The structures of universe reflect those of gravitational potential, which may be determined mainly by the distribution of dark matter. The hot gas in a galaxy tends to gather toward the bottom of the potential well of the galaxy, and has kinetic energy corresponding to the potential depth, and often emits X-ray corresponding to its kinetic energy. Thus X-ray observations of the hot gas are important as a tracer of dark matter. In fact, a cluster of galaxies and an elliptical galaxy were found to be powerful X-ray emitters by the Einstein and ASCA, but the presence of the hot gas in spiral galaxies was not clear until recently.
X-ray emission of spiral galaxies contains several emission components of different origins; e.g. individual X-ray sources (X-ray binaries and supernova remnants), active nuclei, and diffuse emissions related to starburst activity. Thus in order to study the origins of X-ray emissions, it is essential to resolve these components, which can be achieved by combining the spectral and spatial information. By utilizing the high spectral capabilities of ASCA, i.e. a wide energy coverage (0.5 - 10 keV) and a good energy resolution (~120 eV FWHM at 6keV), and moderate spatial resolution (~1 arc minute), we developed a new method of spatially resolved spectroscopy, which enabled us to uniquely distinguish the individual emission component in nearby spiral galaxies, M 51, M 82, M 83, M 106, NGC 253, NGC 1365, and NCG 4631. When we observe these galaxies by using ASCA, we can get the spatial information of a few kpc scale resolutions as well as energy spectrum, and can distinguish the emission components.
We analyzed M 83 to establish the method of space and spectrum combined analysis and estimate the statistical and systematic errors of this method. We took three steps to study the X-ray emission in M 83. The first was to make the total energy spectra of the emission excluding the possible contamination from the off-center bright individual X-ray sources. The second step was to quantify the spatial extent of the emission as a function of X-ray energy. For this purpose, we introduced a new measure, root mean square (rms) radius of the X-ray image, which enables us to remove the effect off the finite angular resolution of the telescope and quantitatively measure the size of X-ray emission region as a function of X-ray energy. The third step was to construct an X-ray emission model which can satisfy both the energy spectrum and the energy dependence of rms radius of the image simultaneously. From this analysis, we discovered that the X-ray emission of M 83 was made of two components and estimated the spatial extent of soft and hard components to be less than 2 arc minutes and about 3 arc minutes, respectively. And we applied this method to remaining six galaxies and found that all seven galaxies contain an extended X-ray emission centered at a position consistent with the nucleus of the galaxy. The combined analysis of these seven galaxies showed that two or three emission components of different energy spectra and different spatial extent (i.e. rms radius) were necessary to consistently explain both the energy spectrum and the energy dependence of the rms radius.
We have divided the emission components observed from the seven selected galaxies into three groups; components I, II, and III(Table 5.16). Each component has different spatial size and energy spectrum. Component III, which has a hardest energy spectrum and a small spatial extent can be consistently interpreted by active galactic nucleus. The nature of the three Seyfert galaxies, M 51, NGC 1365, and M 106, support this interpretation. Component II is further subdivided into two sub groups. Because of the spectral hardness(kT ~ 3 keV when approximated with thermal bremsstrahlung emission) and spatial extent (rms radius ~ 4 kpc), one sub group of emission is considered to be a sum of unresolved binary sources. The ratio of the X-ray to the blue band luminosities is similar to that of M31, which supports this interpretaion. The other sub group of emission contains emission lines of atoms and observed only from galaxies with strong starburst activities. Thus it is considered to be related to this physics.
On the other hand, we can not explain the luminosity and spectrum of component I with previously known classes of X-ray sources in spiral galaxies. The component I, which is soft (kT = 0.1 - 0.9 keV) and extended (rms radius ~ 2kpc), was detected from all seven spiral galaxies. We consider that it is diffuse in nature and originates from a "diffuse core" which fills the potential well of the galaxy. The temperature of the emission nicely coincides with the depth of the gravitational potential of spiral galaxies. Both the temperature and the luminosity are positively correlated with the blue band luminosity of the galaxy. This supports our interpretation. We estimate that the mass and the thermal energy of the "diffuse core" can be supported by supernova(SN) explosions if the SN rate is 0.01 yr-1 per galaxy or higher. We conclude that observed seven spiral galaxies contain extended hot gas which is in balance with the gravitational potential of galaxies. This is the most important results in this thesis.