Photometric Variations as Small Perturbations in Aerosol Content

The quality of profile fitting of resolved stars depends ultimately upon the accuracy with which spectral differences of the sources are retrievable within the data, because the radiation color of well-separated known sources can serve as an indicator of the origin of the optical depth variations one observes during the night. The particularities of the whole sky imager (WSI) detector and optical system are such that the data suffer from lack of the spatial resolution required in a common astronomical observation. We construct an empirical point spread function (PSF) produced by the optical system and detector, and fit it to the data, frame by frame. Star radiance extracted with profile fitting proves to be more accurate than that from simple aperture photometry. A bonus of the method is the possibility to construct “flat-fields”, i.e., an average of frames with stars extracted and inter-pixel variations smoothed. Aperture Photometry vs. Profile Fitting Extinction calculation from starlight measurements depends crucially on the precision of inferring the star magnitude from sampled (pixelized) images. Two main methods of star photometry (or a combination of them) are frequently used: (1) aperture photometry, and (2) profile fitting. Aperture photometry assumes that a star brightness is exactly measured as a positive signal above the surrounding sky brightness. One chooses a number of pixels (aperture) around the known star position, and deduct from their summed intensity a mean value of the surrounding “sky” which is measured in an annulus centered on the star, but is not influenced by star’s wings. Caveat: “Evaluating the diffuse component of the observed brightness at a given position in a crowded star field is extremely difficult. There are complexities concealed within the word <>. Even the meaning of the phrase “diffuse