An all-sky light pollution model for global-scale applications that embraces a full range of cloud distributions

Abstract

Light pollution has been traditionally modeled using clear or completely overcast conditions. Usually, atmospheric conditions are more complex and involve variable cloudiness. To predict light pollution in a realistic atmosphere, we developed a model for computing the artificial night sky brightness over the sky hemisphere, considering the presence of different types of clouds and the cloud fraction. The model is applied to the city of Žilina, Slovakia, which has moderate levels of light pollution and a population of approximately 80,000. We performed simulations for various aerosol optical depths, distances from the city, cloud types, and cloud coverages (from clear to completely overcast). Results show that above the simulated city, the clouds can amplify the zenith artificial radiance by more than 15 times and the irradiance incident at ground level by more than 4 times compared to clear-sky conditions. Outside the city, however, the presence of clouds can have a screening effect, lowering the artificial zenith radiance. Additionally, an analysis performed in photopic units demonstrated that over the urban area, amplification caused by low clouds can generate amplification factors of up to 27 in zenith luminance and 17 in horizontal illuminance. These amplification factors are obtained for a moderately urbanized environment; in highly urbanized areas, even stronger amplification effects might be expected. The model can be used to explain observational data collected by light pollution monitoring networks, particularly at sites where the combination of amplifying and darkening effects of clouds generates ambiguous brightness outcomes.