Analysis of sky luminance experimental data and comparison with calculation methods

Abstract

Data of sky luminance measured in Osaka (Japan) and in Arcavacata di Rende (Italy) during a year were analysed and compared with the values predicted by the Perez, Igawa and CIE calculation methods. The best predictions of absolute, relative and zenith luminances were obtained by the CIE method. If the measured luminances in a locality are not available, only by the Perez and the Igawa method is possible to predict the sky luminance distribution under different meteorological conditions. Of these two methods, the Igawa method appears to be more accurate for the prediction of the absolute luminance, and the Perez method for the predictions of the relative luminance. It should be stressed that, before comparison with the calculation models, the sky luminances measured in each scanning have to be corrected in order to make the horizontal illuminance obtained by integration of the luminances on the sky vault equal to the value of the horizontal illuminance measured by photometers. If such a correction is not made owing to the absence of this latter measurement, the errors of the predictions obtained by the models become larger, but they can be reduced by applying a new methodology proposed by the authors.