多种卫星降水产品在长江流域的数据质量评价  

Evaluation of data quality for multiple satellite precipitation products in Changjiang River Basin

作　　者：余菲 沈定涛 谭德宝[3] 古洪己 肖舒婷 YU Fei;SHEN Dingtao;TAN Debao;GU Hongji;XIAO Shuting(KeyLaboratory for Geographical Process Analysis&Simulation of Hubei Province,Central China Normal University,Wuhan 430079,China;College of Urban and Environmental Sciences,Central China Normal University,Wuhan 430079,China;Changjiang River Scientific Research Institute,Changjiang Water Resources Commission,Wuhan 430010,China)

机构地区：[1]华中师范大学地理过程分析与模拟湖北省重点实验室,湖北武汉430079 [2]华中师范大学城市与环境科学学院,湖北武汉430079 [3]长江水利委员会长江科学院,湖北武汉430010

出　　处：《人民长江》2025年第2期65-74,共10页Yangtze River

基　　金：国家自然科学基金项目(42077438);中央高校基本科研业务费项目(CCNU24JC029,CCNU22QN019);云南省重大水利科技项目(2022YNYKJ10-01)。

摘　　要：准确的降水资料对防汛抗洪、水文预报以及灾害监测等至关重要。为系统分析卫星降水产品在长江流域的适用性,从降水精度特征、降水探测能力和极端降水捕捉能力3个方面对GSMaP、PERSIANN、CMORPH、TRMM、IMERG和CHIRPS共6种卫星降水产品进行综合评价,并对比分析了其在非汛期和汛期中的表现。结果表明:①在降水精度特征方面,GSMaP在非汛期和汛期均展现出最高的精度,与地面数据的一致性显著优于其他产品,其他卫星数据在长江上游区域表现不佳,与地面数据的相关性较弱;进入汛期后,尽管其他5种卫星产品在长江中下游精度有所提升,但整体相关性仍然不强,且误差显著增大。②在降水探测能力方面,GSMaP的命中率(POD值)普遍较高,尤其在洞庭湖流域表现突出,且误报率最低;IMERG和PERSIANN在特定区域表现相近,CHIRPS和TRMM的整体探测能力较弱。③在极端降水捕捉能力方面,CHIRPS、CMORPH、IMERG和TRMM普遍高估降水量,而GSMaP与地面数据更吻合;在汛期,极端降水事件显著增加,部分地区强降水总量超过2000 mm,所有卫星产品均能捕捉这些变化,但CHIRPS、CMORPH、IMERG、PERSIANN和TRMM普遍在金沙江和岷江流域的西北部低估强降水、在东南部高估强降水。研究成果可为长江流域洪水预警和水资源管理提供支撑。Accurate precipitation data are crucial for flood control,hydrological forecasting,and disaster monitoring.To systematically assess the applicability of satellite precipitation products in the Changjiang River basin,six satellite precipitation products,namely GSMaP,PERSIANN,CMORPH,TRMM,IMERG,and CHIRPS,were comprehensively evaluated from three aspects of precipitation accuracy,precipitation detection capability,and extreme precipitation capture ability,with comparison made between their performance in non-flood season and flood season.The results show that:①In terms of precipitation accuracy,GSMaP exhibits the highest accuracy in both the non-flood and flood seasons,with significantly better consistency with ground data than the other products.Other satellite data perform poorly in the upper reaches of the Changjiang River,with a weak correlation to ground data.Although the accuracy of the satellite products(except GSMaP)are improved in the middle and lower reaches of the Changjiang River during the flood season,the overall correlation remains weak,and the errors significantly increase.②In terms of precipitation detection capability,GSMaP generally has a higher detection rate(POD value),especially in Dongting Lake Basin,with the lowest false alarm rate.IMERG and PERSIANN show similar performance in specific regions,while CHIRPS and TRMM have worse overall detection capabilities.③In terms of extreme precipitation capture ability,CHIRPS,CMORPH,IMERG,and TRMM generally overestimate precipitation amounts,while GSMaP shows better agreement with ground data.During the flood season,extreme precipitation events multiply significantly,with some areas experiencing total precipitation exceeding 2000 mm.All satellite products capture these changes,but CHIRPS,CMORPH,IMERG,PERSIANN,and TRMM tend to underestimate extreme precipitation in the northwest of the Jinsha River and Minjiang River basins while overestimating it in the southeast.The findings provide valuable data support for flood early warning and water resource

关 键 词：卫星降水产品 精度评估 多尺度评价 极端降水 长江流域 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] P332.1[自动化与计算机技术—控制科学与工程]