Abstract: Astronomical observations are increasingly limited by light pollution, which is a product of the over-illumination of the night sky. To predict both the angular distribution of scattered light and the ground-reaching radiative fluxes, a set of models has been introduced in recent decades. Two distinct numerical tools, MSNsRAu and ILLUMINA, are compared in this paper, with the aim of identifying their strengths and weaknesses. The numerical experiment comprises the simulation of spectral radiances in the region of the Canary Islands. In particular, the light fields near the Roque de los Muchachos and Teide observatories are computed under various turbidity conditions. It is shown that ILLUMINA has enhanced accuracy at low elevation angles. However, ILLUMINA is time-consuming because of the two scattering orders incorporated into the calculation scheme. Under low-turbidity conditions and for zenith angles smaller than 70° the two models agree well, and thus can be successfully applied to typical cloudless situations at the majority of observatories. MSNsRAu is well optimized for large-scale simulations. In particular, the grid size is adapted dynamically depending on the distance between a light source and a hypothetical observer. This enables rapid numerical modelling for large territories. MSNsRAu is also well suited for the mass modelling of night-sky radiances after ground-based light sources are hypothetically changed. This enables an optimum design of public lighting systems and a time-efficient evaluation of the optical effects related to different lamp spectra or different lamp distributions. ILLUMINA provides two diagnostic geographical maps to help local authorities concerned about light-pollution control. The first map allows the identification of the relative contribution of each ground element to the observed sky radiance at a given viewing angle, while the second map gives the sensitivity, basically saying how each ground element contributes per lumen installed.

Abstract: The emission function from ground-based light sources predetermines the skyglow features to a large extent, while most mathematical models that are used to predict the night sky brightness require the information on this function. The radiant intensity distribution on a clear sky is experimentally determined as a function of zenith angle using the theoretical approach published only recently in MNRAS, 439, 3405–3413. We have made the experiments in two localities in Slovakia and Mexico by means of two digital single lens reflex professional cameras operating with different lenses that limit the system's field-of-view to either 180º or 167º. The purpose of using two cameras was to identify variances between two different apertures. Images are taken at different distances from an artificial light source (a city) with intention to determine the ratio of zenith radiance relative to horizontal irradiance. Subsequently, the information on the fraction of the light radiated directly into the upward hemisphere (F) is extracted. The results show that inexpensive devices can properly identify the upward emissions with adequate reliability as long as the clear sky radiance distribution is dominated by a largest ground-based light source. Highly unstable turbidity conditions can also make the parameter F difficult to find or even impossible to retrieve. The measurements at low elevation angles should be avoided due to a potentially parasitic effect of direct light emissions from luminaires surrounding the measuring site.