Real-time anomaly detection in sky quality meter data using probabilistic exponential weighted moving average

Light pollution is a problem that impacts many elements of human life and the environment, including astronomical observations. The authors of this work offer a unique method for detecting anomalies in night sky brightness data recorded using a Sky Quality Meter (SQM). This equipment has been widely utilized in light pollution research worldwide, yielding massive data. However, there is the possibility of experiencing abnormalities or outliers throughout the data collection process due to natural occurrences or measurement errors. This study uses the probabilistic exponential weighted moving average algorithm to find anomalies in SQM data received from Timau Observatory by simulating the streaming procedure on SQM data using Apache Kafka technology. Finally, this study intends to shed fresh knowledge on night sky brightness and light pollution dynamics. The authors could locate and analyze unusual or suspicious phenomena that had previously gone unreported using the anomaly detection approach. These findings can help us better understand light pollution and its environmental and human life effects. Still, they can also help us establish strategies and policies that will reduce light pollution in the future. Furthermore, this work illustrates the potential of anomaly detection as a powerful tool for data analysis in various domains, encouraging the use of this approach in future research.