Abstract

Although it is known since at least 1905¹⁾ that the skylight polarization changes drastically when the totality of a solar eclipse starts, it lasted till 1961 till this effect was recorded instrumentally ²⁾. Since then, only a few more instrumental observations have become available ^3-7). However, the information obtained prior to 1973 is too limited for getting a quantitative understanding of the phenomenon, since the observers restricted their measurements to one single point in the sky, located at 90° from the sun in the solar vertical ^2-6). This situation changed in 1973, when Shaw recorded a very complete set of sky radiance observations during the 30 June eclipse, which is still of an unmatched quality ⁷⁾. His measurements included also a polarization scan in the solar vertical, so that we got for the first time a general impression of the state of polarization at eclipse conditions. It turned out that the degree of polarization during eclipse has a symmetry with respect to the zenith, with a minimum value in the zenith and a maximum polarization rather close to the horizon. This outcome fascinated me for years and I wondered if it is possible to calculate the polarization at eclipse conditions from Shaws on-side intensity measurements and his polarization measurements outside totality by means of a simple two-step scattering model. The results of an attempt for this are rather encouraging and presented in this lecture.

© 1986 Optical Society of America