基于雷达组网拼图的定量降水估测算法业务应用及效果评估  

Operational application and evaluation of the quantitative precipitation estimates algorithm based on the multi-radar mosaic

作　　者：勾亚彬[1,2,3] 刘黎平[1] 杨杰[4] 吴翀[4] 

机构地区：[1]中国气象科学研究院灾害天气重点实验室,北京100081 [2]中国科学院大学,北京100049 [3]杭州市气象局,杭州310051 [4]成都信息工程学院,成都610225

出　　处：《气象学报》2014年第4期731-748,共18页Acta Meteorologica Sinica

基　　金：973项目(2012CB417202);气象行业专项(GYHY201306004);国家自然基金项目(41175038;41205078)

摘　　要：基于雷达组网实时的定量降水估测(QPE)及实时评估系统在浙江省杭州市气象局成功实现了业务应用,在评估雷达定量降水估测业务应用效果的同时,根据雷达反射率因子垂直廓线(VPR)特征,探讨分析了不同类型降水过程中雷达定量降水估测的误差源。系统联合杭州、宁波、舟山、温州、金华及衢州6部新一代天气雷达的基数据资料,以及覆盖浙江省且经反距离加权(Inverse Distance Weights,IDW)法实时质量控制的雨量计观测资料,采用先雷达组网拼图再降水估测的方案,集成Z-R关系法和最优插值法反演与校准雷达定量降水估测数据场。4次不同类型降水过程的评估结果表明:(1)在地物遮挡严重的浙江西北部和雷达覆盖较差的浙江南部,降水估测的雷达反射率因子如果源于0℃层亮带,会导致雷达定量降水估测严重高估;如果源于浅薄层云云系的云顶,会造成雷达定量降水估测严重低估。(2)多种降水类型云系并存,但使用相对单一的Z-R关系,会导致梅雨和台风期间雷达定量降水估测的局部高估或低估。(3)伴随飑线系统的强对流以及台风系统的非对称性也是导致雷达定量降水估测误差的重要原因。(4)联合Z-R关系和最优插值法,有效地降低了雷达定量降水估测的系统误差,但仍然存在大量的局部误差。The real-time system of QPE （Quantitative Precipitation Estimates） together with the real-time evaluation based on the multi-radar mosaic has been successfully implemented in the operation at the Weather Bureau of Hangzhou in Zhejiang Prov- ince. A comprehensive evaluation is presented in this paper and meanwhile the error sources are analyzed according to the char- acteristics of vertical profile of radar reflectivity using the four different precipitation events. The system unites the basic data from the six CINRAD （China New Generation of Weather Radar） at Hangzhou, Ningbo, Zhoushan, Jinhua and Quzhou and the gauge data within Zhejiang Province which are quality-controlled in real time by the IDW （Inverse Distance Weights） meth- od, and in order to reduce systematic and local errors, both the dynamic Z-R relationship and optimal interpolation method are integrated into the system to calibrate the QPE field. From the verification and analysis of the four events, some conclusions are as follows.. （1） In northwestern and the southwestern Zhejiang, if the reflectivity sources from the bright-band （the top of the cloud）, it results in the overestimation （underestimation） of radar QPE of the stratiform rainfall system. （2） In the case of the coexistence of different rainfall types the use of relative uniform Z-R relationship will result in serious local overestimation and underestimation of QPE during the Meiyu front/typhoon. （3） The severe convection with the squall line and the asymme- try of typhoon are important reasons of the error in the radar QPE. And, （4） although the combination of the Z-R relationship with the optimal interpolation can effectively reduce the systematic error, some large local error remains in the QPE.

关 键 词：雷达三维组网拼图 定量降水估测 实时业务应用及效果评估 

分 类 号：P414[天文地球—大气科学及气象学]