Optically Thick Winds from Degenerate Dwarfs. I. Classical Nova of Populations I and II. Mariko Kato

Transcription

1 Optically Thick Winds from Degenerate Dwarfs. I. Classical Nova of Populations I and II Mariko Kato Department of Astronomy, Keio University, Kouhoku-ku, Yokohama, 223 Japan mariko@educ.cc.keio.ac.jp ABSTRACT Twenty six sequences of optically thick wind solutions have calculated which mimic time-dependent evolution of classical novae of populations I and II. The peak of the new opacity around log T = 5:2 due to iron lines is found to be strong enough to accelerate the winds even in very low iron abundance such as Z=0.001 for massive white dwarfs ( 0:8M ). The old population novae show the slow light curve, the long X-ray turn-o time, the small expansion velocity and the small wind mass-loss rate. The X-ray turn-o time is a good indicator of the white dwarf mass because of its strong dependence on the white dwarf mass and weak dependence on the populations. The white dwarf mass is estimated to be 0:6 M for GQ Mus and 1:0M for V1974 Cyg. The systematic dierence of the wind velocity is predicted between novae in globular clusters and in galactic disk. Twenty six tables are presented in the computer readable form of CD-ROM that consists of solutions of the optically thick wind and the static for the decay phase of classical novae with composition of X=0.35, C=0.1 and O=0.2 and heavy elements content Z=0.001, 0.004, 0.02, 0.05 and 0.1 for the white dwarf masses of 0.4, 0.6, 0.7, 0.8, 0.9, 1.0, 1.2 and 1.35 M. These tables list the characteristic values of the envelope such as the photospheric temperature, the velocity, the wind mass-loss rate and uxes of four wavelength bands. The updated OPAL opacity (1996) is used. Subject headings: globular clusters: general { stars; abundances { stars: mass-loss { novae, cataclysmic variables { stars; Population II 1

2 1. Introduction The optically thick wind is a continuum-radiation driven mass-loss in which the acceleration occurs deep inside the photosphere (Friedjung 1966). It occurs in various stages of stellar evolution such as nova outbursts, helium nova outbursts, the planetary nebula nuclei, the last helium shell ash and X-ray bursts (Finzi & Wolf 1971, Ruggles & Bath 1979, Kato 1983a, Kato 1983b, Quinn & Paczynski 1985, Kato & Hachisu 1994, Kato 1996) and also suggested in Wolf- Rayet stars and in very massive stars(kato 1985, Kato & Iben 1992). Among these phenomena, the optically thick wind theory has been developed mainly in the study of nova outbursts and has recently succeeded in reproducing theoretical light curves of the decay phase of novae (Kato 1994, Kato & Hachisu 1994, Kato 1995). The recent progress in nova outburst theory is due to the OPAL opacity (Rogers & Iglesias 1992, Iglesias & Rogers 1993, Iglesias & Rogers 1996). The OPAL opacity has a large peak deep inside to the photosphere and it causes strong wind mass loss that reduces nova decay time-scale. Using the OPAL opacity, Kato has rstly succeeded in reproducing reliable light curve of classical nova, nova Cygni 1978 (Kato 1994). Kato and Hachisu (1994) calculated many light curves for various set of the white dwarf mass and the chemical composition of the envelope, and showed that the light curve of nova depends primary on the white dwarf mass and the secondly on the chemical composition. The fast nova corresponds to massive white dwarfs and the slow nova less massive white dwarfs ( Kato 1994 for nova Cygni 1978; Kato 1995 for nova Mus 1983, four recurrent novae and slow nova RR Pic; Kato 1997a for symbiotic novae). In this way, the variety of the light curve and the evolutional time scale of nova decay phase have been explained by the optically thick wind theory as the variety of the white dwarf mass and the chemical composition of the envelope. The fastest and the slowest time scale of nova phenomena just correspond to the mass distribution of the white dwarfs. The similar conclusion is also derived from the time-dependent work by Prialnik & Kovetz (1995) with comprehensive sets for three parameters of the white dwarf mass, the mass accretion rate and the white dwarf temperature. The resultant time-scales of nova outbursts cover the entire region of the observed speed class of novae. In this way, one of the most important long-standing problems in nova theory, i.e., what determines the nova speed class, has been essentially solved. The next step of nova study, therefore, may begin with the application of these theoretical knowledge to numerous number of novae. Statistical information of nova binaries such as the white dwarf mass, the chemical composition of ejecta, dierences of these values among nova subclasses and dierences of nova properties between populations will be important clues for the study of origin and evolution of binary systems. Novae have been discovered in the Galaxy, LMC, M31, M33, NGC5128 and the giant elliptical galaxies and a spiral galaxy in Virgo cluster (Hubble 1929, Arp 1956, Payne-Gaposchkin 1957, Della Valle et al. 1994, Pritchet & van den Bergh 1987, Ferrarese et al. 1996) and a few objects have been suggested to associate with the galactic globular clusters (Payne- Gaposchkin 1957, Cohen 1985). Recent works (Della Valle et al. 1994, Duerbeck 1990) have reported the statistical dierences of nova properties between those appeared in the bulge and the disk of the galaxies and also in dierent Hubble type of galaxies. These discoveries of novae in dierent type of galaxies with dierent chemical circumstances arise an theoretical attention on the dierences of nova properties between the new and the old populations. In the theoretical viewpoint, the nova speed class is essentially determined by the strength of the wind acceleration which is mainly governed by the white dwarf mass and the chemical composition of the envelope. Beside the parameters such as the white dwarf mass, the dominant source of the wind acceleration is iron lines in the new opacity. Therefore, we expect the characteristic dierence of nova outburst among dierent stellar populations. The aim of this paper is to show the dierences of novae in populations I and II and to provide a way to estimate the white dwarf mass from light curve tting or X-ray turn-o time. The white dwarf mass distribution of nova system in dierent populations will be an important information for the study of binary evolution and stellar evolution. There are several dierent types of novae; the classical nova, the slow nova, the recurrent nova, the last He shell ash (e.g. Sakurai's object: Duerbeck &Benetti 1996) and the He nova (Kato, Saio & Hachisu 1989). These objects are very similar phenomena in the theoretical point of view with only the dierences in nuclear fuel, the chemical composition of the envelope and other parameters. In addition to 2

3 these objects, other type of phenomena can be deal with in a same scheme of the optically thick wind theory, i.e., the evolution of the central star of planetary nebulae (Kato 1997b) and the wind mass loss from the accreting white dwarf in a close binary that lead to type Ia supernova (Hachisu, Kato & Nomoto 1996). Expecting wide applications of the optically thick winds to such various types of phenomena, this series of papers has been planned to provide numerical data with a wide range of chemical composition and the white dwarf mass. Paper I, the present paper, is designed to provide light curves of classical novae of population I and II, Paper II will present comprehensive set of light curves for recurrent nova outbursts. Section 2 of this paper will give a brief description of the method and assumptions for optically thick winds, and eects of the iron content thorough the opacity on the wind solution. The light curves, the X-ray turn-o time, and the tables are presented in section 3. Discussions and conclusions follow. 2. Z-dependence of Optically Thick Wind Solutions The structure of the mass-losing envelope around a degenerate core is obtained by solving the equations of motion, continuity, diusion and energy conservation with the assumptions of steady-state and spherical symmetry. The wind mass loss rate is determined as an eigenvalue of the boundary-value problem. No mechanism is taken into account that causes wind acceleration outside the photosphere. When no optically thick wind occurs, the envelope is represented by hydrostatic solutions with the mixing length parameter of 1.5. The equations, numerical methods and approximations are described in detail in Kato & Hachisu 1994 (hereafter referred as KH94). The updated OPAL opacity (Iglesias & Rogers 1996) is used in the present series of papers. The eects of updated OPAL opacity is discussed in section 4. The envelope structure is numerically obtained in the region between the photosphere and the bottom of the envelope. Hydrogen nuclear burning occurs at the bottom of the envelope that is equal to the surface of the white dwarf. The chemical composition of the envelope is assumed to be uniform with X = 0:35; C=0.1, O=0.2 for hydrogen, carbon and oxygen and with heavy elements of Z = 0:001, 0.004, 0.02, 0.05 and 0.1. Note that this heavy elements content, Z, is for a mixture of 19 heavy elements that includes additional carbon and oxygen (Iglesias & Rogers 1996). For the white dwarf radius, the Chandrasekhar radius is assumed for M wd 1:3 M and values taken from Nomoto, Thielemann, & Yokoi (1984) for 1.35 M. Figure 1 shows the opacity run for the ve wind solutions on a 1.0 M white dwarf with dierent Z values. The iron peak of the opacity around log T 5:2 is so prominent and still recognizable even in case of Z= This large peak causes strong acceleration of the optically thick wind. Figure 2 shows the quick increase of the velocity around 0.5 R(= log r = 10:542) where the opacity rapidly increases. The velocity reaches a terminal value far inside to the photosphere. This terminal velocity is almost the same in Z=0.02 and 0.05 models, but substantially small in the Z=0.001 model. The prole of the temperature and the density are also very similar except the small deviation of the Z=0.001 model. Figure 3 shows the distribution of the diusive luminosity and the Eddington luminosity L Edd = 4cGM WD = throughout the envelope. The diusive luminosity decreases outward especially where the Eddington luminosity quickly decreases, i.e., where the opacity quickly increases outward. This decrease in the diusive luminosity is due to energy consumption for driving the wind, and then the high Z models show larger energy decrease to cause massive wind. The optically thick wind occurs in a wide region of the H-R diagram as shown in gure 4. In the dashed region, the optically thick wind occurs and no static solution exists. The wind occurs in M WD 0:4 M for Z 0.1, M WD 0:5 M for Z = 0:02 and 0.05, M WD 0:6 M for Z=0.004, and M WD 0:8 M for Z= No wind occurs in the static region (solid part) in gure 4. The low surface luminosity in low temperature models with large Z is due to the large energy consumption to drive the massive wind as shown in gure 3. Figure 5 shows the wind velocity at the photosphere for several sequences appeared in gure 4. The wind velocity increases with the surface temperature and has a maximum value at log T 5:15. At a given surface temperature, the wind velocity is small for less massive white dwarfs and also for small Z values. The ratio of the velocities in 1.0 M models with Z=0.1 and is 3.8 and the ratio of 1.35 M and 0.6 M with Z=0.1 is 1.5 at log T = 4:5. Thus the dependence of the velocities on Z is stronger than in the white dwarf mass. 3

4 The total mass decreasing rate, which is the summation of the mass decreasing rate due to the wind mass loss and the nuclear burning, is plotted against the surface temperature (Figure 6) and the envelope mass (Figure 7). For the same white dwarf mass the large wind mass loss rate is obtained for high Z models. The ratio of the wind mass loss rate of Z=0.1 to models in 1.0 M amounts 2.4 at log T = 4:6 in gure Light curve and X-ray turn-o time 3.1. Decay phase of novae Nova outbursts are triggered by unstable hydrogen shell ash in the envelope around a white dwarf. In the rising phase, the envelope quickly expands with increasing luminosity and the strong mass-loss begins. After the optical maximum the envelope settles down to a thermal equilibrium and afterward the photospheric temperature increases in time which causes the decrease in the visual magnitude. In this decay phase, the nova evolution can be followed by the quasi-evolutional sequence as described in detail in KH94. This sequence is constructed from the steady mass-loss solutions, the static solutions and the solutions with constant envelope mass as in gure 4. In this H-R diagram, the star moves leftward with decreasing envelope mass due to the wind mass loss and nuclear burning. The photospheric temperature increases in time. After the wind stops, the star continues to move leftward due to nuclear burning until hydrogen nuclear burning extinguishes at the point denoted by the dot in gure 4. Beyond this point, the track is calculated with the assumption of constant envelope mass. The theoretical light curves are calculated for each sequence from the mass decreasing rates due to the wind and the nuclear burning, and the magnitudes estimated from the photospheric blackbody temperature (KH94). As the star moves leftward in the H-R diagram, the photospheric temperature rises in time and the mean energy of photon emission becomes higher. Therefore, the UV ux increases with the optical ux decrease, and after the UV maximum, the extreme UV and supersoft X-ray uxes increase. Figures 8 and 9 show the theoretical light curves for various set of the white dwarf mass and Z. The supersoft X-ray ux increases in the later phase after the optical magnitude drops to M v > 0. These light curves develop quickly in massive white dwarfs than in less massive stars because of the small envelope mass and the large mass-loss rate. For the same white dwarf mass, stars with small Z evolve slowly due to the small mass-loss rate. The dierence of the decline rate in the light curves due to the dierence of Z (Z=0.1 to 0.001) is smaller than the dierence due to the white dwarf mass by 0.2 M Tables for envelope solutions The characteristic values of the envelope solutions in the sequence with 1.0 M white dwarf and Z = 0:02; 0:004 and 0.1 are listed in tables 1-3. Among twenty six sequences calculated, only three are tabulated here with reduced row number. The sequences calculated are for 1.35 M, 1.0 M and 0.6 M with Z=0.001, 0.004, 0.02, 0.05 and 0.1, for 1.2 M and 0.8 M with Z=0.001, 0.004, 0.02 and 0.05, for 0.9 M with Z=0.02 and for 0.4 M with Z=0.02. All these twenty six sequences will be published in computer-readable form in the AAS CD-ROM Series, Volume XX(unknown). Each table contains one sequence with specied white dwarf mass and Z. Tables here are divided into two parts, Table A and Table B, due to the limit of space. Table A shows the characteristic values of the envelope. The rst four columns show the photospheric values of the temperature, the luminosity, the radius, and the velocity. Next the table lists the envelope mass, the mass decreasing rates due to the optically thick wind and hydrogen nuclear burning. The last two are the radius and the velocity at the critical point. The solutions are listed along the sequence of the H-R track toward leftward and then ordered as the wind phase, the static phase and the stage after hydrogen burning extinguishes. For example, the rst 8 solutions in Table 1, i.e., log T ph = to are the wind solutions. The solution with asterisk, log T ph =5.446, is the static solution just after the wind stops (the maximum static solution in KH94). The static phase continues until hydrogen burning extinguishes at log T ph =5.848, after that the sequence is followed by the solutions of constant envelope mass. These three dierent phases are easily distinguished in the tables. The solutions after hydrogen extinguishes are shown only for sequences of Z=0.02. Table B lists the light curve data, i.e., the evolutional time in units of year, the bolometric and visual magnitudes and the uxes of UV, EUV and supersoft X-ray, i.e., F UV = log(l UV =4D 2 ), F EUV = 4

5 log(l EUV =4D 2 ), F SSX = log(l SSX =4D 2 ), where UV, EUV and SSX are the ux of ultraviolet ( A), extreme ultraviolet ( A), supersoft X-ray ( A), respectively, and D = 1.0 kpc is the arbitrarily assumed distance. The uxes under -30 are denoted by -30 in these tables. The example of tting is shown in KH94 and Kato (1994) by the case of Nova Cygni X-ray turn-o time Figures 8-9 and Tables 1-3 show that a nova becomes a bright EUV/supersoft X-ray source in the later phase after the optical ux drops. These strong short-wavelength uxes continue after the wind massloss stops and drop shortly after the hydrogen burning extinguishes. Figure 10 shows the X-ray turn-o time for various white dwarf mass and Z value. The X-ray turno time, from the optical discovery or from the optical peak to the X-ray ux turn-o, is approximated here, as the time interval of nova decay phase from log T = 4:0 to the extinguish point of hydrogen burning. The optical peak does not exactly correspond to the stage of log T = 4:0, but the evolutional speed is very rapid around there due to the strong wind mass loss as shown in tables 1-3 and we can neglect the small time interval between the optical peak and the starting point of our sequence of solutions. In many hydrogen shell ashes the rising phase is very short compared with the later phase and then this X-ray turn-o time is also a good approximation to the total duration time, i.e., from the ignition of hydrogen burning to the end of hydrogen burning. As shown in gure 10 this X-ray turn-o time is very sensitive on the white dwarf mass and varies from 0.1 years in 1.35 M to 10 years in 0.6 M. As the dependence on Z is weak, this X-ray turno time is a good indicator of the white dwarf mass. Figure 11 shows the period for classical novae being a bright supersoft X-ray source in its decay phase i.e., log F ssx 07:0. This time-scale can also be used to estimate the white dwarf mass. Nova Muscae 1983 (GQ Mus), discovered in 1983 January, was detected by EXOSAT and ROSAT as a luminous supersoft X-ray source. The X-ray ux is black-body-like and consistent with the Eddington ux of one solar mass star. This X-ray ux turned o in the period between 1992 February and 1993 August (Shanley et al. 1995). The ten years of the X-ray turn-o time give an estimate of the white dwarf of Nova Mus 1983 to be about 0.6 M from gure 10 with Z=0.02. The theoretical light curves of visual, UV and supersoft X-ray bands for M are consistent with the observational data (Kato 1995). Another classical nova, Nova Cygni 1992 (V1974 Cyg) was observed by ROSAT to be a very bright supersoft X-ray source. This X-ray ux reaches the at peak at about 300 days and begins to decline at days after the outburst (Krautter et al. 1996). Using gure 10 we can estimate the white dwarf mass to be about M with Z=0.02. In this way the X-ray turn-o time is a useful tool to estimate the white dwarf mass Novae of populations I and II Novae have been discovered in dierent places of galaxies in dierent Hubble type as mentioned in section 1. The Fe abundance changes place to place, typically from Z=0.001 of globular clusters to Z= of bulge population. In such various chemical circumstances we expect the systematic dierence of nova properties. As already shown, novae of populations I and II differ in the wind velocity, the decline rate of the light curve, and the X-ray turn-o time. As these values depend dierently on nova parameters, we can estimate both of the white dwarf mass and Fe abundance if sucient data are available. The X-ray turn-o time as well as the shape of the light curves is a good indicator of the white dwarf mass. When the optical data is insucient to determine the parameters uniquely, the X-ray turn-o time is especially useful. For example, the optical light curve of 1.0 M with Z=0.001 white dwarf are very similar to that of 0.9 M with Z=0.02 in the early bright phase (M v -1) and may be dicult to distinguish each other from the tting with scattered observational data. Their evolutional speed becomes dierent in the very later phase, which we hardly know from optical data but do from the X-ray turno time. The dierence between the two stars is suf- ciently large: 1.5 years of 1.0 M with Z=0.001 and 2.0 years of 0.9 M with Z=0.02. Therefore, the X- ray turn-o time is the eective way to estimate the white dwarf mass. The other parameter, the Fe abundance, appeals its presence most remarkably in the expansion velocity. Figure 12 shows the maximum value of the veloc- 5

6 ity in gure 5 as a function of Z and the white dwarf mass. This maximum velocity depends strongly on the iron content and weakly depends on the white dwarf mass. For example, the maximum velocity in less massive white dwarf of population I nova, i.e., 0.6 M with Z=0.02, is 1.4 times larger than that of the very massive population II novae of 1.35 M with Z= For the same white dwarf mass the ratio of the velocities of Z=0.1 to is 3.6 in 1.35 M and 4.4 in 1.0 M. From this strong dependence on Z we can expect a detectable statistical dierence in the mean expansion velocity among novae in globular clusters, LMC, the disk and the bulge of the galaxies. Nova T Sco, observed in 1860, has been suggested its possible association with globular cluster NGC 6093 (Payne-Gaposchkin 1957, Cohen 1985). The drop in brightness seems to have been fairly rapid, three magnitudes in 21 days. Such a rapid evolution is consistent with our M white dwarf models with Z=0.001 from gure 8 or from tables in CD ROM for details. No other information such as the expansion velocity are available unfortunately, and future discovery of novae in globular clusters will be of very interesting. The statistical dierence among novae in bulge, the disk, LMC and galaxies of dierent Hubble types is suggested (Della Valle et al. 1994) from the distribution of maximum luminosity and the rate of decline. They showed that the disk novae are fast and bright whereas the bulge novae are slow and faint. Many of the novae in LMC are very fast which seems to be a contrast to those appeared in M31. As the bulge stars generally show larger iron abundance than in the disk and LMC, these dierences of fast nova distribution can be attributed as the dierence of the white dwarf mass distribution rather than iron abundance from the present results. In other words there seems to be the preference of massive white dwarfs in low Fe circumstances. The suggested massive white dwarf in T Sco, possible association to the globular cluster arises further attention from this point of view. The distribution of the white dwarf mass of nova system is not well known in our galaxy, LMC, globular clusters, nor in extragalaxies, and it will be very interesting and useful clue for the study of the white dwarfs and the binary evolution. 4. Discussions and conclusions The updated OPAL opacity (Iglesias & Rogers 1996) includes 7 additional elements than in the rst version of the OPAL opacity (Rogers & Iglesias 1992, Iglesias & Rogers 1993) and has larger enhancement than in the rst version by 10 percent at the iron peak as shown in gure 13. This change of the opacity causes little change in the envelope structure such as the distribution of the temperature and the density, but percent increase of the expansion velocity and percent increase of the mass loss rate. The period of the static phase is hardly changed. As a result the X-ray turn-o time is decreased by about 10 percent with the updated OPAL opacity. Our main results are summarized as follows; 1. Using the optically thick wind theory the light curves of the decay phase of classical novae are obtained for very wide region of iron abundance, from Z=0.001 to 0.1, and the white dwarf mass. Optically thick wind occurs in M WD 0.8 M for Z=0.001, M WD 0.6 Mfor Z=0.004 and M WD 0.5 M for Z=0.02 and 0.05 and M WD 0.4 M for Z=0.1. This means that mass ejection during nova phenomena will occur in a very wide range of Fe content, that corresponds to all of the Hubble type of galaxies and globular clusters. 2. The time scale of nova light curves, the wind mass loss rate and the wind velocity depend on the Fe abundance. For small Fe abundance, the light curve develops slowly and the X-ray turn-o time is long, e.g., it varies as 0.95, 1.2 and 1.5 yr for Z=0.1, 0.02, and for 1.0 M white dwarf. The wind mass loss rate of Z=0.001 is as small as 1=2 of that of Z=0.05 in 1.0 M with the same envelope mass. For Z=0.001, i.e., population II, the less massive white dwarfs M have no wind mass ejection. 3. The X-ray turn-o time depends strongly on the white dwarf mass and weakly on the iron abundance. Therefore, the X-ray turn-o time and X-ray lifetime of novae yield a good estimate of the white dwarf mass. For GQ Mus and V1974 Cyg the white dwarf mass is estimated to be 0:6M and 1:0 M, respectively. The X-ray turn-o time is a useful tool for the future statistical study of the possible Fe dependence of the white dwarf mass distribution. 4. The statistical dierence of the wind velocity is predicted between novae in globular clusters and in the galactic disk. The expansion velocity of nova of old population (Z = 0:001) is substantially small 6

7 to be about 1/3 of that of the galactic disk nova (Z=0.02). 5. Twenty six tables are presented in the computer readable form of CD ROM that will be useful for light curve analysis of classical novae with dierent Fe abundances. The author is grateful to Izumi Hachisu for stimulating discussions. This work has been supported in part by the Grant-in-Aid for Scientic Research ( , ) of the Japanese Ministry of Education, Science and Culture. REFERENCES Arp, H., C., 1956, AJ, 61, 15 Cohen, J. G ApJ, 292,90 Della Valle, M., Rosino, L., Bianchini, A. & Livio, M., 1994, A&A, 287, 403 Duerbeck, H. W., 1990, in Physics of Classical Novae, eds. Cassatella, A., & Viotti, R. (Berlin: Springer) p.34 Duerbeck, H. W.& Benetti, S. 1996, ApJ, 468 L111 Ferrarese, L., Livio, M., Freedman, W., Saha, A., Stetson, P. B., Ford, H. C., Hill, R. J. & Madore, B., F ApJ, 468, L95 Finzi, A. & Wolf, R.A. 1971, A&A, 11, 418 Friedjung, M. 1966, MNRAS, 132, 317 Hachisu, I., Kato, M. & Nomoto K. 1996, ApJ, 470, L97 Hubble, E., 1929, ApJ, 69, 103 Iglesias, C.A. & Rogers, F. J., 1993, ApJ, 412, 752 Iglesias, C.A. & Rogers, F. J., 1996, ApJ, 464, 943 Kato, M. 1983a, PASJ, 35, 33 Kato, M. 1983b, PASJ, 35, 507 Kato, M. 1985, PASJ, 37,19 Kato, M A&A, 281, L49 Kato, M. 1995, in Cataclysmic Variables, eds. Bianchini, A., Della Valle, M. & Orio, M. (Dordrecht: Kluwer) p.243 Kato, M. 1996, in Supersoft X-ray Sources, ed. Greiner, J. (Berlin: Springer) p.15 Kato, M. 1997a, in Physical Processes in Symbiotic Binaries, ed. Mikolajewska, J.,in press Kato, M. 1997b, in Advances in Stellar Evolution eds. Rood, R.T. & Renzini, A. (Cambridge:Cambridge U. Press),in press Kato, M. & Hachisu, I., 1994, ApJ, 437, 802 (KH94) Kato, M. & Iben, I. Jr., 1992, ApJ, 394, 305 Kato, M., Saio, H. & Hachisu, I., 1989 ApJ, 340, 509 Krautter, J., Ogelman, H., Starreld, S, Wichmann, R. & Pfeermann, E ApJ, 456, 788 Nomoto, K., Thielemann, F., & Yokoi, K. 1984, ApJ, 286, 644 Payne-Gaposchkin, C., 1957, The Galactic Novae (Amsterdam: North-Holland) p.190 Pritchet, C. J. & van den Bergh, S., 1987 ApJ, 318, 507 Prialnik, D. & Kovetz, A., 1995, ApJ, 445, 789 Quinn, T., & Paczynski, B. 1985, ApJ, 289, 634 Rogers, F.J. & Iglesias, C.A, 1992, ApJS, 79,507 Ruggles, C.L.N, & Bath, G.T. 1979, A&A, 80, 97 Shanley, L, Ogelman, J. S., Gallagher, J.S., Orio, M. and Krautter, J. 1995, ApJ, 438, L95 Fig. 1. The run of the updated OPAL opacity for the wind solutions of 1.0 M white dwarf with Z = 0:1; 0:05; 0:02, and (from upper to lower). The radius of the critical point (KH94) is assumed to be R cr = 0:5R (denoted by dots) for all the solutions. The left end of each curve corresponds to the photosphere. Fig. 2. The change of the velocity, the temperature, and the density of three solutions plotted in gure 1; Z =0.05 (dashed), 0.02 (solid) and (dotted). The dot denotes the critical point. The thin line shows t he escape velocity V esc = (GM WD =r) 1=2. This 2-column preprint was prepared with the AAS L A TEX macros v4.0. 7

8 Table 1 Table 1A. Envelope solutions for classical nova of 1.0 M with Z=0.02. T ph a L ph b R ph c v ph d 1M e _ M wind f _M nuclear f R cr g v cr h a in units of log T (K) b in units of log L (ergs s 01 ) c in units of log R ph (R ) d in units of log v ph (cm s 01 ) e in units of log 1M (M ) f in units of log j M _ j (M yr 01 ) g in units of R cr (R ) h in units of log v cr (cm s 01 ) 3 the maximum static solution 8

9 Table 2 Table 1B. Envelope solutions for classical nova of 1.0 M with Z=0.02. t (yrs) M bol M v F UV a F EUV a F SSX a 0.000e e e e e e e e e e e e e e e a in units of F = log(l=4d 2 ); D=1kpc 3 the maximum static solution Table 3 Table 2A. Envelope solutions for classical nova of 1.0 M with Z= T ph L ph R ph v ph 1M _M wind _M nuclear R cr v cr

10 Table 4 Table 2B. Envelope solutions for classical nova of 1.0 M with Z= t (yrs) M bol M v F UV F EUV F SSX 0.000e e e e e e e e e e e e e e e e Table 5 Table 3A. Envelope solutions for classical nova of 1.0 M with Z=0.1. T ph L ph R ph v ph 1M _M wind _M nuclear R cr v cr

11 Table 6 Table 3B. Envelope solutions for classical nova of 1.0 M with Z=0.1. t (yrs) M bol M v F UV F EUV F SSX 0.000e e e e e e e e e e e e e e e e The outermost point of each curves corresponds to the photosphere. Fig. 3. The diusive luminosity (thick curves) is plotted together with the Eddington luminosity (thin) for the same solutions as in Figure 2. There appears a strong super-eddington region around log r(cm) = 10:8 (for Z=0.05; dashed) and for (Z=0.001; dotted), which correspond to the opacity peak due to iron lines. The lled circle denotes the critical point. Fig. 4. Theoretical H-R tracks for 1.35, 1.2, 1.0, 0.8, 0.6 and 0.4 M white dwarfs. The heavy elements abundance other than additional carbon (0.1) and oxygen (0.2) is assumed to be Z=0.001, 0.004, 0.02, 0.05 and 0.1 from upper to lower for 1.35, 1.0 and 0.6 M, Z=0.001, 0.02 and 0.05 for 1.2 and 0.8 M. Optically thick wind occurs in the dashed region. No wind occurs in 0.6 M with Z=0.001 and 0.4 M with Z=0.02. Hydrogen burning extinguishes at the lled circles. Fig. 5. The wind velocity at the photosphere is plotted against the surface temperature. Thick curves denote the ve sequence for 1.0 M of Z=0.1, 0.05, 0.02, and from upper to lower. Thin solid curves are for 1.35 M with Z=0.1 (upper) and (lower), the two thin dotted curves 0.6 M with Z=0.1 (upper) and Z=0.004 (lower), and the thin dashed curve 0.8 M with Z= Fig. 6. The total mass decreasing rate, that is the summation of the wind mass loss rate and the mass decreasing rate due to hydrogen nuclear burning, is plotted against the photospheric temperature. The white dwarf mass is attached to each set of the curves. For 1.35 M Z=0.1, 0.02 and from upper to lower, and the other solutions are the same as those plotted in gure 4. No wind mass loss occurs in 0.6 M with Z=0.001 (the lowest curve). Fig. 7. The total mass decreasing rate is plotted against the envelope mass. The number attached below the curves denotes the white dwarf mass. The solutions are the same as those plotted in gure 4. For each white dwarf mass Z decreases downward; for example, in 1.35 M, Z=0.1, 0.05, 0.02, and from upper to lower. No wind occurs in 0.6 M with Z=0.001 (dotted curve). Fig. 8. Theoretical light curves for 1.35, 1.2 and 1.0 M with various heavy elements content. For the visual light curves the curves are, for from left to right, 1.35 M (solid) with Z=0.1, 0.02 (thick) and 0.001, 1.2 M (dotted) with Z=0.05, 0.02 and 0.001, 1.0 M (solid) with Z=0.1, 0.05, 0.02 (thick), and The light curves of supersoft X-ray band (30A- 11

12 100 A) are also shown for 1.2 M with Z = 0:02 and ve cases for 1.0 M. The optically thick wind stops at the point at which the supersoft X-ray ax suddenly rises (e.g. t = 218 days in 1.0 M with Z = 0:02). Fig. 9. The same gure as in gure 7 but for 0.8, 0.7 and 0.6 M. The visual light curves denote, from left to right, 0.8 M (dotted) with Z=0.05, 0.02 and 0.001, 0.6 M (solid) with Z=0.1, 0.05, 0.02, and (thick) and 0.7 M with Z=0.001 (thick dotted). The supersoft X-ray uxes are plotted for 0.6 (left) and 0.7 M with Z= Fig. 10. The X-ray turn-o time is plotted against the white dwarf mass. Z is assumed to be 0.001, 0.004, 0.02, 0.05 and 0.1 from upper to lower. Fig. 11. The same as in gure 11 but for the lifetime of novae as bright supersoft X-ray sources (F SSX 07). The supersoft X-ray ux does not reach -7.0 in 0.4 M with Z= Fig. 12. The maximum wind velocity for a white dwarf during the course of nova decay phase is plotted against the white dwarf mass. The white dwarf mass is attached to each curve. Fig. 13. Comparison of the rst and the updated OPAL opacities of the nova wind solutions of Z=0.02 and (lower two curves). Solid: the updated OPAL opacity (1996), Dashed: rst version of the OPAL opacity (1992). The other parameters are the same as those in gure 1. 12