Multi-faceted light pollution modelling and its application to the decline of artificial illuminance in France

Abstract

Artificial Light At Night (ALAN) has been increasing steadily over the past century, particularly during the last decade. This leads to rising light pollution, which is known to have adverse effects on living organisms, including humans. We present a new software package to model light pollution from ground radiance measurements. The software is called Otus 3 and incorporates innovative ALAN diffusion models with different atmospheric profiles, cloud covers and urban emission functions. To date, light pollution modelling typically focused on calculating the zenith luminance of the skyglow produced by city lights. In Otus 3 we extend this and additionally model the horizontal illuminance on the ground, including the contributions from skyglow and the direct illumination. We applied Otus 3 to France using ground radiance data from the Visible Infrared Imaging Radiometer Suite (VIIRS). We calibrated our models using precise sky brightness measurements we obtained over 6 years at 139 different locations and make this dataset publicly available. We produced the first artificial illuminance map for France for the periods of 2013-2018 and 2019-2024. We found that the artificial ground illuminance in the middle of the night decreased by 23 % between these two periods, in stark contrast to the global trend.

Figures (13)

• Figure 1: Schematic representation of the Otus 3 pipeline: ground radiances are transformed to Ground Luminous Emittance (GLE), assuming a pre-defined emission function. The GLE is convolved with diffusion kernels, to calculate the Artificial Skyglow Illuminance and the Artificial Skyglow Luminance. The GLE is also used to calculate the Artificial Direct Illuminance of lamps shining directly on the ground. The natural dark-sky emission from stars and airglow are added to obtain the Total Ground Illuminance and the Total Zenith Luminance. The model parameters, such as atmospheric profile and emission function, can be adjusted for each pixel assigning different Model IDs. Artificial light maps are calculated for each model and summed, which is indicated graphically by the stacking of different maps behind each other. The procedure is repeated for each analysed time step. For more details, see the main text.
• Figure 2: Artificial skyglow luminance (dashed lines) and illuminance (straight lines) as a function of distance. Markers indicate distances at which the simulations were done with the SkyGlow software. The lines were derived from a 4$^{th}$-order polynomial spline interpolation to these simulations. The latter were done for a square 500m $\times$ 500m region with a Ground Luminous Emittance of $1~lm/m^2$. The left panel is shown on a logarithmic distance scale, while the right panels is shown in a linear distance scale to visualize the emission around zero.
• Figure 3: This figure shows the asymmetry factor of the ground radiances observed by the VIIRS-DNB instrument between 2019 and 2024. Higher values indicate larger near-nadir radiances compared to the off-nadir radiance. The black areas show regions where ground radiance was removed during the image cleaning, due to the presence of dominant city halo emission (see text for more details).
• Figure 4: Clear Sky Brightness plot showing the evolution of the median NSB in two-hour windows at a site within the International Dark Sky Reserve of Morvan, France. In total, 14 698 clear-sky measures from 127 different nights were used. Above each point the median NSB value, the number of individual measures used to compute it and the average zenithal galactic latitude are provided. The size of each point relates to the number of measures and its colour to the average galactic latitude at zenith.
• Figure 5: Measured CSB values as a function of simulated zenith brightness. The black dashed line shows the points where both are equal. The gray dashed lines show the best-fit natural sky zenith luminance of $0.214~mcd~m^{-2}$. The colour shows the mean galactic latitude of the Milky way during the CSB data taking.