GW Librae is a quite extraordinary star, or pair of stars. It is a dwarf nova of the WZ Sge type, characterized by extremely rare and bright outbursts. It is also a pulsating white dwarf of the ZZ Ceti type. However it is the dwarf nova behaviour that is the subject of these observation notes. All observations were carried out with the 0.41 m reflector at Woodridge Observatory.
     At quiescence, GW Lib is at magnitude ~18.5V, but on April 13 it had reached 8.25V (see first plot below), an increase in brightness of 10,000 times. Such an increase led its discoverers in 1983 to think it was a nova, but spectra taken later showed it was a dwarf nova instead. These spectra also showed it had an orbital period of 76.78 m (1.28 h, 0.0546 d), the shortest known for a dwarf nova, and thus classifying it as UGSU, the SU UMa type of dwarf nova. UGSUs are characterised by occasional longer brighter outbursts marked by oscillations called superhumps. Since GW Lib had not had an outburst since discovery, which is unusually long for a dwarf nova and indicates a slow rate of mass accretion off the secondary star onto the accretion disk, there has been great interest in this outburst.
    In this plot from April 13, there are rather irregular oscillations of ~0.1 mag and indications of a periodicity of ~0.055 ± 0.005 d (1.32 h, 79.2 m), measured using Peranso's ANOVA and PDM routines. This period is the same as for the superhumps (below). It is not normal for UGSUs to show this periodic behaviour in the interregnum period after onset of maximum brightness and before the onset of superhumps.
    The gap at JD 04.18 is due to meridian flip, and the steep changes in brightness either side may be spurious. A B passband plot at the same time (not shown here) had a mean mag of ~7.95 but no difference in shape, indicating the outburst is uniformly blue.
    There are also fluctuations of ~0.02 mag with a period of very roughly 15 minutes, well above the noise level and visible in both the V and B plots, so probably not spurious. There is a good run of them from JD 04.13 to 04.17, visible better when the plot is significantly expanded horizontally.

The next observing run was on April 23, and this time though fainter (~10.05V) the outburst showed the triangular superhumps characteristic of UGSU superoutbursts, with 0.1 mag amplitude. Their period P_s = 0.055 ± 0.005 d (1.32 h, 79.2 m) which is scarcely longer than P_orb. Superoutbursts are believed due to a precessing elliptical accretion disk, and in this case P_s/P_orb = 1.007, the smallest I have been able to find in the literature, but to be expected for such a short orbital period. See e.g. B. Warner, Cataclysmic Variable Stars (Cambridge 1995), Table 3.3 p.132.

The next night was much the same, though a little fainter at ~10.1V. However the superhumps were stronger with an amplitude of nearly 0.2 mag, and the same P_sh. Superhumps often show colour change, being bluest (hottest) at minimum. However a V-I plot for this night's data (not shown here) showed no such periodicity.
The final observations of GW Lib were obtained the night after, April 25th. After that the weather closed in. The only noticeable change was that the mean brightness had declined another 0.1 mag.