MOS方法在机场气象要素客观预报中的应用

Application of Model Output Statistics into Objective Element Forecast at Airport

利用WRF中尺度数值模式和模式输出统计(MOS)方法,研究建立乌鲁木齐机场逐时温度、相对湿度的回归预报模型,并尝试针对冬季低云、低能见度等天气建立分类预报模型,通过对统计模型的检验可以看到:逐时温度绝对差为1.09～2.33℃;逐时相对湿度绝对差为4.7%～9.6%;11月至翌年2月低云量>5分量分类预报准确率为76.94%～83.13%,TS评分54.27%～66.50%;11月至翌年2月能见度≤800 m的分类预报准确率为89.83%～92.04%,TS评分为29.09%～46.05%。该方法预测效果较好,因此可以尝试使用本方法为日后航空气象业务提供机场客观预报指导产品。

Using WRF mesoscale model and the Model Output Statistics（MOS） method,the Urumqi Airport’s regression prediction model for the hourly temperature and relative humidity and classification prediction model for the winter low level clouds and low visibility weather are established.The results show that the absolute error of hourly temperature is 1.09 ℃～2.33 ℃ and Hourly relative humidity’s absolute error is 4.7% ～9.6%;from November to February,classifier forecast accuracy rate for the low level clouds cover5 component is 76.94%～83.13%,and the TS score is 54.27%～66.50%;the classification prediction accuracy rate for visibility≤800 m is 89.83% ～92.04%,and the TS score is 29.09%～46.05%.We can see MOS method predictsis better than the WRF products,so we try to use this method to provide the airport objective forecast guidance products for aviation meteorological services in the future.