is one of the few techniques in cosmologyin which our ability to detect the presence of matter does not requirethat the matter be luminous. Lensing can address the important issuesof the nature and the distribution of dark matter on a wide rangeof scales. For example, the measurement of weak lensing, in whichfaint background galaxies would be expected to have correlated orientationsdue to the gravitational distortions induced by the foreground massdistribution,

OCR for page 25
Cosmology: A Research Briefing can be studied by careful analysis of high-quality images coveringlarge fields of view. Recent theoretical analysis describes how thismethod can provide a solid measurement of the matter-clustering amplitudeof the foreground mass distribution. Elongated images are more likelyto occur near large concentrations of matter, such as clusters ofgalaxies, where several spectacular examples have already been observed(see Figure 7). On small scales (stellar masses or smaller), the most effective techniqueis observation of microlensing, the variations in flux detected froma background source due to the focusing of rays caused by the passageof a massive object through the beam. By searching for microlensingevents in the direction of nearby concentrations of stars, threeseparate research groups have found spectacular examples of largeflux amplification (up to a factor of 14!). These events might implythat the dark matter comprising the halo of our galaxy is dominatedby compact stars that were never sufficiently massive to ignite theirnuclear furnaces and become luminous, or that there exist many moreremnants of normal stars in an extended disk of our galaxy than hadbeen anticipated. The time behavior of one well-documented microlensingevent in the direction of the nearby Large Magellanic Cloud is shownin Figure 8 . The study of gravitational lensing has already provided importantresults in cosmology. Improved instrumentation and data-handlingtechniques are allowing astrophysicists to recognize and exploitsubtle manifestations of gravitational lensing, as well as the strikingexamples of strong lensing. With the planned capabilities of futuregenerations of instruments, and the growing interest in gravitationallensing as an observational tool, it is likely that this trend willcontinue. Figure 8. The first well-documented microlensing event, detected when a faintobject of appoximately 0.1 solar mass crossed very close to the lineof sight between Earth and a star in the Large Magellanic Cloud (LMC).The gravitational lensing caused by the object has focused and intensifiedthe light detected from an LMC star in the background, causing thepeak (Ared) wavelengths recorded as a function of time. (Courtesy of CharlesAlcock for the MACHO Project.)

OCR for page 25
Cosmology: A Research Briefing V. PHYSICS OF THE EARLY UNIVERSE In contrast to the observational studies of the CMBR, galaxies, andlarge-scale structure, the field of the physics of the early universeinvolves concepts that are less familiar, more theoretical, and moredaunting. A little background, supplementary to the sidebars, ishelpful. As described in the sidebars (pp. 5 and 8), the universe is coolingand decreasing in density as it expands. Since temperature is simplya measure of mean energy per particle, the energy available for particleinteractions is also declining. If we imagine running the universe's clock backward toward zero, before the first 100 seconds, the CMBRwould be blindingly hot and energetic, and more and more energeticevents would become possible, including the creation of multitudesof elementary particles that are not stable at the present energydensity of the universe. An important concept of modern physics is the phase transition, theidea that the nature of the interactions of particles can changewith available energy. The freezing and boiling of water are familiarexamples of phase transitions. In the early universe, the natureof physical law itself is thought to have undergone a series of phasetransitions, with enormous consequences for the physics of that era.The history of the early universe and the laws of physics are intimatelyintertwined—each one can be illuminated by studying the other. Moreover,energies in the Big Bang reached up to 1014 times higher than any conceivable terrestrial accelerator, and theseenergies probe realms that are inaccessible to our laboratory experiments.For this reason the Big Bang is sometimes called the “poor man's particleaccelerator.” The aim of early-universe cosmology is to trace the successive transitionsof forces and particles from the earliest, fiercely hot moments ofthe Big Bang to the epoch of atom formation at 4,000 K (see section II on the cosmic microwave background radiation). Along the way, atcertain critical temperatures, precipitous changes in the state ofthe universe occurred. Some of these changes have observable consequences,called relics, that persist to this day and provide key tests ofmodels. Some of these relics—the ratio of photons to atoms and thenature of the large-scale fluctuations in the matter distributionas detected by the COBE satellite and studies of large-scale structure—are discussed above. Other fundamental questions, such as why theuniverse is homogeneous on the largest scales, and what is the natureof dark matter, also appear to have their answers in the physicsof the early universe. The sidebar “The Early Universe” (p. 8) treats the first 100 seconds after the Big Bang. During thisperiod, the universe was opaque to all forms of electromagnetic radiation,and so astronomers cannot make direct observations of the eventsat these early times. Nevertheless, important experimental and observationalinformation can be obtained that leads to reasonable physical inferencesabout events and conditions during this period. Progress in the studyof the early universe has been spectacular in recent years. The earliest state of the universe that can be addressed by physicaltheories is called the quantum gravity era (see sidebar, “The Cosmic Picture,” p. 5), because during that era the temperature and density of theuniverse were so high that gravity must be described by a quantumfield theory of some kind. The four forces of nature (gravity, theweak force of radioactive decay, electromagnetism, and the strongnuclear force) were probably completely unified during this era,and space and time could not be differentiated. After this almostunimaginably remote epoch, the universe cooled sufficiently for gravityto be described by Einstein's general relativity theory, but thetemperature was still sufficiently high that the other three forcesof nature remained unified (the grand unification era). Many theoristsbelieve the phase transition that marked the end of the grand unificationera was followed by a period of inflation. (Inflation is discussedbelow.)

OCR for page 25
Cosmology: A Research Briefing As time progressed and the universe cooled further, additional phasetransitions occurred, such as the end of the symmetry between weakinteractions and electromagnetic interactions and the transitionfrom free quarks to quarks bound into hadrons. (Protons and neutronsare the most familiar example of hadrons.) At very high temperatures quantum interactions create and destroyparticles, leading to an equilibrium in the number density of eachparticle type. However, particle interaction rates depend on temperature,and as the universe cooled, interactions occurred less frequently.When an interaction rate became small relative to the expansion rateof the universe, the equilibrating process effectively ended. Theabundances of the reacting particles were thereafter fixed. Althoughthe expansion rate of the universe slowed as it became cooler andless dense, particle interaction rates decreased even more rapidly. Although the abundance of atoms relative to photons is likely tobe a relic from an earlier, hotter epoch whose physics is not yetwell understood, the conditions that existed when the universe was1 second old can now be explored in the high-energy physics laboratory.Cosmologists are therefore more confident in their modeling of theconstituents and the physical processes at work. For example, bythe time the universe was 1 second old, it had cooled and expandedto the point where neutrinos, whose weak interaction with other matterdecreases as a function of temperature, effectively ceased interactingwith matter. This change acted to stabilize the ratio of neutronsto protons. With further cooling, positrons (antielectrons) annihilatedmost of the electrons, and neutrons quickly attached to the protons,forming all the deuterium (an atomic nucleus consisting of one protonand one neutron) and most of the helium now present in the universe. Thus, many interactions that were close to thermal equilibrium inthe early universe later froze out at a predictable epoch, and leftrelics, some of which survive to the present day (see Table 1). One example is the abundance of primordial helium and deuterium,discussed in more detail below. Primordial Nucleosynthesis and Dark Matter Along with the Hubble expansion and the cosmic microwave backgroundradiation, one of the pillars of the Big Bang theory is its successfulprediction of the abundances of the light elements deuterium, helium,and lithium. The Big Bang theory says that when the universe wasabout 1 second old and had a temperature of 1010 K, nuclear processes should have started that eventually yieldedcertain well-specified abundances for these light elements (see Table 1). The abundances of these light elements have all been found tobe in agreement with the predictions of Big Bang theory within theaccuracy of the measurements. Even the abundance of lithium relativeto hydrogen, predicted to be 1 part in 10 billion, matches the observations.Furthermore, the Big Bang theory predicts that the abundances willfit well only if there are no more than three families of neutrinos—a condition that was confirmed recently at the Large Electron-Positron(LEP) Collider in Geneva, Switzerland. Thus, the Big Bang theory's detailed predictions, even though they are based on the natureof the universe when it was only 1 second old, have been confirmedby observations and experiments. The light elements shown in Table 1, with abundances ranging from about 23 percent for helium to 1 partin 10l0 for lithium, all fit with the Big Bang theoretical predictions onone condition—that the one adjustable parameter of the theory, ΩB (the ratio of the mean density of ordinary matter (baryons) to thecritical density in the universe), has a value between 0.01 and 0.1.The nucleosynthesis calculations thus imply either that (1) the totaldensity of the universe is much less than the critical density andit will never stop expanding, or (2) the dominant component of theuniverse is not ordinary (baryonic) matter. The value of ΩB predicted by primordial nucleosynthesis can be compared to that

OCR for page 25
Cosmology: A Research Briefing Table 1. Observed Relic Values Relic When Measured Observed Values 1H (Hydrogen) 1960s ∼76% by mass 2H (Deuterium) 1970s >1.8 × 10−5 relative to hydrogen 3He (Helium-3) 1970s <6 × 10−5 relative to hydrogen 4He (Helium-4) 1960s ∼23±1% by mass 7Li (Lithium-7) 1980s 1.5±0.5 × 10−10 relative to hydrogen Number of neutrino families 1990 Nv = 2.99±0.02 observed in the universe. The mass density of luminous material instars and galaxies is small, Ωvisible < 0.0l, while the hot gas in galaxy clusters, which astronomers can detectby its x-ray emission, contributes perhaps Ωgas ~ 0.03. The sum of these values lies in the range consistent withprimordial nucleosynthesis. At the same time, dynamical models basedon the relative motions of galaxies, and the way spiral galaxiesrotate, argue that galaxies have more mass than is seen in theirdetected stars and gas. These dynamical arguments imply that eachgalaxy has an invisible halo of dark matter that is about 10 timesthe visible mass. Moreover, consideration of large-scale flows seemsto indicate a still larger amount of dark matter on scales much largerthan single galaxies. Perhaps there is even enough to be consistentwith a total matter density of Ω ~ 1, or at least Ω > ΩB. This implies that there must exist some unknown form of matterthat dominates the mass density of the universe—an awkward situationfor cosmologists. There are other theoretical reasons to expect that Ω = 1, or in other words that the total mass density is exactly equalto the critical value that just closes the universe. Ω is an unstablequantity in an expanding universe. If Ω is below 1 it will rapidlybecome much less than 1 as expansion proceeds. Conversely, if Ω isgreater than 1, it will grow to values much greater than 1. Onlyif Ω = 1 does it stay at 1; all other values diverge to either zeroor infinity. A finite, non-zero value of Ω today, other than Ω = 1, implies that it musthave been extremely close to 1 at the beginning of the universe.Cosmologists have puzzled over this fine-tuning problem for decades,but just in the past decade or so, considerations of the early universehave motivated a sensible resolution to this question—inflation. Epoch of Inflation and Grand Unified Theories of Matter After the Big Bang, the temperature of the early universe was sohigh that the four fundamental forces of nature are believed to havebeen merged. In the grand Unification era that followed, the grandunified theory (GUT) predicts that all the forces except gravitywere of equal strength. Modern theories of these forces involve aconcept known as symmetry breaking, in which the lowest-energy state(the vacuum) is not symmetric at the low temperatures of the presentuniverse. As time progressed, the temperature decreased, and thevacuum underwent a phase transition from a symmetric state of higherenergy. The higher energy of the “false vacuum” can in principle act like a non-zerocosmological constant, A ≠ 0, which, according to Einstein's general relativity theory, can drivean extremely rapid,