Indirect imaging of the accretion stream

We apply our technique for indirect imaging of the accretion stream to the polar HU Aquarii, using eclipse prooles observed when the system was in a high accretion state. The accretion stream is relatively luminous, contributing as much or more to the overall system brightness as the accretion region on the white dwarf. We model the eclipse prooles using a model stream consisting of a ballistic trajectory from the L1 point followed by a magnetically-channeled trajectory that follows a dipole eld line out of the orbital plane. We perform model ts using two geometries: a stream that accretes onto both footpoints of the eld line, and a stream that accretes onto only the footpoint of the eld line above the orbital plane. The stream images indicate that the distribution of emission along the stream is not a simple function of the radial distance from the white dwarf. The stream is redirected by the magnetic eld of the white dwarf at a distance (1.0{1.3)10 10 cm from the white dwarf; this implies a mass transfer rate in the range 8{7610 16 g s ?1. The absorption dips in the light curve indicate that the magnetically-entrained part of the stream moves from 42 to 48 from the line of centres over the three nights of observation. This is in close agreement with the results of the one-footpoint models, suggesting that this is the more appropriate geometry for these data. The stream images show that in almost all sections of the stream, the ux peaks in B and is successively fainter in U , V and R. 1 BACKGROUND In the strongly magnetic (and thus disc-less) cata-clysmic variables, the polars (or AM Herculis stars), the accretion ow from the red dwarf secondary is controlled at large distances from the white dwarf surface by white dwarf's strong magnetic eld (10{240 MG). In these systems, the properties of the accretion ow far from the white dwarf are not well understood. This is because we lack a detailed theoretical understanding of the interactions between the dense stream plasma and the white dwarf's magnetic eld, and because emission from the stream is diicult to isolate from other sources of emission in the system, making observational investigations less straightforward. Progress has been made by studying the three main observational signatures of the accretion stream: the strong and variable UV, optical and IR emission lines (e.g.