Capture and Analysis of Heavy Pollution Process in an Environmental Air Quality Forecast in North China

In recent years, a series of cross-year heavy pollution episodes have occurred consecutively in the central and southern regions of North China, specifically in Henan Province and the central-western regions of Shandong Province. The uniqueness of such heavy pollution episode is closely related to the unique climate conditions that have emerged in recent years. Taking a cross-year heavy pollution episode from the end of 2022 to the beginning of 2023 as an example, the potential roles of five meteorological factors which may influence the environmental air quality during the heavy pollution episode were analyzed in the present study from the perspective of forecasting experience. Different levels of meteorological products were positioned and classified in different intervals, attempting to grasp the methods for capturing heavy pollution episode by classifying meteorological factors in North China during winter. Meanwhile, numerical simulation prediction technology were employed to simulate heavy pollution episode. The NAQPMS model was used for predicting and evaluating the accuracy, which was further validated by comparing with the accuracy of the AQI level manually corrected by forecasters. As a result, heavy pollution episode is more likely to take place when the high altitude of the area is controlled by ridge or pressure equalization fields, the ground is controlled by pressure equalization fields, the wind speed is relatively lower with the direction being not west or north, and the humidity is relatively higher and no heavy precipitation. The NAQPMS numerical model simulations on the 72 hour forecast in the present study show varying capture success rates for different provinces in North China. While the overall capture success rates for the whole region of the forecasters is relatively lower at the initial stages of the process, with only 50% for central and northern Henan, but up to 100% for southern Henan.