摘要
针对缺少实测数据导致格陵兰冰面湖水深反演及精度验证较难开展问题,提出了一种基于交叉比对的冰面湖水深反演及精度评价方法,实现了格陵兰消融期内9期时间序列WorldView影像上的冰面湖水深反演和储水量变化监测。引入北极数字高程模型开展基于形态拟合法的时间序列冰面湖水深反演,基于非线性物理模型拟合参数实现基于物理模型的水深反演,两种方法水深反演结果R^(2)平均可达0.728。9期冰面湖水深反演结果平均相对误差优于20.43%,均方根误差平均优于0.69 m。经冰面湖储水量计算,整个消融期内冰面湖总储水量累计可达5331.35×10^(4)m^(3)。通过冰面湖面积、最大水深和储水量变化监测,发现了冰面湖的融水输送现象,为研究格陵兰冰盖表面融水的存储、输送和释放机制提供有益参考。
Due to lacking of measured data,it is usually difficult to realize the water depth retrival for supraglacial lakes(SGLs)on greenland ice sheet(GrIS)based on satellite imageries as well as the accuracy analysis for the bathymetric results.In this study,we proposed a method to carry out SGLs water depth retrival for 9 time series high resolution images and introduced a cross validation method to ensure the bathymetric accuracy.SGLs volumes were calculated and the change detection was performed during the whole ablation period of GrIS.First,time series SGLs water depth retrival was realized by shape-fitting based on proper ArcticDEM.Second,the parameters were estimated by nonlinear fitting based on a physical bathymetric model and water depth retrival for time series images were realized.Based on cross validation,the average R^(2) of water depth retrival results of the two methods was 0.728,indicating a high consistency of the bathymetric results.The average of the mean relative error(MRE)for the bathymetric results based on the shape-fitting method for all of the SGLs on the 9 time series images was greater than 20.43%,and the average root mean square error(RMSE)is greater than 0.69 m.Based on the high-precision SGLs boundaries and water depth data,the volume i.e.the water storage of SGLs on time series images were calculated.During the whole ablation period,the total water storage of all the SGLs were up to 5331.35×10^(4)m^(3).By monitoring the evolution of SGLs’area,maximum water depth and volume characteristics,we found the melt water transmission phenomenon on GrIS surface through other hydrological elements such as moulins and supraglacial rivers,which could provide a reference for the study of the storage,transportation and release mechanism of melt water on the surface of GrIS.
作者
杨晓彤
谷祥辉
赵彬如
张峰
邢喆
YANG Xiaotong;GU Xianghui;ZHAO Binru;ZHANG Feng;XING Zhe(National Marine Data and Information Service,Tianjin 300171,China)
出处
《遥感信息》
CSCD
北大核心
2023年第2期87-95,共9页
Remote Sensing Information
基金
国家海洋信息中心青年科学基金项目(202001003)
国家重点研发计划“科技助力经济2020”重点专项(SQ2020YFF0426316)。
关键词
冰面湖
冰面湖水深
冰面湖体积
北极数字高程模型
变化监测
supraglacial lake
supraglacial lake depth
supraglacial lake volume
Arctic DEM
change detection
