Carnegie Observatories Astrophysics Series, Vol. 2:Measuring and Modeling the Universe, 2004ed. W. L. Freedman (Cambridge: Cambridge Univ. Press)Dark Matter TheoryJOSEPH SILKDepartment of Physics, University of OxfordAbstractI evaluate the dark matter budget and describe baryonic dark matter candidates and theirdetectability. Dark matter issues in galaxy formation theory are discussed, and I review theprospects for detecting nonbaryonic dark matter. Indirect detection via halo annihilationsof the favored dark matter candidate, the SUSY LSP, provides a potential signal. The relicdensity of dark matter particles specifies the annihilation cross-section within model uncer-tainties, and indirect detection provides our optimal strategy for confirming the dark mattercandidate. Galaxy formation simulations suggest that the predicted clumpiness of the darkhalo will facilitate our imaging the dark matter in gamma rays, with cosmic ray signaturesproviding invaluable confirmation. Similarly, the supermassive black hole in the center ofthe Milky Way should present a unique signal amplifier by which we can view the dark mat-ter in neutrinos as well as in gamma rays. The astrophysical uncertainties are so large thatone has no alternative but to look.1.1 IntroductionThere are two types of dark matter: baryonic and nonbaryonic. Several candidatesfor the former exist, but the mass fraction is unknown. Conversely, while we have not yetdetected the leading nonbaryonic matter (cold dark matter; CDM) candidate, the neutralino,we can calculate its mass fraction and interaction cross-section, within model uncertainties.Dark matter dominates the Universe, amounting to90% of the matter density. Bary-onic dark matter, as yet unambiguously identified, comprises up to a third of the baryons,although there are compelling candidates. Elucidating the nature of all of the dark matter isone of the outstanding problems in astrophysics to be addressed over the next decade. Aswill be discussed in this review, research at the interface of dark matter with galaxy for-mation has been particularly active in recent years, but has also raised challenges that areleading some to question the entire dark matter edifice.Dark matter has a venerable history. In the solar system, anomalies in the orbit of Uranuspointed to dark matter in the form of a new planet, and this led to the discovery of the planetNeptune, following the predictions of Adams and Leverrier. The advance of the perihelion ofMercury’s orbit also stimulated searches for dark matter in the form of a new planet interiorto the orbit of Mercury, dubbed Vulcan. This turned out to be a red herring: the orbitalanomalies eventually led Einstein to propose a new theory of gravitation, general relativity.Similarly, there are parallels that may be drawn today. The modified Newtonian dynamics1J. Silk(MOND) theory seeks to modify the law of gravity in order to dispense with the need fordark matter. There is little in the way of compelling theory or data to support such a position,but, at the same time, MOND is tenaciously difficult to kill (e.g., Sanders & McGaugh 2002).It is certainly worth bearing in mind that general relativity has been thoroughly tested onlyin the weak-field limit, and on relatively small scales, now extended up to a Mpc or so bygravitational lensing studies.The modern dark matter problem was first described in 1933 by Fritz Zwicky (Zwicky1933), who noted that in the Coma cluster of galaxies, the ratio of mass to light as measuredby the virial theorem is about 400 M¬L¬. Since individual galaxies are found to haveabout 10 M¬L¬, there is a serious shortfall of luminous matter. When observations ofclusters were extended to X-ray frequencies, a significant component of mass was found thatwas not detected optically. X-ray observations have revealed the presence of a substantialamount of hot intracluster diffuse gas, contributing about 15% of the cluster mass. However,80% of the cluster mass is not accounted for in any known form. Moreover, this unknownmass component must be nonbaryonic, as the primordial nucleosynthesis measure of globalbaryon abundance, combined with the mean matter density, is consistent with the observedcluster baryon content in gas and stars.On larger scales, one can probe dark matter via both Hubble-flow perturbations studiedvia deep redshift surveys and shear maps from weak gravitational lensing. The dark mattercontent dominates the total matter content. The general consensus is thatm03withabout 10% of this being in baryons. Most of the baryons must also be dark, although thereare persuasive arguments about their nature.I now describe in more detail the dark matter budget, describe baryonic dark matter can-didates and their detectability, and then turn to CDM issues in galaxy formation theory. Iconclude with a review of the prospects for detecting nonbaryonic dark matter.1.2 Global Baryon InventoryPrimordial nucleosynthesis of the light elements demonstrates that the baryon frac-tion isb=0040004In effect, this is measured at z109when the light-elementabundances freeze out. There are two other independent measures of the total baryon con-tent of the Universe. The heights of the cosmic microwave background (CMB) acousticpeaks, in particular the odd peaks that correspond to wave compressions and rarefactionson the last-scattering surface scale and at half of this scale, are controlled primarily by thebaryon density at z1000The Ly forest indirectly measures the baryon density at z3,once a correction is made for the predominant ionized fraction, whose density is inferredfrom ionization balance by invoking the ionizing radiation field from quasars. All threemeasures of the baryon density converge onb004At the present epoch, the baryon inventory is rather different, however. Stars accountfor a modest fraction of all the baryons present, about 0.0026 in spheroid stars and 0.0015in disk stars and cold gas, in units of the Einstein-de Sitter density (Fukugita, Hogan, &Peebles 1998). These numbers are in approximate agreement with a recent determination,including all cold interstellar matter and stars in low-mass as well as in massive galaxies inthe local Universe, which yields 000240001, corresponding to 8%4% ofb(Bell etal. 2003). Intracluster gas in rich clusters, mapped in X-rays, accounts for another 0.0026of the closure density, consistent with the