ES Dra was observed in an intermediate state and it is not clear whether the main component is a WD or a disk, as the inclination and WD mass of the system are not known. Because of that, the FUSE spectrum of ES Dra is modeled first with a single WD, and then with a DISK alone. We do not consider any WD+disk model, as there is a degeneracy of the solution (there are many WD+disk models that can fit this spectrum covering a large area in the parameter space). Details on the processing of the observed spectrum and the spectral analysis are given in Godon et al. (2009). The spectrum exhibits some sharp emission lines due to air glow as well as ISM absorption lines (mainly molecular H, but also some metals). There could be some broad emission near 977A (C iii ?) as well as near 990A . The C iii (1175) feature seem to have a blue shifted absorption component and a red shifted emission component. We identify with some reliability absorption lines/features from C ii (1010), Si iv (1066), Si iii (1110 & 1143), Si iv (1123-1128), as these lines are located at the expected wavelengths and agree grossly with the model. The S iv (1063, 1073) lines are contaminated with ISM absorption. Some of the lines that are marked on the figure are present in the theoretical spectrum but are not detected in the observed spectrum (e.g. S iv 1006 & 1100).
The flux (vertical axis) is given in ergs/sec/cm2/A and the wavelength (horizontal axis) is given in Angstroem. For clarity the spectrum is shown on 3 panels: the upper panel from about 900 to 1000A, the middle panel from about 1000A to 1100A, and the lower panel from about 1100A to 1200A. Air glow emission is anoted above each panel with a "+" sign inside a circle. Strong emission lines reaching the upper part of the frame in each pael is marked just above the panel. The Hydrogen series is marked below the x-axis in the upper and middle panel and the hydrogen cut-off limit is clearly marked around 915A with an increasing density of tick marks. The He I (1168A) is marked below the lower panel. Oxygen I lines are shown with an arrow "OI -->" if many (ISM) OI lines are present, they are just annotated with verical tick marks in the lower portion of each panel. ISM molecular hydrogen lines are either individually annotated vertically (e.g. L10R1) or just marked with an arrow "MH -->" and vertical tick marks (usually in the upper part of each panel). Features resulting from fixed pattern noise introduced by FUSE detect ors are indicated with "FPN;.
When a model is shown, the portions of the observed spectrum that have been masked before the spectral fit are in blue, while the rest of the observed spectrum is in red. The model fit is shown the solide line in black (or possibly white). If the model is a composite made of a disk model and a white dwarf model, then these are shown with the dashed- and dotted lines (respectively). Some models include the effect of quasi-molecular HI absorption features, these are labelled "Q-mol" in the spectrum.