SIMBAD references

We have used the radiation field from a starburst population synthesis model appropriate for the brightest H II region of I Zw 18 to perform a photoionization model analysis of this object. We have investigated whether, with the observed nebular density distribution as revealed by the HST images and a total stellar radiation compatible with the observed UV flux, one could reproduce the constraints represented by the observed ionization structure, the He II λ4686/Hβ ratio, the Hα flux and the electron temperature indicated by the [O III] λ4363/5007 ratio. We have found that, even taking into account strong deviations from the adopted spectral energy distribution of the ionizing radiation and the effect of additional X-rays, the photoionization models yield too low a [O III] λ4363/5007 ratio by about 30%. This discrepancy is significant and poses an interesting problem, which cannot be solved by expected inaccuracies in the atomic data. The missing energy may be of the same order of magnitude as the one provided by the stellar photons or lower, depending on the way it acts on the [O III] λ4363 line. Elemental abundance determinations in I Zw 18 are affected by this problem. Before the nature of the missing heating source and its interactions with the nebular gas are better understood it is, however, not possible to estimate the typical uncertainties by which standard empirical methods are affected. Several side-products of our photoionization analysis of I Zw 18 are the following: (o) We have been able to reproduce the intensity of the nebular He II λ4686 using a stellar radiation field consistent with the observed Wolf-Rayet features in I Zw 18. (o) We have shown that the bright NW H II region in I Zw 18 does not absorb all the ionizing photons from the central star cluster, and that about half of them are available to ionize an extended diffuse envelope. (o) The [O i] emission can easily be accounted for by condensations or intervening filaments on the line of sight. (o) The intrinsic Hα/Hβ ratio is enhanced by collisional excitation to a value of 3. Previous estimates of the reddening of I Zw 18 are therefore slightly overestimated. (o) We give ionization correction factors appropriate for I Zw 18 that can be used for more accurate abundance determinations in this object once the electron temperature problem is better understood.