## Appendix A: treatment of the diffuse fieldThe ionized gas also contributes to the radiation field through the
capture of free electrons, hydrogen and helium ions. For a nebula
which is optically thick in the Lyman continuum a good approximation
is to suppose that the photons are absorbed in a local environment
near the place where the recombination has occurred, i.e. the OTS
approximation. This allows the use of the local variables, where and . The recombinations to the excited levels of He The OTS approximation is not as precise in the thermal balance as it is in the ionization balance because it does not take into account the fact that photons emitted after recombination may have different energies so that different penetration depths into the nebula, and thus some of these cannot be absorbed locally. For this reason we have solved for the transport of the diffuse field using the outward approximation (Williams 1967), i.e. each cell may receive ionizing photons from all inner cells. To calculate the radiation intensity at each point of the nebula we have built a grid of rays with impact parameters () at different radius. Then, given that is equivalent to know as in spherical symmetry, . The radiative transport along each ray with impact parameter is then solved: for , and , and with where is the coordinate along the ray () and is the gas source function (we have omitted the dependence on for clarity) and and are correspondly the emissivity and the gas opacity. The mean intensity is then calculated through a weighted summing over the rays: ## Appendix B: line transferAs output, the hydrodynamical models produce a set of magnitudes: density, temperature, velocity, etc, at each spatial location, but these can only be related to observable quantities if they are used to calculate either surface brightness or line profiles. To solve the radiative transport along a line of sight we used the method described by Yorke (1982). The line intensity is given by where , assuming that the continuum contribution to the line is negligible. If the source function varies linearly with optical depth, Eq. (B1) can be integrated to yield The emissivity and the opacity are and where and are the values at the line centre. is the frequency in the ion reference system, which is displaced from the frequency at rest by the Doppler effect, i.e. , assuming , where is the velocity component along the line of sight. After complete redistribution one obtains where is the line thermal width. As the emissivity and the opacity have the same frequency dependence the source function is independent from the frequency, and in the Eq. (B2) only the optical depth remains as a function of the frequency: where is the error function and we have supposed that the velocity component varies linearly in the interval , i.e. . is the coordinate along the line of sight. Finally the surface brightness can be determined from the integration over all frequencies of . © European Southern Observatory (ESO) 1998 Online publication: February 4, 1998 |