京津冀区域重污染天气过程数值预报评估新方法

利用区域空气质量监测数据、空气质量模式数值预报产品及天气图资料,建立了一种适用于区域重污染天气过程预报的评估方法,将其用于评估NAQPMS模式系统对2013年和2014年京津冀地区静稳型、沙尘型和特殊型3类重污染天气过程的预报能力,并探讨了重污染天气过程早报、晚报及漏报的可能气象条件原因,以提高预报准确率.结果表明:数值模式系统提前3 d预报重污染天气过程的预报准确率可达57%,秋冬季预报效果好于其他季节,静稳型预报效果好于沙尘型和特殊型.对模式AQI预报结果统计发现,当预报AQI值达到150以上时,实际发生重污染天气过程的概率较大,如定义AQI等于150作为重污染天气预警临界值,模式预报准确率可提高至70%以上.天气系统对污染过程预报有重要影响,WRF气象模式对中低层天气系统位置及强度预报偏差是导致静稳型污染过程早报和晚报的一个重要原因.

An evaluation method for the operational NAQPMS numerical forecast of heavy pollution in Beijing-Tianjin-Hebei area

An evaluation method on the forecasting ability of heavy pollution in Beijing-Tianjin-Hebei area was established by using regional air quality monitoring data, numerical air quality model forecast product and meteorological data. The forecasting ability of the Nested Air Quality Prediction Modeling System (NAQPMS) on stagnant, sand dust and abnormal conditions was evaluated. In order to improve the accuracy of forecast, possible causes of meteorological conditions on the advanced forecasting, lag forecasting and missing forecasting were investigated. Results showed that three-day ahead forecast accuracy of heavy pollution can reach 57%. Forecasting ability in autumn and winter was better than in other seasons. Forecasting ability in stagnant weather was better than that in sand-dust and abnormal conditions. The statistical analysis on AQI forecast results showed that the possibility of heavy pollution increased when the predicted AQI exceeded 150. If 150 was defined as the threshold AQI value for heavy pollution warning, the prediction accuracy can be enhanced above 70%. Weather system has an important influence on the process of heavy pollution weather forecast. The forecast deviation of the position and intensity of weather system by WRF is an important reason for the advanced and lag forecasting of stagnant pollution.