兰州市冬季霾的统计预报研究

Research on Statistical Forecast of Haze in Winter in Lanzhou City

利用污染扩散方程,在分析了能见度与空气污染浓度之间关系的基础上,建立了兰州市冬季霾的统计预报模型。此预报模型既考虑了污染物排放量和污染浓度的作用,又考虑了气象条件的影响。利用兰州市2003—2008年冬季的能见度资料和同期兰州气象站的地面气象要素、天气形势和T213数值产品,分析了霾的出现与气象条件的定量关系,发现哈密、敦煌、酒泉3站平均海平面气压与兰州市之间气压差增大或是兰州附近的气压梯度增加时,有利于空气污染物的扩散,发生霾的可能性较低;反之,容易出现霾。此外,通过逐步回归,最终确定了冬季兰州市霾的24 h预报方程。使用2009年数据进行试预报检验,结果表明,此预报方法方便实用,具有较好的预报能力,可以用于兰州市霾的预报业务。

Haze is a kind of weather phenomenon when dust, smoke and other dry particles obscure the clarity of sky, and its occurrence is caused not only by air pollution but also by weather conditions. The authors first analyzed the relationship between air visibility and pollutant concentration, and then established the statistical forest model for winter haze in Lanzhou City by simplifying the equation of air pollutant diffusing. The effects of meteorological element, air pollution and visibility were all taken into consideration in this model, so its physical foundation may be more believable than pure statistical models. The relationship be- tween haze and weather conditions in Lanzhou City was analyzed by use of the visibility data, surface and lower layer meteorological element from the Lanzhou Meteorological Station, and the T213 numerical prediction products and the weather map during 2003-2008. The results showed that the reduced air pressure field are favorable for the occurrence of haze in Lanzhou. Finally, a 24-hour haze forecast equation for the Lanzhou City was derived from stepwise regression. This equation was verified by the observation data in 2009, which proved the forecasting results are credible.