BACKGROUND RADIATION, SOFT X-RAY WILLIAM L. KRAUSHAAR "Soft x-rays" means x-radiation that is easily absorbed. Here we shall be discussing the x-ray energy region between 80 and 2000 eV. Adjacent at lower energies is the "extreme ultraviolet" and adjacent at higher energies are "hard x-rays". Abeam of 2000 eV x-rays is attenuated to half-intensity by an air path of about 1.5 cm, whereas an air path of only 0.01 cm is sufficient to attenuate to half-intensity a beam of 80 eV x-rays. The detection technology of soft x-rays is awkward because of this extreme susceptibility to absorption and, of course, all astronomical observations must be carried out well above the Earth's atmosphere. Given the x-ray energy, E(in electron volts), the wavelength, ^ (in angstroms), can be calculated from ^=12,400/E. Very early in the history of x-ray astronomy, it was noted that in addition to the radiation coming from discrete astronomical objects that are largely confined to the plane of our galaxy (the Milky Way), there was also a diffuse component that appeared to be isotropic. These early measurements were generally made at energies greater than 2000 eV where interstellar absorption (absorption by the gas between the stars of the Milky Way) is not significant. The isotropy of the detected radiation implied, therefore, that the diffuse background radiation above 2000 eV had an origin unrelated to the discrete objects in our galaxy and most likely originated in the vast regions beyond. This interpretation of the diffuse background radiation above 2000 eV has persisted over the years. When the first measurements of diffuse soft x-rays near 250 eV were made (here interstellar absorption is an important effect), there were pronounced deviations from isotropy. An appreciation of the possible significance of this anisotropy requires familiarity with a few of the details of the interstellar absorption process. SOFT X-RAY MEASUREMENTS All of the x-ray background measurements made during the decade since discovery have been made with detectors of modest-to-poor energy resolution. Survey data such as those used to prepare Figure 1 have been obtained with proportional counters. Thin filters in front of a counter window, sometimes fabricated as part of the window itself, have been used to limit the response to a fairly narrow band of x-ray energies. The soft x-ray band names and definitions are shown in the two left-hand columns of Table 1. Maps of the entire sky are available for all except the Be band. The most complete data set, of which the C band map shown in the figure is part, has a spatial resolution of about 6^. Be, B, and C refer to beryllium, boom, and carbon as filter materials. INTERSTELLAR X-RAY ABSORPTION The physical process by which soft x-rays are attenuated by the interstellar medium is photoelectric absorption, usually with the ejection of a K-shell electron. Hydrogen and helium are responsible for almost all of the absorption at energies less than 250 eV but at higher x-ray energies, carbon, nitrogen, oxygen, iron (L-shell electron ejection), neon, and magnesium together make the dominant contribution. Interstellar absorption has usually been estimated quantitatively from a measured knowledge of only the atomic hydrogen column density. This is because column densities of atomic hydrogen can be and have been measured in all directions, using the hydrogen hyperfine 21cm radio transition in emission. The column density of most other elements can only be measured in optical or ultraviolet absorption, using a bright star for the background source. X-ray absorption as inferred from 21cm column density measurements relies, therefore, on a knowledge of the abundances of the other elements, relative to hydrogen. Furthermore, 21cm-measured column densities extend through the whole Galaxy and do not terminate on a star. Interstellar absorption mean free paths, expressed as the column density of atomic hydrogen, are shown in Table 1. The mean free paths (in parsecs) assume an atomic hydrogen density of 1cm - 3. Recent evidence indicates that absorption by interstellar gas associated with ionized hydrogen may be comparable in some directions with that associated with atomic hydrogen. The detailed spatial distribution of ionized hydrogen is not known. It is not simply related to that of atomic hydrogen. The 21cm studies have shown that the galactic atomic hydrogen is distributed in a flat disk some 200 pc or so thick and 15,000 pc in radius. The solar system is located about midway through the disk and 8000 pc from the galactic center. Lines of sight near the perpendicular to the galactic plane (galactic latitude, [b], of about 90ø or toward a galactic pole) traverse generally a minimum column density of Nm=10x20 H atoms cm-2, although there are a few special directions with only half that amount. The interstellar gas distribution is quite irregular so that only in very rough approximations does the column density conform to the expected value, Nm/sin/b/. In the galactic plane the column density, Sun to outside the Galaxy, is of order 10x22 H atoms cm -2, sufficient to absorb completely in all but the I and J bands. THE SOFT X-RAY SKY Among the objects evident in all-sky soft x-ray surveys are stars of many types, accreting binary systems and supernova remnants of various ages. In preparing maps for background studies, these more or less discrete objects are usually removed, if possible, to leave maps of the background alone. The map shown in Figure 1 is for the C band and is in galactic coordinates. A prominent feature is the tongue of enhanced intensity that dips down from near the north galactic pole along galactic longitudes near 15ø. It appears also on the B, M1, M2, and I band maps with a tendency for the intensity peak to move toward the galactic plane at higher x-ray energies. An apparently related but not quite coincident feature in the nonthermal radio sky is known as the North Polar Spur. The synchrotron radio emission is thought to originate in a shell at about 100 pc. It is likely that the x-rays are emitted from a region of hot gas interior to the radio-emitting shell. The connected tongue that runs along longitude 300^ also has a radio noise counterpart and the total feature is known as loop I. In those parts of the sky not dominated by loop I, there is, in the C and B bands, a marked tendency for the intensity to be larger at high galactic latitudes. The intensity near the galactic poles is two or three times as great as it is near the galactic plane. The intensity has a negative correlation with the column density of hydrogen, though with consider- able scatter. The finite intensity in the galactic plane is clear evidence for nearby emission. Whereas the Loop I x-ray feature remains prominent on the M1, M2, and I band all-sky soft x-ray maps, the intensity enhancement at high galactic latitudes is lacking. Instead, there is a large irregular high-intensity feature about 40ø in extent near the galactic center. Although there are a few other enhanced regions, possibly relic supernova remnants, the overall picture is one of increasing isotropy as the x-ray energy increases from the M1, to M2, to I, to J band. DISTANCE TO THE DIFFUSE EMISSION SOURCES The B and C band observations might easily be taken to suggest a strong extragalactic or galactic halo source subject to interstellar absorption (to explain the negative correlation of intensity with hydrogen column density) superimposed on a local diffuse source (to explain the intensity in the galactic plane). This seemingly attractive suggestion has serious problems, however, outlined as follows. 1. Shadowing of soft x-rays by the absorbing interstellar gas of external galaxies should be observable, has been searched for, but has not been seen. 2. Intensity fall-off with increasing amounts of galactic gas is observed but only in an overall average sense. Many individual features in the gas distribution, where deep absorption is predicted, show no decrease whatever in x-ray intensity. 3. The steepness of the intensity fall-off with gas column density should vary inversely with the absorption mean free path. Yet the C and B band intensities (and the Be band intensity also, where data are available) appear to fall off as though their mean free paths were all the same and about twice the predicted value for the C band. Extreme clumping of the gas of the interstellar medium could result in these anomalous mean free paths, but the required clumping is not seen. Extragalactic intensities or even upper limits are often of cosmological interest. A power law spectrum 11****** photons (cm2 s sr keV)-1 (E is the photon energy in kiloelectron volts) is a good representation of the extragalactic intensity from 2-6 keV. For the C band the intensity upper limit for x-rays arriving from beyond the absorbing galactic gas is about three times that predicted from the extrapolated extragalactic power law. This upper limit does not include a correction factor estimated to be between 2 and 4 to allow for absorption due to gas associated with ionized hydrogen. (The upper limit measurement was carried out in directions near to but not coincident with that of the Small Magellanic Cloud.) The corresponding intensity upper limit for the M2 band is about twice that predicted from the extrapolated extragalactic power law. No correction factor is appropriate here. As far as the galactic emission is concerned, shadowing by interstellar clouds a known distance away would serve to give direct indicators of distance to the soft x-ray emitting gas. This shadowing has not been observed and the best distance indicator comes from an indirect argument, as follows. If there were no absorbing material between us and the emission source, the C and B band maps would reflect just variations in the source strength and so, except for minor effects of source energy spectra, the intensities would track and the maps would look much the same. For small but variable (in direction) amounts of intervening absorber, say 10x20 H atoms cm-2 or so, the amount of absorption and its variation would be very different in the B and C bands. Features, if due to absorption, should be twice as deep in the B band as in the C band. This is not observed and so we conclude that the bulk of the galactic emission must be separated from us by less than 10x20 H atoms cm-2 With an atomic hydrogen density of 1 cm-3, this would correspond to a distance of less than 30 pc. Note, however, that the solar system is in a region of density appreciably less than the galactic plane average of 1 cm-3. EMISSION MECHANISMS The energy spectrum of the isotropic extragalactic radiation between 2 and 6 keV is well represented by a power law. The nature of the source and its emission mechanism is uncertain. If extrapolated into the soft x-ray region, this power law spectrum would provide a significant fraction of the measured intensity in the J, I, M2, and M1 bands. Its extrapolated contribution to the C and B bands would be negligible except in those few special directions where the 21cm H data would predict extraordinarily small absorption of an extragalactic source. M-dwarf stars, which have known x-ray emission properties and a very large space density, may contribute an apparent diffuse intensity of as much as 25% of that measured in the M bands. The contribution predicted in the lower energy bands is only a few percent. For the remaining emission source, the available evidence favors radiation from very hot regions of interstellar gas. Ultraviolet measurements have shown that O VI, the oxygen ionization state favored at 0.3 million degrees, exists along the lines of sight to many stars. Furthermore, in the M1 band evidence for emission lines of O VII, uniquely characteristic of gas at 2 million degrees kelvin, have been observed from the North Polar Spur and from the general background radiation. Emission lines have not yet been observed in the other soft x-ray bands. Additional Reading Bochkarev, N.G.(1987) The structure of the local interstellar medium, and the origin of the soft x-ray background. Soviet Astron. 31 (No. 1) 20. Cox, d.p. and Reynolds, R.J.(1987). The local interstellar medium. Ann. Rev. Astron. Ap. 25 303. McCammon, D., Burrows, D.N., Sanders, W.T., and Kraushaar, W.L.(1983). The soft x-ray background. Ap. J. 269 107. McCammon, D., and Sanders, W.T.(1990). The soft x-ray background and its origins. Ann. Rev. Astron. Ap. 28 657. Tanaka, Y. and Bleeker,J.A.M.(1977). The diffuse soft x-ray sky. Space Science Rev.2O 815. See also Background Radiation, Gamma Ray; Background Radiation, X-Ray; Interstellar Medium, Hot Phase.