2002ApJ...579..760C

We examine the AAVSO light curve of U Gem from 1908 to 2002, with particular focus on the 1985 October outburst. This outburst was longer than any other seen in U Gem by about a factor of 2 and appears to be unique among all dwarf nova outbursts seen in systems with orbital periods longer than 3 hr in that one can measure the decay timescale during the initial slow decay. This rate is ∼26±6 days mag^–1. Using estimates of the rate of accretion during outburst taken from Froning et al., one can show that ∼10²⁴ g of gas was accreted onto the white dwarf during the outburst, which constrains the surface density in the outer accretion disk to be ∼600 g.cm^–2. The observed timescale for decay is consistent with that expected in U Gem, given its orbital period and disk mass at the time the outburst began. The data are not of sufficient quality to be able to ascertain a deviation from exponentiality in the decay light curve (as in the SU UMa stars' superoutbursts). When coupled with the viscous time inferred from the (short orbital period) SU UMa stars, the U Gem viscous timescale lends support to the standard model in which the decays in dwarf novae can either be viscous or thermal, with the ratio between them being roughly h/r, where h is the vertical pressure scale height in the disk. Indeed, dwarf novae are the only systems for which one can be reasonably certain of the identification of ``viscous'' and ``thermal'' decays.