Modeling the reflection effect in AA Dor
AA Dor is a bright (B=10.8 mag) eclipsing Algol-type subdwarf+brown-dwarf (sdO+dM) binary in constellation Doradus. The hot (T=42000 K) sdO star has a radius of 0.21 R⊙ and irradiates its companion, a roughly half as large brown dwarf (T≈3000 K) orbiting at a distance of 1.2 R⊙. Due to the small orbit the inner hemisphere of the brown dwarf is heated to about 18000 K. A V-band light curve has been obtained by Hilditch et al. (2003), that confirms a strong reflection effect and the existence of a hot spot on the companion. For a detailed overview of the system please visit the extensive works done by Thomas Rauch et al., who also obtained phase resolved high-quality UVES spectroscopy covering the entire 6.3 hrs orbital period.
The optical spectrum of AA Dor is dominated by the bright sdO star and only weak and smeared features of the companion can be visible as line profile variations of the sdO star. With careful re-reduction of the UVES data Vučković et al. (2016) (see also our poster) produced a high signal-to-noise mean sdO spectrum. By subtracting the mean spectrum from each individual spectra and combining the residuals in velocity space the authors extracted the spectrum of the irradiated hemisphere of the brown dwarf. This spectrum is full of metal emission lines and broad Balmer emissions that have central self-absorption. My Tlusty atmosphere models could reasonably reproduce most of these features and provided the temperature and density structure of the irradiated spot as in Figure 1.
In Figure 2 the B-band flux is calculated from the specific intensities. The hot spot is so much brighter that the brown dwarf is practically invisible on the night side. Figure 3 shows the temperature distribution of an old model. This model used the atmospheric temperature and density structures and integrated the optical depth inward. The surface intensity was calculated at different orbital phases. Obviously the hottest layers are on the top of the atmosphere. However, these tenuous layers are only visible when we look at them at grazing angles. Figure 2 and 3 suggest that the surface intensity is a complicated function of orbital phase and difficult to describe with a general limb-darkening law.