Identifying distinct metrics for assessing night sky brightness

Abstract

ABSTRACT Studying light pollution is an interest to researchers in a wide range of fields including astronomy, biology, civil engineering, ecology, and social science. Consequently, numerous sky brightness metrics have been developed over the years. However, what metrics are truly representative of the night sky quality and unique to the measured feature? The US National Park Service Night Skies Program has collected more than 1500 sets of night skies data throughout the United States. For each data set, a maximum of 56 metrics were measured through the combination of the captured images, Sky Quality Meter readings, and visual observations. This paper analyses these measurements and identifies a distinctive set of night sky brightness metrics based on the principal component analysis. Three major findings emerge. First, the commonly used metrics, such as the zenith brightness, horizontal illuminance, maximum vertical illuminance, all-sky light pollution ratio, Bortle class, and limiting magnitude, are highly correlated. Secondly, the observed sky brightness often offers a good estimate of the artificial light level despite the natural varying night sky background. Thirdly, a set of six metrics that consists of the zenith brightness and sky brightness percentiles are more distinctive when used to concisely describe night sky characteristics. These findings suggest that long-term night sky monitoring can be efficiently carried out by measuring the sky brightness percentiles on the observed all-sky images.