基于GRACE数据同化的水文模型构建方法分析与展望  

Analysis and prospects of hydrological model construction methods based on GRACE data assimilation

作　　者：罗增良 董雯雯 王伦澈 丁相毅[3] 黄河清 吴云龙 钟玉龙 LUO Zengliang;DONG Wenwen;WANG Lunche;DING Xiangyi;HUANG Heqing;WU Yunlong;ZHONG Yulong(School of Geography and Information Engineering,China University of Geosciences,Wuhan 430074,China;Engineering Research Center of Natural Resource Information Management and Digital Twin Engineering Software,Ministry of Education,Wuhan 430074,China;State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin,China Institute of Water Resources and Hydropower Research,Beijing 100038,China)

机构地区：[1]中国地质大学(武汉)地理与信息工程学院,武汉430074 [2]自然资源信息管理与数字孪生工程软件教育部工程研究中心,武汉430074 [3]中国水利水电科学研究院流域水循环模拟与调控国家重点实验室,北京100038

出　　处：《地理科学进展》2024年第11期2136-2146,共11页Progress in Geography

基　　金：国家自然科学基金项目(42201038)。

摘　　要：GRACE卫星反演的水储量信息为研究水循环过程提供了高精度数据,但如何把GRACE卫星水储量数据同化到现有分布式水文模型的径流模拟过程,已成为限制GRACE数据应用于水文模型、提高径流模拟精度的关键科学问题。论文在综述国内外研究进展的基础上,归纳总结GRACE水储量数据同化的常用方法、现有数据同化方式的原理及优缺点,以及不同数据同化方式提高水文模型精度的程度。进而,总结GRACE水储量数据同化存在的难点问题,并展望可能的破解途径。梳理结果表明,GRACE水储量数据同化面临的难点问题包括水文模型精细的建模需求与GRACE数据粗糙时空分辨率的矛盾问题,以及水文模型单独模拟水储量分量与GRACE数据表示陆表总的水储量变化的矛盾问题;综合前人在不同流域的研究结果表明,通过GRACE水储量数据同化,水文模型水储量模拟值的均方根误差减少了约5%~40%,相关性系数提高了约10%~50%,但径流模拟精度的提高幅度比较有限,相关性系数提高了约2%~16%。研究结果将为推动GRACE卫星水储量数据同化的水文模型开发及应用研究提供理论和方法参考。The water storage information retrieved by GRACE satellite provides high-precision data for the study of water cycle process.However,assimilating terrestrial water storage change data of GRACE satellite into existing distributed hydrological models for streamflow simulation has become a key scientific challenge in utilizing GRACE data to improve the accuracy of water storage and streamflow simulations.Based on a review of current research progress,this article summarized the commonly used methods for assimilating GRACE water storage change data into distributed hydrological model,the principles and advantages/disadvantages of existing data assimilation approaches,and the extent to which different data assimilation methods improve the accuracy of hydrological modeling for water storage and streamflow simulations.Furthermore,this study identified the problems associated with assimilating GRACE water storage change data into distributed hydrological models and then proposed possible solutions to these problems.The results of this study indicate that the challenges in assimilating GRACE water storage change data into distributed hydrological models include:1)the contradiction between the high spatial and temporal modeling requirements of hydrological models and the low spatial and temporal resolutions of GRACE data and 2)the discrepancy between the individual simulation of water storage components in distributed hydrological models and the representation of total terrestrial water storage changes by GRACE data.Findings from previous studies in different river basins manifest that assimilating GRACE water storage data can reduce the root mean square error of water storage simulations by approximately 5% to 40% and increase the correlation coefficient of water storage simulations by about 10% to 50%.However,the improvement in streamflow simulation accuracy is relatively limited,with an increase in the correlation coefficient of approximately 2% to 16%.This research provides theoretical and methodological references for

关 键 词：GRACE卫星水储量数据 分布式水文模型 陆表水储量 数据同化 径流模拟 

分 类 号：P334.92[天文地球—水文科学]