The best way to conference proceedings by Francis Academic Press

Web of Proceedings - Francis Academic Press
Web of Proceedings - Francis Academic Press

Evaluation and intervention of light pollution risk level

Download as PDF

DOI: 10.25236/icmmct.2023.009


Xu Chuang, Li Mengchen, Liu Biao

Corresponding Author

Li Mengchen


With the prosperity of cities, problems such as global light pollution are becoming more prominent, hence the need for scientific monitoring and assessment. This paper mainly studies the comprehensive evaluation index and intervention strategy of light pollution, and establishes the evaluation and prediction model based on TOPSIS and wavelet neural analysis. This paper first defines the concept of light pollution risk level. Secondly, based on the ERA5 data, the TOPSIS model optimized by AHP and entropy weight method is used for the preliminary assessment of the risk level. Then, based on the night remote sensing data of Luojia No. 1 satellite, the VIIR map was deeply analyzed by using graph theory algorithm, and the evaluation index was further optimized. Finally, an evaluation system based on luminosity, radiation, spectrum, perspective, space time and region is established. At the same time, for the prevention and control of light pollution, based on the established index system, this paper adopts the stepwise regression analysis algorithm to obtain the factors significantly related to the risk level of light pollution, and input them into the wavelet neural network. Since different intervention measures correspond to different parameters, this paper predicts the change of light pollution risk level under different measures by modifying the parameters. Further development of optimal intervention models.The results showed that the risk levels of light pollution were urban community, suburban area, rural area and protected area. In particular, over the past few decades, regions such as Europe have suffered higher levels of light pollution, mainly between 30° and 60° north latitude.


Light pollution, Topsis; Neural network; Graph theory; Stepwise regression analysis