Data analysis techniques in light pollution: A survey and taxonomy

Abstract

One of the most pressing issues facing astronomy today is the growing threat of light pollution. Light pollution affects not only astronomical observations but also sustainability in the social and environmental sense. Light pollution has been reported to cause environmental changes by altering the circadian rhythm of organisms such as birds. In this work, we conducted a systematic review of data analyses on light pollution in the literature to assist researchers and those interested in light pollution. The results of the systematic review can be divided into four distinct phases, which are research objective, data collection, data preprocessing, and data analysis. Simple popularity for each phase shows the most popular approaches are measurement as a research objective at 47.46%, ground-based sensors for data collection at 31.91%, image preprocessing at 51.61%, and statistics & machine learning for data analysis at 64.29%. The most popular combination of each phase is a measurement objective with ground-based sensors for data collection without data preprocessing or analysis. This implies that a not insignificant number of studies seek to obtain ground measurements without further analysis of the data. Data analysis as an integral part of the effort for understanding light pollution needs to be used efficiently and effectively by all stakeholders in the pursuit of sustainability.