冰流速是估计南极物质平衡的重要参数，长时间序列的冰流速观测对于评估全球气候变化影响具有重要意义。近代遥感卫星主要获取1973年后的数据，导致对更早期的冰流速参数认识不足。美国解密的第一代胶片卫星数据填补了1960年代的卫星遥感空白，为拓展南极遥感观测时间尺度提供了可能。在已有预研基础上，本项目将突破早期胶片卫星影像成像质量差、姿态数据不足和同时期地面信息缺乏的限制，构建基于核线影像动态视差分解的南极冰流速反演新方法，开发面向极地冰雪环境的弱纹理影像特征提取和密集匹配算法，实现冰流速度由核线影像空间向地面物方空间的投影转换，并进行误差分析，最终建立一套早期胶片卫星影像的现代化摄影测量遥感处理新方法，实现南极冰流场重建。项目将选择典型冰架冰川区域开展实证研究，分析冰流速演化规律。本研究对深化南极冰盖长时序演化过程认知，减少冰盖变化估计不确定性，提升全球海平面上升贡献预测精度等方面具有重要意义。

Ice flow velocity is an important parameter for estimation of Antarctica mass balance, and long time series observations will be of great significance in evaluation of the global change effects. Modern satellite remote sensing began since 1973, resulting in the poor understanding of the earlier ice flow parameters. Declassification of the first satellite imagery in film format by US filled the gaps of satellite data in 1960s, providing the possibility to extend the time scale in Antarctica remote sensing. Based on the basis of pre-research, this project will overcome the limitations of earlier film-based satellite images, such as low imaging quality, insufficient attitude parameters and lack of ground information in the same time period. We will develop a novel approach to interpret the ice flow velocity in Antarctica based on the dynamic parallax decomposition in epipolar image, followed by a dense image matching algorithm for the low-texture images in polar ice-snow environment. We will then transform the parallax displacement of ice flow velocity from epipolar image space to the ground space, followed by the error analysis. Finally, a modern photogrammetric-remote sensing method will be developed for the early film-based satellite imagery, to reconstruct ice flow field in Antarctica. Study areas will be selected in typical glaciers and ice shelves to further verify the proposed method, and analyze the evolution of ice flow velocity. This project is of significance in deep understanding of long time series evolution in Antarctica, reducing the uncertainty of ice sheet evaluation, as well as improving the prediction accuracy on contribution of global sea level rise.