Dwarf nova SS Cygni
The dwarf nova SS Cygni was discovered in 1896, the second following U Gem in 1855. As shown in Fig. 2a, explosions occur at an interval of 50 days or so and brightness in visible light increases from 12.4 magnitudes at the quiescent phase to a maximum of 7.7 magnitudes. It is believed that the drastic change in brightness is caused by the thermal instability of the disk. At the outer edge of the disk, only two stable states are possible: one where the temperature is lower than several thousands kelvin with neutral hydrogen; and another where the temperature is higher than 10,000 kelvin with ionized hydrogen. In the former state, matter from the companion star accumulates on the disk’s outer edge because the viscosity of the disk is low and, accordingly, the matter’s angular momentum is not fully released. At this time, the inner area of the disk becomes dark because of insufficient mass accretion. In the latter case, the disk’s outer-edge viscosity is high so that the accumulated mass accretes at once towards the white dwarf through the disk, causing the disk to shine brightly. The former is the quiescence while the latter is the outburst. In the case of dwarf novae, the rate of mass accretion from companion stars is in between the two stable states and therefore the dark and bright states alternate.
Ultraviolet and X-ray wavelengths, shorter than visible light, are emitted from deeper within the accretion disk. It is assumed that X-rays in particular are emitted from the “boundary layer” that forms between the innermost edge of the disk and the white-dwarf surface. White dwarfs generally rotate at a far lower speed than the Kepler velocity at their surface. In order for matter in the disk rotating at Kepler speed to fall onto the white dwarf, rotation energy corresponding to the velocity difference against the white-dwarf surface must be de-accelerated by friction and be transformed to thermal energy. Then the energy must be released in the form of radiation. The rapid de-acceleration and heatup take place at the boundary layer. As shown in Fig. 2b, the accretion disk that is about 100,000 kelvin at the entrance of the boundary layer at quiescence with lower matter mass-accretion ratio is heated in the boundary layer to over 100 million kelvin and inflates. At that moment, the disk emits X-rays. Meanwhile at outburst, when the mass-accretion ratio is high, since both disk density and cooling efficiency are high, the disk is kept geometrically thin even if it enters the boundary layer. Furthermore, even on the surface of the white dwarf, its temperature is still kept low at 100,000 kelvin or so. It has been thought, therefore, that X-ray radiation does not take place.
Fig. 3 shows the X-ray spectra of the SS Cygni at quiescence and outburst as measured by the SUZAKU satellite. Looking at the enlarged view around the iron emission line, the characteristic X-ray of 6.4keV from the neutral iron is observed in addition to the characteristic X-rays of 6.7keV and 7.0keV from helium-like and hydrogen-like ionized irons. The 6.4keV line is emitted when the iron atom, which is exposed to continuous X-ray continuum in the environment of lower than 1 million kelvin, loses one innermost core electron and, in place of that, one electron at higher-level orbit falls to fill the vacancy. Examining the figure carefully, it can be found that a broad width component is included in the 6.4keV bright line in addition to a narrow one. The narrow one is where the X-rays from the boundary layer illuminate the iron on the white-dwarf surface and the broad one is where they illuminate the iron in the accretion disk. The reason for the spread of energy width is thought to be the Doppler Effect caused by the disk’s Kepler rotation.
From the narrow width 6.4keV bright line, we can estimate the spatial spread of the boundary-layer plasma. The higher the boundary layer plasma spreads over the white dwarf, the smaller the angle viewing the white dwarf (acting as a reflector) is. As a result, reflection efficiency decreases and the bright-line intensity weakens. Using this principle, we estimated the spread of the boundary layer. As a result, it was first revealed that the spread is at most 15% of the radius of the white dwarf. From the broader component, we can estimate to what height the boundary layer distributes from the orbital plane. We are now analyzing this matter in detail.
In the spectrum at outburst in visible light, a low-temperature emission component from the accretion disk is clearly observed in the energy level of 0.5keV or less. This component is not found in the spectrum at quiescence so it must be true that the disk falls closer to the white dwarf only at outburst. Hard X-rays, however, which should not be emitted at outburst, are also clearly observed by SUZAKU, although weaker than that at quiescence. It is not clear from where the hard X-rays at outburst are emitted. However, we obtained a clue to answer the question by SUZAKU’s observation of the iron bright line. As shown in Fig. 3, the broader component dominates in the 6.4keV iron emission line. This implies that most reflection from the boundary layer comes from the accretion disk. We are now proceeding with the quantitative simulation to study the origin of this broader 6.4keV emission line and the spatial spread of the boundary layer at outburst. As discussed above, X-ray spectroscopic observations by SUZAKU are now contributing to the elucidation of the boundary layer that has been until now incompletely understood.