The approach to stray light suppression in the UVCS is based on the design of a rocket-borne UV coronagraph [Kohl et al. (1978), Kohl et al. (1980)]. In the following, an analysis of the stray light properties of the UVCS is described [see Romoli et al (1993) for a more detailed description of the analysis for the white light channel]. The analysis is based, in part, on measurements of component characteristics. It is shown that internal and external occulters, light traps, baffles, and telescope mirrors are expected to keep the stray light on the entrance slits at an acceptable level.
The problem is well defined in Figure 13 , where the radiances of the solar disk and corona in units of the Sun-center radiance are plotted as a function of the heliocentric height. The HI Ly- radiance in a coronal hole [Strachan (1990)] is plotted, as well as the combined contributions of the visible K and F coronae in a quiet region [Allen (1964)]. The comparison shows that the required stray light suppression in the visible light range is greater than in the UV.
In a coronagraph the solar disk radiation is the source of the stray light. For convenience, the stray light can be divided into two major components: that resulting from radiation that enters the instrument directly, and that which is diffracted by the edges of the entrance aperture. The first contribution can be reduced with a sunlight trap which would ideally absorb or reflect away 100% of this radiation. The second contribution can be reduced by configuring the optical geometry to prevent the light from reaching the detection system. In the latter case, this must be done at very little expense of coronal radiation (in the first coronagraph the Lyot stop performed this job). The geometry of the UVCS occulted telescope has been designed in such a way that the level of stray light at the spectrometer entrance slits is lower than most coronal signals planned to be detected. Hence, the telescope environment provides the suppression. It is expected that the primary components of stray light at the entrance slit have very little or no polarization. An effective reduction in stray light is then introduced in the WLC when light is passed through the polarimeter and the polarized light component is determined. In addition, the FOV of the polarimeter has been designed to not intercept some of the potentially significant sources of stray light, such as that from the outer edges of the entrance slit baffle, which is suppressed by a second entrance slit baffle.
Hence, the analysis of the stray light described here has been divided, primarily, into two categories: the stray light involving scattering off the sunlight trap and the stray light due to diffraction off the external occulter. Table VI summarizes the sources of stray light that are classified in the two categories plus four other sources.
Figure 14 summarizes the results obtained. The plot shows the expected stray light irradiance level on the slit in units of solar disk irradiance versus the heliocentric height for each analyzed contribution of stray light and their total, compared with the estimated signal irradiance at the entrance slit. The ordinate gives the irradiance in units of photons cm^-2, normalized to the expected solar irradiance at 1 AU. Figure 14a refers to the analysis for HI Ly-, and Figure 14b refers to the analysis for the visible broadband light.
It is clear from Figure 14a that the major contribution to stray light at Ly- is expected to come from non-specular reflections off the mirror surface (the primary contribution was originally diffracted by the primary occulter). For visible light, two contributions are dominant. At lower heliocentric heights () scatter off the internal occulter is dominant, while above the contribution from the mirror surface prevails. For the WLC, the effect of the stray light contribution is to increase the statistical error of the polarized radiance derived from the measurement.