近实时卫星降水数据对南京“20170610”极端性强降水过程的监测分析

Monitoring and Analysis of the Extreme Heavy Rainfall Process on June 10,2017 in Nanjing Using Five Near Real Time Satellite Rainfall Estimations

以高密度的地面雨量站网观测数据为基准,综合分析了TRMM 3B42RT V7、IMERG Early、IMERG Late、GSMaP NRT、GSMaP NRT Gauge共5种近实时卫星降水数据对南京及周边地区2017年6月10日一次破历史记录的极端性强降水过程的监测能力。结果发现,在累积雨量方面,2种GSM aP数据有比较明显的低估,但基本再现了雨量中部高、南北低的空间分布格局;而3B42RT V7和2种IMERG对累积雨量的主要空间分布特征的辨识性较差。在降水强度时序变化方面,所有卫星数据均能正确探测到南京全市与江宁区的"20170610"降水过程,但在区域和网格尺度上对降水强度变化的动态跟踪能力明显不足,定量误差比较突出。5种数据中,GSMaP NRT的综合精度相对较高,而2种IMERG数据的表现尚不及3B42RT V7。总体上,近实时卫星降水数据对中小尺度极端性强降水过程的监测已展现了一定的积极效果,但捕捉降水落区及追踪降水动态变化的能力尚需大力改进。

Near real time satellite rainfall estimation （NRT-SRE） with high spatial-temporal resolution and time- liness provides an effective approach for monitoring extreme rainfall events, while its performance determines the reliability and applicability of the data. This paper comprehensively evaluated the capabilities of five NRT-SRE （including TRMM 3B42 V7, IMERG Early, IMERG Late, GSMaP NRT, GSMaP NRT Gauge） against a dense rain gauge network, for monitoring the record-broken extreme rainfall process on June 10, 2017 over Nanjing and the surrounding areas. The results demonstrated that, although two kinds of GSMaP data present notable un- derestimation of the accumulated rainfall, they basically capture the spatial pattern in which the cumulative rain- fall is heavy in the central region while light in the northern and southern regions. However, 3B42RT V7 and two IMERG data showed poorer ability to identify the main spatial distribution characteristics of cumulative rainfall. Regarding to temporal variation of rainfall intensity, all NRT-SRE can correctly detect the extremely heavy rain- fall process over Nanjing City and Jiangning District, but the dynamic tracking ability is obviously unsatisfactory with severe significant quantitative errors at the regional and grid scales. Comprehensive accuracy of GSMaP NRT is relatively higher among five NRT-SRE, while the performance of IMERG data is still poorer than 3B42RT V7. Overall, the NRT-SRE has exhibited positive performance on monitoring extremely heavy rainfall process at me- dium and small space scales, whereas their capabilities to capture precipitation falling areas and track temporal fluctuation of rainfall intensity need to be enhanced in practical applications.