X-ray burst sources form a subclass of low-mass X-ray binary systems, in which accretion of gas from a secondary star onto the neutron star causes both persistent emission and recurrent bursts of X-rays. Most of the objects are Type I X-ray bursters, in which recurrent thermonuclear flashes develop in the neutron star envelope and raise the effective temperature above K, which is the cause of eruptive X-ray emission. Observational properties and the nature of these phenomena have been discussed in extensive review articles (Lewin & Joss 1981, 1983; Joss & Rappaport 1984; van Paradijs & Lewin 1988; Lewin, van Paradijs & Taam 1993).
X-ray burst spectra contain some information concerning basic parameters of the event. If we assume, that the X-ray emission is generated in a very hot neutron star atmosphere, then the analysis of a burst spectrum should yield values of the effective temperature and surface gravity in the stellar photosphere , with perhaps some hints concerning the chemical composition of matter. The analysis of model atmospheres was extensively described in Mihalas (1978). At the exceeding several MK and photon energies relevant to X-rays, all the classical equations and techniques presented there have to be enhanced by terms describing Compton scattering of photons by a very hot free electron gas (cf. Pomraning 1973; Rybicki & Lightman 1979). Model atmosphere computations including Compton scattering, and the relevant spectral analysis are very complex tasks (eg. London et al. 1986; Ebisuzaki 1987; Babul & Paczyski 1987; Madej 1989, 1991; Titarchuk 1994).
Due to the very limited spectral resolution of archival X-ray burst observations, which is of order 25 %, an interpretation of observational data requires fitting of the observed counts by predicted (or just assumed) theoretical spectra expressed by simple and compact formulae. At present, observational data are routinely fitted by a blackbody spectrum, which yields the observed color temperature . There exist several methods for the determination of the effective temperature of a burst at infinity, if the is known, which were recently reviewed by Lewin et al. (1993). Unfortunately, there exists no useful relation which determines gravity.
In the present paper we attempt to revise previous scales, and to give a method for the estimate of in the photosphere of a neutron star at various phases of a burst. Let us temporarily ignore general relativistic effects (redshift factor), then all theoretical spectra, , and , can be set to the values observed at infinity. The following considerations are restricted to the arbitrary case in which X-ray emitting gas is a mixture of hydrogen and helium alone.
© European Southern Observatory (ESO) 1997
Online publication: July 3, 1998