Shouldn’t it possible to extract useful information from the sky background in the photometrically calibrated images and spectra we take ? If so there could already be a huge (untapped ?) historical source of data covering a range of passbands, locations, air mass and atmospheric conditions
Cheers
Robin