Time-series surface water reconstruction method(TSWR)based on spatial distance relationship of multi-stage water boundaries  

作　　者：Mingyang Li Shanlong Lu Cong Du Yong Wang Chun Fang Xinru Li Hailong Tang Muhammad Hasan Ali Baig Harrison Odion Ikhumhen 

机构地区：[1]International Research Center of Big Data for Sustainable Development Goals,Key Laboratory of Digital Earth Science,Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing,People’s Republic of China [2]College of Resources and Environment,University of Chinese Academy of Sciences,Beijing,People’s Republic of China [3]School of Earth Sciences and Spatial Information Engineering,Hunan University of Science and Technology,Xiangtan,People’s Republic of China [4]School of Geospatial Engineering and Science,Sun Yat-Sen University,Zhuhai,People’s Republic of China [5]Institute of Geo-Information & Earth-Observation (IGEO),PMAS Arid Agriculture University Rawalpindi,Rawalpindi,Pakistan [6]Key Laboratory of Ministry of Education for Coastal Wetland Ecosystems,College of the Environment and Ecology,Xiamen University,Xiamen,People’s Republic of China

出　　处：《International Journal of Digital Earth》2022年第1期2335-2354,共20页国际数字地球学报（英文）

基　　金：The research was funded by the National Natural Science Foundation of China[grant no 42171283];the Major Science and Technology Projects of Qinghai Province[grant no 2021-SF-A6];the Second Tibetan Plateau Scientific Expedition and Research Program(STEP)[grant number 2019QZKK0202];Strategic Priority Research Program of the Chinese Academy of Sciences[grant number XDA19090120].

摘　　要：Spatiotemporal continuity of surface water datasets widely known for its significance in the surface water dynamic monitoring and assessments,are faced with drawbacks like cloud influence,which hinders the direct extraction of data from time-series remote sensing images.This study proposes a Time-series Surface Water Reconstruction method(TSWR).The initial stage of this method involves the effective use of remote sensing images to automatically construct multi-stage surface water boundaries based on Google Earth Engine(GEE).Then,we reconstructed regions the reconstruction of regions with missing water pixels using the distance relationship between the multi-stage water boundaries in previous and later periods.When applied to 10 large rivers around the world,this method yielded an overall accuracy of 98%for water extraction,an RMSE of 0.41 km2.Furthermore,time-series reconstruction tests conducted in 2020 on the Lancang and Danube rivers revealed a significant improvement in the image availability.These findings demonstrated that this method could not only be used to accurately reconstruct the surface water distribution missing water images,but also to depict a more pronounced time variation characteristic.The successful application of this method on GEE demonstrates its importance for use on large scales or in global studies.

关 键 词：Google earth engine sentinel-2 surface water reconstruction time-series surface water data 

分 类 号：P33[天文地球—水文科学]