2006ApJ...648..541L

Observations of U Gem with FUSE confirm that the WD is heated by the outburst and cools during quiescence. At the end of an outburst, the best uniform-temperature WD model fits to the data indicate a temperature of 41,000-47,000 K, while in midquiescence, the temperature is 28,000-31,000 K, depending on the gravity assumed for the WD. Photospheric abundance patterns at the end of the outburst and in midquiescence show evidence of CNO processing. Improved fits to the spectra can be obtained assuming there is a hotter, heated portion of the WD, presumably an accretion belt, with a temperature of 60,000-70,000 K occupying 14%-32% of the surface immediately after outburst. However, other relatively simple models for the second component fit the data just as well, and there is no obvious signature that supports the hypothesis that the second component arises from a separate region of the WD surface. Hence, other physical explanations still must be considered to explain the time evolution of the spectrum of U Gem in quiescence. Strong orbital-phase-dependent absorption, most likely due to gas above the disk, was observed during the midquiescence spectrum. This material, which can be modeled in terms of gas with a temperature of 10,000-11,000 K and a density of 10¹³/cm3, has a column density of ∼2x10²¹/cm2 at orbital phase 0.6-0.85 and is probably the same material that has been observed to cause dips in the light curve at X-ray wavelengths in the past. The discrepancy described by Naylor et al. between the radius of the WD derived, on the one hand, by the UV spectral analysis and the distance to U Gem and, on the other hand, by the orbital elements and the gravitational redshift, remains a serious problem.