2013ApJ...769...94Z

The shape of OB-star ionizing spectral energy distributions (SEDs) is a critical component in many diagnostics of galaxy and interstellar medium properties. To quantitatively examine the shape of the OB-star SED, we compare long slit observations of single-star, Large Magellanic Cloud H II regions to the predictions from CLOUDY photoionization simulations that use CoStar, TLUSTY, and WM-basic stellar atmosphere models as the ionizing source. For each atmosphere model, we run grids of H II region simulations with the effective temperature (T_eff) of the star as a free parameter. The best SEDs from each atmosphere code are found by matching the predicted emission-line spectra with those observed from the nebulae. By assuming a clumpy gas distribution, all atmosphere codes are able to reproduce the observed emission lines, except at the highest energy transitions ≳ 40 eV. Taking into account both low and high energy transitions, we find that simulations using WM-basic produce the best agreement with the observed line ratios. The rates of ionizing photons from different atmosphere models vary systematically with the relative hardness of the SEDs. However, in general the rates produced by the model SEDs, for standard log(g) = 4.0 models, are consistent with the rates derived from the Hα luminosities. We find that our effective temperatures inferred from the nebular ionization balance are consistent with those predicted by conventional photospheric-based calibrations from the literature. We suggest that future spectral type to T_eff calibrations can be constructed from nebular data.