Measuring the brightness of the night sky has become an increasingly important topic in recent years, as artificial lights and their scattering by the Earth's atmosphere continue spreading around the globe. Several instruments and techniques have been developed for this task. We give an overview of these, and discuss their strengths and limitations. The different quantities that can and should be derived when measuring the night sky brightness are discussed, as well as the procedures that have been and still need to be defined in this context. We conclude that in many situations, calibrated consumer digital cameras with fisheye lenses provide the best relation between ease-of-use and wealth of obtainable information on the night sky. While they do not obtain full spectral Quality Meter" continue to be a viable option for long-term studies of night sky brightness and for studies conducted from a moving platform. Accurate interpretation of such data requires some understanding of the colour composition of the sky light. We recommend supplementing long-term time series derived with such devices with periodic all-sky sampling by a calibrated camera system and calibrated luxmeters or luminance meters.

Artificial light at night can be harmful to the environment, and interferes with fauna and flora, star visibility, and human health. To estimate the relative impact of a lighting device, its radiant power, angular photometry and detailed spectral power distribution have to be considered. In this paper we focus on the spectral power distribution. While specific spectral characteristics can be considered harmful during the night, they can be considered advantageous during the day. As an example, while blue-rich Metal Halide lamps can be problematic for human health, star visibility and vegetation photosynthesis during the night, they can be highly appropriate during the day for plant growth and light therapy. In this paper we propose three new indices to characterize lamp spectra. These indices have been designed to allow a quick estimation of the potential impact of a lamp spectrum on melatonin suppression, photosynthesis, and star visibility. We used these new indices to compare various lighting technologies objectively. We also considered the transformation of such indices according to the propagation of light into the atmosphere as a function of distance to the observer. Among other results, we found that low pressure sodium, phosphor-converted amber light emitting diodes (LED) and LED 2700 K lamps filtered with the new Ledtech’s Equilib filter showed a lower or equivalent potential impact on melatonin suppression and star visibility in comparison to high pressure sodium lamps. Low pressure sodium, LED 5000 K-filtered and LED 2700 K-filtered lamps had a lower impact on photosynthesis than did high pressure sodium lamps. Finally, we propose these indices as new standards for the lighting industry to be used in characterizing their lighting technologies. We hope that their use will favor the design of new environmentally and health-friendly lighting technologies.

The Sky Quality Meter (SQM) has become the most common device to track the evolution of the brightness of the sky from polluted regions to first class astronomical observatories. A vast database of SQM measurements already exists for many places in the world. Unfortunately, the SQM operates over a wide spectral band and its spectral response interacts with the skys spectrum in a complex manner. This is why the optical signals are difficult to interpret when the data are recorded in regions with different sources of artificial light. The brightness of the night sky is linked in a complex way to ground-based light emissions while taking into account atmospheric-induced optical distortion as well as spectral transformation from the underlying ground surfaces. While the spectral modulation of the skys radiance has been recognized, it still remains poorly characterized and quantified. The impact of the SQMs spectral characteristics on the sky brightness measurements is here analysed for different light sources, including low and high pressure sodium lamps, PC-amber and white LEDs, metal halide, and mercury lamps. We show that a routine conversion of radiance to magnitude is difficult or rather impossible because the average wavelength depends on actual atmospheric and environment conditions, the spectrum of the source, and device specific properties. We correlate SQM readings with both the Johnson astronomical photometry bands and the human system of visual perception, assuming different lighting technologies. These findings have direct implications for the processing of SQM data and for its improvement and/or remediation.

No abstract