Modelling and Researching Light Pollution Risk

Abstract

With the acceleration of urbanization, the negative impact of light pollution on human production and life and biological growth is increasing. This paper develops four models to quantify the intensity of light pollution, formulate effective strategies to mitigate the negative impact, and improve human awareness. The STNLP model is developed for light pollution by firstly using AHP to determine the indicator weights and concluding that geography is a widely applicable indicator to describe the risk level of light pollution, then building the STNLP model between geography and light radiation brightness, and finally using the Erdos-Renyi model to test the effect of the STNLP model. This paper proposes three intervention strategies, builds a stepwise regression wavelet neural network model, and concludes that policy adjustments can effectively mitigate light pollution by comparing the three-dimensional scatter plots before and after the corrections.