Fraser/Abstract This proposal requests research support for optical remote sensing of twilight. Like many other researchers, we are interested in using twilight's optical structure to evaluate the verisimilitude of atmospheric scattering models. However, unlike others, we propose to do our testing with a remote sensing system that combines: a) multiple spectral channels with broad bandwidths, and b) very high spatial and temporal resolutions. This remote sensing system is human trichromatic color vision, which has the obvious disadvantage that it is essentially subjective. However, trichromatic analogues of color vision, such as color film and digital imaging systems, let us map their objective and reproducible responses into those of a typical naked- eye observer. Other advantages gained by using photography and digital imaging include the ability to easily analyze: a) spectra from large portions (or all) of the twilight at once, b) the temporal evolution of twilight spectra at very high resolutions, and c) the spatial details of twilight polarization. Specifically, we plan to use digital image analysis of color slides, occasionally corroborated by spectroradiometry, to explore the photometric, polarimetric, and colorimetric structure of twilight. With a variety of data inversion techniques, we will then assess how well various atmospheric scattering models account for twilight as seen by the naked-eye observer (we include here observers equipped with simple linear polarizers). In the past, twilight's ephemeral nature has restricted researchers to gathering only limited amounts of photometric and polarimetric data, and then usually only as a means of constructing atmospheric soundings. Colorimetric studies of twilight are even rarer. As worthwhile as twilight-derived atmospheric soundings are, here we propose a quite different goal: measuring and modeling twilight's largely unquantified visual structure.
