The X-Ray Spectrum of WZ Sagittae

We carried out an Advanced Satellite for Cosmology and Astrophysics (ASCA) observation of WZ Sagittae, to obtain the first X-ray spec- trum and to search for an X-ray counterpart near the well-known optical signal of 27.87 s. At around the same epoch, we carried out a similar search in ultraviolet radiation with the Faint Object Spectrograph (FOS) aboard the Hubble Space Telescope (HST). These experiments were done to test the hypothesis that the signal arises from rapid white dwarf rotation. The result generally confirmed that hypothesis, because signals were apparently detected at 27.87 s in both UV and hard X-ray light. But the detailed results were somewhat confusing, not exactly as simple as we dreamed in the proposal. The X-ray signal only emerges with fairly subtle methods of data analysis: either a "period-biased" search, or a sorting of X-ray photons into hard/soft - thus the detections are slightly marred by the taint of a posterz'ori reasoning. The UV data suffer a different but also worrisome problem: although 27.87 s is definitely detected, it's weaker than two other signals, at 28.22 and 29.69 s. That's quite surprising - we expected that the UV power spectrum would be simple, on the grounds that the UV flux distribution is nearly a pure white dwarf. Considered in isolation, such behavior from a white dwarf should be considered evidence for the pulsation hypothesis. The X-ray spectrum was interesting, mostly by virtue of what it did not show. It was essentially a pure thermal bremsstrahlung with kT = 5 keV, with only a very weak Fe K complex. This probably signifies origin near the central white dwarf and heating from accretion (propellor heating farther out would give lower densities and more cooling by emission lines). We published a paper reporting most of these details (Patterson et al. (1998), PASP, 110, 403), Much of the world accepts now that WZ Sge contains a rapidly spinning white dwarf - the fastest spin of any accreting white dwarf.