The emission factor method is often used to estimate the atmospheric pollution intensity in the past, in which exists a few defects such as the lack of emission source category, the complex computing and the suspicious atmospheric emission data sources on account of social enterprises' stealing pollutant emission. This study established the simple and rapid-computing atmospheric pollution emission intensity estimation model by using all urban areas' air quality data under high levels of pollution weather. Tyson polygon is adopted to define the regional characteristic length and air box model estimation method is used to determine the source of emissions of pollutants at a high rate of computing speed in this study. Simulation results indicated that the estimated emission intensity of NO_x, SO₂ and CO in the study area is 436.33, 154.82 and 8 566.82 t/d. According to the model accuracy test, the relative error of intermediate value is below 30% which indicates that the atmospheric emission intensity Algorithm can be used in simulation and prediction of regional atmospheric pollutants emission.