合肥市大雾预报方法建立与应用

利用主成分分析方法对2012～2014年合肥市高速公路沿线交通气象站日最低能见度资料统计出的大雾观测样本进行研究，应用因子荷载点聚图将合肥市县大雾划分为2个区：中南区、北区。基于PP法统计ECMWF模式输出产品与大雾之间的相关性，全市和分片分别确定与大雾密切相关的高影响因子，利用等级分类和逐步回归建立大雾预报模型。在回归结果的判定阈值和消空指标选定的情况下，通过研发的大雾天气精细化预报系统每日定时输出合肥市大雾预报格点产品。经过前期业务化运行和预报效果检验表明：数值模式产品释用方法在有无大雾预报技巧方面较WRF模式明显占优，技巧评分大幅提升，而两类典型大雾天气过程预报效果检验则可以更直观地看出数值模式产品释用的预报方法效果要更好。

Establishment and Application of Dense Fog Prediction Method in Hefei City

Using the method of Principal Component Analysis(PCA) to study the dense fog observation samples collected from daily minimum visibility data of traffic meteorological stations from 2012 to 2014 in Hefei City, by means of the factor-load scatter diagram to divide the dense fog into two zones: Central and Southern Area and Northern Area. Using PP method to calculate the correlation between ECMWF mode output data and dense fog, the meteorological factors closely related to the dense fog were selected in the whole city and region, establish the dense fog forecasting model by classification and stepwise regression. In the case of determining of regression result and emptying index, through the development of dense fog fine prediction system, daily output of Hefei City''s dense fog forecast grid products. After the previous period of the operation and forecasting results show that the method of numerical interpretative forecast is obviously superior to the WRF model in dense fog forecasting skills, and the skill score is greatly improved, test results of two typical dense foggy weather processes can be more intuitive to see that the method of numerical interpretative forecast is better.