Light pollution refers to the excessive or improper use of artificial light. Global light pollution poses environmental and health problems. In order to measure and put forward reasonable suggestions to alleviate these phenomena, we established a light pollution evaluation model to calculate the risk degree of light pollution in a region, and sensitively put forward improvement suggestions to alleviate the degree of local light pollution according to the severity of local indicators. In order to strengthen the correlation between the indicators and the evaluation objectives in the evaluation model, we constructed the relationship between the degree of economic development and the night light data. Then, we analyzed the light pollution evaluation model from six perspectives: GDP, population density, forest area, terrain, temperature and precipitation. Combined with the data over the years, we made a preliminary prediction of the weights of each index, then carried out a consistency test on the weights of each index to judge their rationality, and finally conducted a sensitivity analysis.