1 Introduction

In AMHerculis stars, or polars, a strongly magnetic white dwarf ( MG) accretes from a Roche-lobe filling late type secondary star. The flow of matter leaving the secondary star is threaded by the magnetic field once the magnetic pressure exceeds the ram pressure in the accretion stream. The kinetic energy is converted into heat in a shock near the footpoints of the flux tubes on the white dwarf surface. For low and intermediate local mass flow densities ( ), this shock occurs above the white dwarf surface, and the cooling thermal bremsstrahlung and cyclotron radiation can escape unhindered into space. For high mass flow densities ( $\dot m>10\,\mbox{$\rm g\;cm^{-2}\,s^{-1}$ }$), the shock is submerged in the white dwarf atmosphere, and the primary accretion luminosity is reprocessed into the soft X-ray regime.

The strong magnetic field of the white dwarf synchronises its rotation with the binary orbital period ( $\mbox{$P_{\rm spin}$ }=\mbox{$P_{\rm orb}$ }$). As an observational consequence, the emission of polars is strongly modulated at the orbital period in almost all wavelength bands. While the shape of the hard X-ray light curve of AMHerculis (and most other polars) can be easily interpreted in terms of the changing geometric aspect of the hot plasma below the shock, the optical through infrared emission may show a more complicated phase dependence, as different components within the binary contribute to the observed emission at a given wavelength, and as the cyclotron emission from the accretion column is subject to wavelength-dependent beaming.

In this paper we develop a simple quantitative model which takes into account the various emission sites in AMHerculis, and which can quantitatively describe the observed B and V band high state light curves with a minimum of free parameters.