The task of smog control in China is still arduous compared with developed economies. To address the problems associated with the uncertain of smog pollution, this paper establishes a new EGM (even grey model) (1,1) prediction model based on the kernel and degree of greyness under the condition that the possibility function is known. The original EGM(1,1) prediction model based on the interval grey number sequence is constructed under circumstances where the possibility function is unknown. For testing the proposed model, the daily AQI data of Nanjing, Jiangsu Province, China, was selected. Also, this paper uses another five forecasting models and compares the results with the new model. The results show that the daily AQI index of Nanjing presents a significant downward trend excluding seasonal factors, and the prediction accuracy of the new EGM (1,1) model is also higher than that of other forecasting models.