2013ApJ...773...55S

We report the discovery by M. Linnolt on JD 2,455,665.7931 (UT 2011 April 14.29) of the sixth eruption of the recurrent nova T Pyxidis. This discovery was made just as the initial fast rise was starting, so with fast notification and response by observers worldwide, the entire initial rise was covered (the first for any nova), and with high time resolution in three filters. The speed of the rise peaked at 9 mag.day^–1, while the light curve is well fit over only the first two days by a model with a uniformly expanding sphere. We also report the discovery by R. Stubbings of a pre-eruption rise starting 18 days before the eruption, peaking 1.1 mag brighter than its long-time average, and then fading back toward quiescence 4 days before the eruption. This unique and mysterious behavior is only the fourth known (with V1500 Cyg, V533 Her, and T CrB) anticipatory rise closely spaced before a nova eruption. We present 19 timings of photometric minima from 1986 to 2011 February, where the orbital period is fast increasing with P/{dot}P = +313,000 yr. From 2008 to 2011, T Pyx had a small change in this rate of increase, so that the orbital period at the time of eruption was 0.07622950±0.00000008 days. This strong and steady increase of the orbital period can only come from mass transfer, for which we calculate a rate of (1.7-3.5)x10^–7 M_☉/yr. We report 6116 magnitudes between 1890 and 2011, for an average B = 15.59±0.01 from 1967 to 2011, which allows for an eruption in 2011 if the blue flux is nearly proportional to the accretion rate. The ultraviolet-optical-infrared spectral energy distribution is well fit by a power law with f_ν∝ν^1.0, although the narrow ultraviolet region has a tilt with a fit of f_ν∝ν^1/3. We prove that most of the T Pyx light is not coming from a disk, or any superposition of blackbodies, but rather is coming from some nonthermal source. We confirm the extinction measure from IUE with E(B - V) = 0.25±0.02 mag.