智能网格实况融合分析产品在武汉军运会中的应用检验分析

Analysis and Assessment on the Application of Intelligent Gridded Real-time Fusion Analysis Products During the Wuhan Military Games

为检验智能网格实况融合分析产品在武汉军运会的分析应用效果,以武汉市150个地面气象自动站逐小时观测数据为基础,与2019年9-11月“全国智能网格实况融合分析产品(V1.0)”2 m气温、相对湿度、10 m风速、降水量和能见度进行对比分析。结果表明:(1)气温数据的相关系数为0.972~0.988,ME、MAE和RMSE分别为-0.026~0.166℃、0.578~0.732℃和0.841~1.067℃,实况分析产品与地面观测数据基本一致。(2)相对湿度数据的相关系数为0.891~0.938,ME、MAE和RMSE分别为-0.433%~0.559%、4.843%~5.172%和6.472%~7.318%,能很好地反映全市的月相对湿度变化。(3)风速数据的相关系数为0.541~0.687,ME、MAE和RMSE分别为-0.492~-0.303 m/s、0.839~0.913 m/s和1.129~1.263 m/s,实况分析产品月平均风速偏小。(4)能见度数据的相关系数为0.937~0.967,ME、MAE和RMSE分别为0.793~1.029 km、1.485~2.182 km和2.001~2.937 km。实况分析产品能见度数据在无雾时与地面观测结果非常接近,但实况分析产品月平均能见度偏大。(5)降水量数据的相关系数为0.489~0.886,ME、MAE和RMSE分别为0.001 mm左右、0.005~0.030 mm和0.082~0.206 mm。晴雨准确率高,随着降水量级增大,漏报率和空报率增大,大量级降水漏报明显。实况分析产品降水空间分布接近地面观测结果,但削弱了降水的强度,降水极值和区域月降水量偏小。

To test the application effect of intelligent gridded real-time fusion analysis products in the 2019 Wuhan Military Games,based on the hourly observation data from 150 surface automatic weather stations(AWS)in Wuhan,this paper compares and analyzes the 2 m temperature,relative humidity,10 m wind,precipitation and visibility with the“National Intelligent Grid Real-time Fusion Analysis Product(V1.0)”from September to November 2019.The results show that:(1)The correlation coefficient of 2 m temperature is between 0.972 and 0.988,and the ME,MAE and RMSE of temperature are-0.026-0.166℃,0.578-0.732℃ and 0.841-1.067℃respectively,which means that the gridded 2 m temperature is basically consistent with the surface observation data.(2)The correlation coefficient of relative humidity is 0.891-0.938,and the ME,MAE and RMSE of relative humidity are-0.433%-0.559%,4.843%-5.172%and 6.472%-7.318%respectively,which can well reflect the monthly relative humidity changes in the whole city.(3)The correlation coefficient of 10 m wind speed is 0.541-0.687,and the ME,MAE and RMSE of wind speed are-0.492 to-0.303 m/s,0.839-0.913m/s,and 1.129-1.263 m/s respectively.Relatively,the monthly average wind speed of gridded data is smaller.(4)The correlation coefficient of visibility is 0.937-0.967.The ME,MAE and RMSE of visibility are 0.793-1.029 km,1.485-2.182 km and 2.001-2.937 km respectively.The gridded visibility is very close to the observed data when there is no fog,but the monthly average visibility of gridded data is larger.(5)The correlation coefficient of precipitation is 0.489-0.886,and the ME,MAE and RMSE of gridded precipitation are about 0.001mm,0.005-0.030 mm and 0.082-0.206 mm respectively.The verification results of precipitation show that the gridded data performs well for sunny and rainy weather.But with the increase of precipitation level,the rates of missing forecast and false alarm forecast are elevated,with precipitation at higher grades obviously missed.The spatial distribution of the gridded precipitation is close to the observed data,but the intensity of precipitation is weakened,and the precipitation extreme value and regional monthly precipitation are relatively small.