郑州CINRAD风暴追踪产品误差统计分析

Error Statistical Analysis of Storm Tracking Production of CINRAD in Zhengzhou

利用2009—2016年5—9月郑州新一代天气雷达资料,对5615个单体回波按不同预报时效、不同生命史和不同类型强对流的风暴追踪信息产品进行统计,较系统地分析了单体回波的距离误差和方向误差,探讨了实况与预报的偏差方向。结果表明:30和60 min距离误差分别为9. 3和16. 9 km,方向误差变化不大,随着预报时效的延长,距离误差增大;生命史≤1h、1～2 h和超过2 h的30 min预报平均距离误差分别为9. 8、8. 8和7.7 km,方向误差分别为25. 2°、25. 7°和22.8°;生命史在1～2 h和超过2 h的60 min预报平均距离误差分别为17. 3和15. 9 km,方向误差为24. 4°和22. 7°,随着单体生命史的延长,方向和距离误差呈减小趋势;西北气流型、西南气流型两种类型单体回波和整体误差一致,两种类型回波单体30和60 min方向和距离误差基本相当;不同类型、不同预报时效统计均为实况偏于预报路径右侧的次数多于偏左的次数,30和60 min偏右比例分别为57. 5%和55.6%。

Based on the new generation weather radar data of Zhengzhou from May to September during 2009-2016, this paper checked the storm tracking information products of 5615 cell echoes according to different forecast time, different life cycles and different types of severe convection, processes, and also systematically analyzed the distance errors and direction errors of storm tracking information products of cell echoes and discussed the direction deviation between real condition and forecast products. The results show that the distance errors of 30 and 60 min products are 9.3 and 16.9 km respectively, and the direction error is not so big. The error of distance increases with the extension of lead time of forecasts; The average distance error of the 30 min forecast products of cell echoes with life cycles less than 1 h, 1-2 h and more than 2 h are 9.8, 8.8 and 7.7 km, and the azimuth errors are 25.5%, 25.7% and 22.8% respectively. The ave- rage distance error of the 60 min forecast products with life cycles of 1-2 h and more than 2 h are 17.3 and 15.9 km, and the direction errors are 24.4% and 29.. 7%, respectively. With the extension of life cycles of the cells, the errors of direction and distance are decreasing, the errors of northwest airflow type and the southwest airflow type are entirety the same, and the direction and distance errors of the two types of echoes are almost the same in 30 and 60 min. The errors of the southwest airflow type are slightly smal-ler than those of the northwest airflow type. The number of different types and forecast efficiency are thatthe observation data biased to the right of the forecast path are more than that of the left side, and the pro-portion of 30 and 60 min biased to right are 57.5% and 55.6%, respectively.