COSMIC RAYS, ORIGIN RICHARD E. LINGENFELTER The question of the origin of cosmic rays has been the focal point of cosmic ray studies since their discovery nearly a century ago, and a wide variety of sources, ranging from solar to cosmological, have been suggested over the years. Although the question is still not fully answered, the observational constraints now at least allow us to strongly restrict the possibilities. In this entry we will concentrate on the question of the sites and energetics of cosmic ray sources, leaving the details of their acceleration and propagation to other entries. The existence of cosmic radiation was first suggested in 1900 by Charles T.R. Wilson, following his discovery of atmospheric ionization, when he proposed that extraterrestrial gamma rays might be responsible for the ionization. Twelve years later, with balloon-borne ionization detectors showing that the ionization rate increased with altitude and had no diurnal variation, Victor F.Hess demonstrated the extraterrestrial and extrasolar origin of the ionizing radiation, which soon came to be known as "cosmic rays." These rays were generally thought to be gamma rays until the variation of their flux with geomagnetic latitude was discovered in 1927. This clearly established that they were not gamma rays, but charged particles deflected by the magnetic field. Subsequent observations from balloons and satellites have shown that the cosmic rays are mostly relativistic protons with a significant fraction of heavier nuclei and electrons. The energy spectrum and anisotropy of the cosmic rays have now been measured over more than 10 decades in energy all the way up to **** eV. Studies of the observed cosmic ray energy spectrum, anisotropy, and composition, together with related gamma ray and radio observations, now suggest that the bulk of the cosmic rays observed below **** eV are of galactic origin, while those at higher energies are primarily of nearby extragalactic origin. These suggestions, nonetheless, are still debatable. GALCTIC COSMIC RAYS The observational constraints on the cosmic rays with energies ****** eV strongly suggest a galactic origin for the bulk of the cosmic rays and a self-consistent argument can be made. The simplest argument for a galactic origin of at least some of the cosmic rays comes from the cosmic ray electrons, which have been observed to have a power-law energy spectrum of ********** extending up to energies of ***** eV. Because such relativistic electrons suffer energy losses by Compton scattering on the ****** microwave background radiation, the highest-energy electrons have a lifetime against energy loss of only *******. Thus they cannot have traveled more than *******, even if they traveled in straight lines. Since the overall anisotropy of the cosmic rays at hat energy is *********, their source is probably no more than ******* away, because the anisotropy is essentially the ratio of th rectilinear distance divided by the total distance traveled. Thus the cosmic ray electrons clearly must be of galactic origin, and that could suggest a similar origin for protons of similar energies. Similar estimates of the volume of space in which the bulk of the cosmic rays are contained and of their lifetime within that containment volume can be made from a number of different observations. A containment volume of at least galactic disk dimensions with roughly the local energy density is required to account for the observed diffuse galactic flux of high energy (*******) gamma rays, if they result primarily from the decay of pions produced by nuclear interactions of the cosmic rays with the interstellar gas. A minimum containment volume several times that of the galactic disk is required by measurements of cosmic ray nuclei of secondary origin, produced by spallation of the primary cosmic rays in nuclear interactions with ambient matter. The cosmic ray abundances of secondaries **, ***, **, **, and B all require that the bulk of the comic rays must have gone through a mean column depth ****** of****** ****. Moreover the cosmic ray abundances of radioactive secondaries, *****, ***, and ****, with half lives of ********, **** ***, and ***** *** yr, further require that these interactions must have taken place over a mean time *********** yr. Thus the mean density of the ambient matter in which the cosmic rays are contained and interact is ********* ***********. This density is significantly less than the ****** *********** local mean density of gas in the disk, implying a minimum cosmic ray containment volume *** or more times that of the galactic disk. Such a size is also consistent with the measured anisotropy ********* for the bulk of the cosmic rays (Fig. 1). Because the anisotropy is essentially ****/**, the ratio of the rectilinear distance * divided by the total distance traveled **, then the mean distance *****, which for the above values gives a distance *****, or several times the scale height of the disk gas. This very small anisotropy of the bulk of the cosmic rays can be understood in terms of diffusive propagation resulting from spattering of the charged cosmic rays by irregularities in the galactic magnetic field. In the simplest diffusion treatment, the mean size of the irregularities is **************, which, for the above values of *, *, and *, gives *******. This size is typical of interstellar distances in the disk. Diffusive scattering should be effective as long the cosmic ray gyroradius is less than the mean *********. Such a gyroradius, ***************, where *** is the energy in units of **** eV, corresponds to a cosmic ray proton energy of ********** eV in a ***** galactic magnetic field, and could thus account for the observed breaks in both the cosmic ray energy spectrum (Fig. 2) and anisotropy (Fig.1) at ***********eV. As the energy increases, the diffusive scattering becomes less efficient, the anisotropy of the cosmic rays increases, and their containment time decreases inversely with anisotropy ******** *********. Thus cosmic rays produced with a source energy spectrum **** would be observed within the containment volume with an equilibrium energy spectrum ***************. The observed energy spectrum and anisotropy are, in fact, quite consistent with such a relationship for a single power-law source spectrum of ******* all the way from **** to **** eV. This observed correlation between the energy spectrum and anisotropy also provides the only direct evidence for a galactic origin of the bulk of the cosmic rays. For with the exception of the indirect evidence of the cosmic ray electrons, none of the other observations places any upper bound on the size of the containment volume that would require a galactic origin. The limits on the isotropic background of high energy (*******) gamma rays are not inconsistent with cosmic rays of the local energy density filling the entire Universe and interacting with the intergalactic gas to produce pion decay gamma rays, as long as the density of the gas was less than ******* of that required to close the Universe. Although only a negligible amount of matter could be traversed in such densities by the extragalactic cosmic rays even in the Hubble time, the bulk of the ******** of matter that locally observed cosmic rays must have passed through could still have been traversed in the galactic disk, just as if they originated in the disk. The inverse correlation between the observed energy spectrum and anisotropy, however, is expected only within a containment volume from which cosmic rays are escape and not for cosmic rays all of space. The maximum size of the containment volume for the bulk of the cosmic rays is limited by the minimum anisotropy of ***********************, since the cosmic ray containment time can not exceed the Hubble time c/Ho. Thus the bulk of the observed cosmic rays must be contained in a volume no larger than that of the Local Group of galaxies, in which our Galaxy makes up **** of the observable mass. One further argument for an even more restricted galactic disk origin of the bulk of the cosmic rays is the remarkable similarity in magnitude of the cosmic ray energy density and the magnetic field energy density in the galactic disk, both of which are of the order of *************. This would be expected if the cosmic rays are produced within the galactic disk and are contained by the galactic magnetic field. For if the energy density, or pressure, of the cosmic rays exceeded that of the magnetic field they could not be contained by it. Such a similarity would not be expected if the cosmic rays originated outside of the galactic disk. Furthermore, the turbulent energy density of the interstellar gas in the galactic disk is also of the same order of magnitude as that of the cosmic rays and the magnetic field. This suggests a coupling and possible equipartition of energy between the three, supporting suggestions that the cosmic rays could be accelerated by stochastic processes in the galactic disk, as will be discussed further. There is, however, one further energy density that is also of the same order of magnitude, that of the 2.7-K microwave background radiation, and this puzzling similarity has not yet found any ready explanation with either a galactic or an extragalactic origin of the cosmic rays. Independent of the size of the galactic cosmic ray containment volume or the cosmic ray lifetime within it, however, a straightforward determination can be made of the combined cosmic ray luminosity of all galactic sources required to maintain the measured, local cosmic ray energy density of ***************. The determination of this power depends only on the measured energy density and average amount of matter traversed by the cosmic rays, and the total mass of interstellar matter in which the cosmic rays can interact. In particular, the galactic cosmic ray luminosity is ********, the total cosmic ray energy in whatever galactic volume V the local energy density w fills, divided by the mean cosmic ray containment time * within that volume. However, the mean time * is simply *******, where x is the average amount of matter per unit area that the cosmic rays have traversed, and the average density of that matter seen by the cosmic rays, p, is equal to **/*, the total mass of gas Mg in the containment volume V. Thus, the galactic cosmic ray luminosity is simply **********. Taking a mean amount of matter ************ traversed by cosmic rays, determined from the relative abundances of secondaries produced by nuclear spallation, and a total mass of interstellar gas ********** *********,equal to ****** of the mass of the Galaxy, then ************ **. This is the total rate of cosmic ray production in the Galaxy, required to maintain the local cosmic ray energy density ****** erg ***, independent of both the containment volume and time. There are a variety of galactic sites that have been suggested as the source of the cosmic rays, but they most likely all derive their energy in one way or another from supernovae. Galactic supernovae are estimated to occur on the average about once every *30 yr and release some ****** erg just in the kinetic energy of their expanding ejecta, corresponding to a time-averaged energy release of ****** erg ** This is an order of magnitude greater than that required in cosmic rays, and thus various models have been suggested for cosmic ray acceleration in the initial explosion of the supernova, in supernova shock waves as they subsequently expand into the interstellar medium, and later yet in stochastic scattering by turbulent irregularities resulting from the dissipation of the supernova energy in the interstellar medium. Supernovae may also leave behind a comparable amount of rotational energy in rapidly spinning (***** period) magnetic neutron stars which could generate an enormous electric field and directly accelerate cosmic rays. Subsequent accretion of gas onto such compact objects can also lead to the acceleration of particles to energies of at least **** eV, as is evident from observations of Cygnus X-3, Hercules X-1, and other ultrahigh energy sources. EXTRAGALCTIC COSMIC RAYS The arguments for an extragalactic origin of the highest-energy cosmic rays are simpler and more straightforward. Unlike the cosmic rays at much lower energies, which are nearly isotropic, those at the highest energies are extremely anisotropic. As can be seen in Fig.1, the measured anisotropy approaches unity at energies ***** eV. Although there is no clear data on the composition of these cosmic rays, at such energies the gyroradius of protons in the galactic magnetic field of a few microgauss becomes comparable to the dimensions of the Galaxy. Thus, such cosmic ray protons should suffer little deflection in the magnetic field and travel in nearly straight lines, so that their arrival directions should point close to the direction of their origin. The energy-weighted mean direction of the highest-energy cosmic rays is not in the direction of the inner part of the Galaxy, or even close to the galactic disk, as might be expected if these cosmic rays came from a galactic source. Instead, these highest energy cosmic rays come from a mean direction within **** of the North Galactic Pole. This is close to the direction of the Virgo supercluster of galaxies, the center of which lies at a distance of *******, and the direction is even closer to that of the mean of the galaxies weighted by their mass divided by their distances squared. Estimates of the magnitudes of intergalactic magnetic fields suggest that these cosmic rays would also not be significantly deflected over such distances. Thus their arrival directions seem to point to a nearby extragalactic origin. Photopion production by interactions of these cosmic rays with the ***** microwave background radiation further constrains their possible sources. In the rest frame of these ultrarelativistic particles the microwave photons are blueshifted by the particle lorentz factor to energies of the order of 200 MeV, sufficient to produce pions which can carry away a significant fraction of the particle energy. Thus the maximum distance that the highest-energy cosmic rays could have traveled through the microwave radiation without losing most of their energy is only about 30 Mpc. An estimate of the cosmic ray An estimate of the cosmic ray luminosity required for such an extragalactic source can be made from their energy spectrum. For anisotropy is not the only difference between the highest-energy cosmic rays and those at energies ******; their energy spectrum is also quite different, supporting the possibility that they might have a separate origin. As can be seen in Fig. 2, for several decades below this energy the spectrum roughly follows a power law in energy, ** with an index **********, but above ***** eV the spectrum abruptly flattens to a shape that can be crudely approximated by a power-law index *********. If such a power-law spectrum extended on down to much lower energies ( *******) typical of the bulk of the cosmic rays, they would make up only ****** of the flux at those energies, and have a local energy density ******* erg cm**. If these cosmic rays fill a volume of radius ***** Mpc, comparable to the distance of the Virgo supercluster, with a containment time ********* yr between the light-travel time and the Hubble time, then the cosmic ray luminosity of the supercluster would have to be **** **************** erg s**. The lower luminosity, divided among the ***** *** galaxies in the cluster, would require an average galactic luminosity of ****** erg ***, only *10% of that required for our Galaxy, if it is the source of bulk of the cosmic rays, as was discussed above. The mean value is comparable to the bolometric luminosity of the Seyfert galaxies, NGC4151 and NGC1068, which lie in the cluster. The upper bound would require local quasistellar objects. Additional Reading Cesarsky, C.J.(1980). Cosmic ray confinement in the Galaxy, Ann. Rev. Astron. Ap. 18 289. Hillas, A.M.(1984). The origin of ultra-high-energy cosmic rays. Ann. Rev. Astron. Ap. 22 425. Setti, G., Spada, G.; and Wolfendale, A.W., ed.(1981). Origin of Cosmic Rays, IAU Symposium No.94. Reidel, Dordrecht. Shapiro, M.M., ed.(1986). Cosmic Radiation in Contemporary Astrophysics, Reidel, Dordrecht. Shapiro, M.M. and Wefel, J.P., ed.(1988). Genesis and Propagation of Cosmic Rays. Reidel, Dordrecht. Simpson, J.A.(1983). Elemental and isotopic composition of the galactic cosmic rays. Ann. Rev. Nucl. Part. Sci. 33 323. See also Background Radiation, Microwave; Cosmic Rays, Acceleration; Cosmic Rays, Propagation; Magnetohydrodynamics, Astrophysical; Supernova Remnants, Evolution and Interaction with the interstellar Medium.