摘要
"灾害天气短时临近预报系统"(SWAN)在我国气象预报业务中已得到广泛应用。文章介绍其中雷达定量降水估算QPE算法(RASIM方法)的技术与特点,选取湖北省6部S波段多普勒天气雷达在2012年探测的暴雨天气过程资料,系统性评估了SWAN系统中QPE产品实用性,初步分析了产生估算误差的原因。评估表明:整体而言该方法在湖北省使用效果较好,平均绝对误差率小于30%;探测距离的增加对S波段雷达QPE精度影响不大;各雷达对30mm以上降水的估算平均绝对误差率较小,但估算结果较实况偏弱,随着雨量(雨强)增大,低估的比例也增大。就单部雷达而言,宜昌雷达估算降水误差最大,武汉雷达受附近建筑遮挡影响次之,恩施雷达估算降水误差最小。
The Severe Weather Automatically Nowcast System （SWAN） has been widely used in weather nowcasting operation. The technology and features of radar quantitative precipitation estimation （QPE） are introduced. The practicability of QPE is evaluated, and the causation of QPE error is preliminarily analyzed, selecting the data during rainstorm weather from 6 S-band Doppler weather radar sets in 2012 in Hubei Province. The results show that the mean absolute error rate （MAER） is below 30% overall; the increasing range of radar has little influence on QPE precision; the MAER is keeping small in above 30 mm cases, while QPE is underestimated comparing with rain gauge records; moreover, the proportion of the underestimated is increasing in heavier rainfall. Concerning a single radar set, the worst is in Yichang due to the radar beam blocking of mountains, and worse in Wuhan due to the radar beam blocking of buildings, and the best is in Enshi.
出处
《气象科技》
北大核心
2015年第3期380-386,共7页
Meteorological Science and Technology
基金
公益性行业专项GYHY201306008
GYHY201306078
湖北省气象局科技发展基金项目2013Q03资助
