Based on the forecasting products of Community Multiscale Air Quality Model (CMAQ), observation of air pollutants and the conventional ground meteorological data from January 2007 to June 2010 in Fuzhou City, Fujian Province, China, the models were developed to forecast daily air pollutant concentration for various weather systems by statistical-dynamic forecast method and multivariate linear stepwise regression. The models featured both CMAQ forecast products and various other forecast parameters. The results showed that there were 7 weather systems influencing Fuzhou City: continental high, subtropical high, shear, warm sectors convergence, upper trough, typhoon and tropical convergence. The air quality was poor under the control of warm sectors convergence, upper trough or continental high weather systems. In contrast, when Fuzhou was under the control of subtropical high or typhoon, the air quality was better. The p-value associated with the forecast functions of air pollutant level was 0.000, so the models were statistically significant. The models were tested on the air pollutant data in Fuzhou from June to December 2010 by back substitution. The forecast accuracy of the models on contamination index level of PM10 reached 71.3%, while the forecast accuracy on SO2andNO2 were 100%. The comprehensive score of daily air quality forecast in the city of Fuzhou was 88.8 points on average.%利用CMAQ(Community Multiscale Air Quality Model)模式预报产品和福州市2007年1月至2010年6月大气污染物的观测资料以及常规地面气象观测资料,根据动力统计相结合的预报方法,通过多元线性逐步回归,建立不同天气系统下CMAQ模式产品和多类预报因子相结合的日污染物浓度预报模型.结果表明,影响福州市的天气系统共分为大陆高压、副热带高压、切变、暖区辐合、高空槽、台风和热带辐合带7类天气型.在暖区辐合、高空槽和大陆高压控制下,福州市的空气质量较差,而副热带高压和台风系统影响时,福州市的空气质量最好.日污染物浓度预报方程置信度均为P=0.000,模型有统计学意义.利用模型对2010年7~12月福州市各污染物浓度进行预报效果回代检验,模型对PM10的污染指数等级预报正确率达到了71.3%,对SO2和NO2的级别预报正确率达到了100%,日预报综合评分平均达88.8分.
展开▼