Abstract: Based on observational data of sunrise and meteorological data, such as cloud cover, visibility, and sunshine duration collected from the National Meteorological Station in Longquanyi District of Chengdu from December 2022 to November 2023, the sunrise conditions were divided into three levels: the sky on the bright day, the sun in the cloud and the cloud blocking day. Through correlation analysis, the main meteorological factors influencing sunrise visibility and their corresponding weights are identified, and the prediction model of sunrise index is established. The results show that the main meteorological factors affecting sunrise are total cloud cover, low cloud cover and horizontal visibility, and there are seasonal differences in the degree of influence, so the factor weights must be adjusted accordingly. The prediction model shows high prediction accuracy in practical application. The correlation coefficient between the prediction results and the actual index is 0.835, which is statistically significant at the 0.01 level.