HUBBLE-SPACE-TELESCOPE OBSERVATIONS OF THE DWARF NOVA Z-CHAMAELEONTISTHROUGH 2 ERUPTION CYCLES

We have obtained the first high-speed photometry of the eclipsing dwar f nova Z Cha at ultraviolet wavelengths with the Hubble Space Telescop e. We observed the eclipse roughly every 4 days over two cycles of the normal eruptions of Z Cha, giving a uniquely complete coverage of its outburst cycle. The accretion disk dominated the ultraviolet light cu rve of Z Cha at the peak of an eruption; the white dwarf, the bright s pot on the edge of the disk, and the boundary layer were all invisible . We were able to obtain an axisymmetric map of the accretion disk at this time only by adopting a flared disk with an opening angle of simi lar to 8 degrees. The run of brightness temperature with radius in the disk at the peak of the eruption was too flat to be consistent with a steady state, optically thick accretion disk. The local rate of mass flow through the disk was similar to 5 x 10(-10) M. yr(-1) near the ce nter of the disk and similar to 5 x 10(-9) M. yr(-1) near the outer ed ge. The white dwarf, the accretion disk, and the boundary layer were a ll significant contributors to the ultraviolet flux on the descending branches of the eruptions. The temperature of the white dwarf during d ecline was 18,300 K < T-wd < 21,800 K, which is significantly greater than at minimum light. Six days after the maximum of an eruption Z Cha had faded to near minimum light at ultraviolet wavelengths, but was s till similar to 70% brighter than at minimum light in the B band. Abou t one-quarter of the excess flux in the B band came from the accretion disk. Thus, the accretion disk faded and became invisible at ultravio let wavelengths before it faded at optical wavelengths. The disk did, however, remain optically thick and obscured the lower half of the whi te dwarf at ultraviolet and possibly at optical wavelengths for 2 week s after the eruption ended. A bright region, which we identify with th e boundary layer, was visible at the equator of the white dwarf in the B band. The upper limit to the temperature of the boundary layer is 1 8,200 K, and at this low temperature it was a minor contributor to the ultraviolet flux. It is possible to explain the invisibility of the b oundary layer at both maximum and minimum light, while still allowing it to be visible on the descending branch of eruptions, by invoking a rapidly rotating white dwarf. The equatorial velocity of the white dwa rf must be greater than similar to 0.23 times the Keplerian velocity a t the surface of a spherical white dwarf. The bright spot became clear ly visible at ultraviolet wavelengths immediately after the end of the eruptions. Its distance from the white dwarf and thus the radius of t he accretion disk was R(D)/R(L1) similar to 0.56. The temperature of t he white dwarf was somewhat elevated but was not greater than 17,500 K . By the third week after eruptions the eclipse looked like a simple o ccultation of an unobscured, spherical white dwarf by a dark secondary star. The center of the accretion disk was, therefore, optically thin at ultraviolet wavelengths and the boundary layer was too faint to be visible. The distance of the bright spot from the white dwarf and hen ce the radius of the disk decreased slowly through minimum light, reac hing R(D)/R(L1) similar to 0.47 just before the first eruption and sim ilar to 0.52 just before the second. The temperature and ultraviolet f lux from the bright spot increased from the end of one eruption to the beginning of the next, its temperature reaching 16,700 +/- 600 K at t he end of quiescence. The temperature of the white dwarf in quiescence was 15,700 +/- 550 K; if it is fully limb darkened, its radius was R( wd)/R(L1) = 0.0254 +/- 0.0013, and if it is not limb darkened, its rad ius was R(wd)/R(L1) = 0.0222 +/- 0.0013.