高级搜索

荆州双偏振天气雷达定量降水估测精度及影响因素分析

Analysis of the accuracy and influencing factors of quantitative precipitation estimation using dual-polarization weather radar in Jingzhou

  • 摘要: 天气雷达反演的定量降水估测(Quantitative Precipitation Estimation,QPE)产品具有空间覆盖范围广、时间分辨率高的优点,对降水的监测、预报及灾害预警具有重要作用。针对2023年湖北梅雨期降水过程,以国家级自动气象站雨量计实测小时雨量为标准值,分析荆州双偏振天气雷达对地面降水的判识性能、QPE反演精度及其影响因素。结果表明:(1) 雷达QPE识别地面降水的命中率为74.7%,空报率为28%,临界成功指数为57.9%;雷达QPE识别暴雨的能力较好,中雨次之,对大雨和小雨的识别能力较差。(2) 雷达QPE和雨量计实测小时雨量具有较好的一致性,两者的相关系数、平均偏差、相对平均偏差、均方根误差分别为0.91、−0.27 mm、−13%、1.96 mm,雷达QPE整体低估地面降水;雷达QPE与雨量计小时雨量的平均偏差和均方差随雨强增大而增大,但两者的相关系数在暴雨级别时高于其他等级降水。(3) 天气雷达波束遮挡会造成雷达反射率因子低估,导致定量降水估测算法对地面降水的判识性能下降,部分区域的降水判识性能可下降20%左右;雷达QPE反演精度受反射率因子采样高度的影响相对较小,其中雷达QPE与雨量计小时雨量的平均偏差随采样高度变化不大,但随着采样高度增加,雷达QPE的离散程度和异常值出现较大的波动,容易低估地面降水。

     

    Abstract: Quantitative precipitation estimation (QPE) products derived from weather radar have the advantages of wide spatial coverage and high temporal resolution. Accurate and reliable QPE is crucial for precipitation monitoring and forecasting as well as disaster warning. Based on the data from the dual-polarization weather radar at Jingzhou station and hourly precipitation of rain gauge at national automatic weather stations (AWSs) in the Jianghan Plain during the 2023 Meiyu period, the identification performance and accuracy of dual-polarization weather radar QPE including its influencing factors is analyzed. The results are as follows. (1) the hit rate of radar QPE in identifying ground precipitation is 74.7%, the false alarm rate is 28%, and the critical success index is 57.9%. The radar QPE performs well in identifying heavy storm rain, followed by moderate rain, yet its ability to depict heavy and light rains needs to be improved. (2) The radar QPE and rain gauge have good consistency in measuring hourly precipitation, with a correlation coefficient (CORR) of 0.91, a mean error (ME) of -0.27 mm, a relative mean error (RME) of -13%, and a root mean square error (RMSE) of 1.96 mm. As a whole, the radar QPE underestimates ground precipitation. The ME and RMSE of hourly precipitation between radar QPE and rain gauge increase with increasing rainfall intensity, but the CORR between them is higher in heavy storm rain than in other grades of precipitation. (3) Radar beam blockage can lead to underestimation of radar reflectivity, which in turn reduces the performance of the quantitative precipitation estimation algorithm in detecting surface precipitation, with the identifying performance in some areas decreasing by about 20%. Additionally, the radar QPE accuracy is relatively less affected by the sampling height of the reflectivity factor, and the ME of hourly precipitation between radar QPE and rain gauge presents little change with varying sampling height. However, as the sampling height increases, the dispersion and outliers of radar QPE fluctuate greatly, making it easy to underestimate ground precipitation.

     

/

返回文章
返回