White Dwarf Heating and Subsequent Cooling in Dwarf Nova Outbursts

Abstract

We follow the time dependent thermal evolution of a white dwarf (WD) undergoing sudden accretion in a dwarf nova outburst, using both simulations and analytic estimates. The post-outburst lightcurve clearly separates into early times when the WD flux is high, and late times when the flux is near the quiescent level. The break between these two regimes, occurring at a time of order the outburst duration, corresponds to a thermal diffusion wave reaching the base of the freshly accreted layer. Our principal result is that long after the outburst, the fractional flux perturbation about the quiescent flux decays as a power law with time (and {\it not} as an exponential). We use this result to construct a simple fitting formula that yields estimates for both the quiescent flux and the accreted column, i.e. the total accreted mass divided by WD surface area. The WD mass is not well constrained by the late time lightcurve alone, but it can be inferred if the accreted mass is known from observations. We compare our work with the well-studied outburst of WZ Sge, finding that the cooling is well described by our model, giving an effective temperature $T_{\rm eff}=14,500 {\rm K}$ and accreted column $Δy\approx10^6 {\rm g cm^{-2}}$, in agreement with the modeling of Godon et al. To reconcile this accreted column with the accreted mass inferred from the bolometric accretion luminosity, a large WD mass $\gtrsim1.1M_\odot$ is needed. Our power law result is a valuable tool for making quick estimates of the outburst properties. We show that fitting the late time lightcurve with this formula yields a predicted column within 20% of that estimated from our full numerical calculations.