偏移量追踪技术不易失相关，是利用光学与SAR卫星影像监测冰川流速的主要手段。该技术通过互相关方法获得像素级偏移量，并通过插值算法达到亚像素级别。实地测量冰川流速不易获取，难以用于验证亚像素级算法精度，因此本文采用图像处理方法对其精度进行分析。本研究以格陵兰Petermann和Kangerlussuaq冰川为例，通过模拟实验设定偏移量场并生成模拟的偏移影像，将COSI-Corr、autoRIFT和ImGRAFT等偏移量追踪软件得到的结果与设定偏移量场对比，而后使用三次函数拟合判断是否存在系统误差，并通过拟合函数的反函数对其进行校正。结果表明COSI-Corr软件的偏移结果的亚像素级系统误差较小，而autoRIFT和ImGRAFT存在一定的亚像素级系统误差且与实验区无关。autoRIFT的亚像素级系统偏差最大，校正后其单方向RMSE平均提升了0.0054 pixels(px)，提升率约为11%；而ImGRAFT(CCF-O)和ImGRAFT(NCC)的单方向RMSE平均分别提升了0.0014 px和0.0012 px，提升率较小。经校正后，autoRIFT精度最优，Petermann和Ka...

偏移量追踪技术不易失相关，是利用光学与SAR卫星影像监测冰川流速的主要手段。该技术通过互相关方法获得像素级偏移量，并通过插值算法达到亚像素级别。实地测量冰川流速不易获取，难以用于验证亚像素级算法精度，因此本文采用图像处理方法对其精度进行分析。本研究以格陵兰Petermann和Kangerlussuaq冰川为例，通过模拟实验设定偏移量场并生成模拟的偏移影像，将COSI-Corr、autoRIFT和ImGRAFT等偏移量追踪软件得到的结果与设定偏移量场对比，而后使用三次函数拟合判断是否存在系统误差，并通过拟合函数的反函数对其进行校正。结果表明COSI-Corr软件的偏移结果的亚像素级系统误差较小，而autoRIFT和ImGRAFT存在一定的亚像素级系统误差且与实验区无关。autoRIFT的亚像素级系统偏差最大，校正后其单方向RMSE平均提升了0.0054 pixels(px)，提升率约为11%；而ImGRAFT(CCF-O)和ImGRAFT(NCC)的单方向RMSE平均分别提升了0.0014 px和0.0012 px，提升率较小。经校正后，autoRIFT精度最优，Petermann和Ka...

偏移量追踪技术不易失相关，是利用光学与SAR卫星影像监测冰川流速的主要手段。该技术通过互相关方法获得像素级偏移量，并通过插值算法达到亚像素级别。实地测量冰川流速不易获取，难以用于验证亚像素级算法精度，因此本文采用图像处理方法对其精度进行分析。本研究以格陵兰Petermann和Kangerlussuaq冰川为例，通过模拟实验设定偏移量场并生成模拟的偏移影像，将COSI-Corr、autoRIFT和ImGRAFT等偏移量追踪软件得到的结果与设定偏移量场对比，而后使用三次函数拟合判断是否存在系统误差，并通过拟合函数的反函数对其进行校正。结果表明COSI-Corr软件的偏移结果的亚像素级系统误差较小，而autoRIFT和ImGRAFT存在一定的亚像素级系统误差且与实验区无关。autoRIFT的亚像素级系统偏差最大，校正后其单方向RMSE平均提升了0.0054 pixels(px)，提升率约为11%；而ImGRAFT(CCF-O)和ImGRAFT(NCC)的单方向RMSE平均分别提升了0.0014 px和0.0012 px，提升率较小。经校正后，autoRIFT精度最优，Petermann和Ka...

The Karakoram Anomaly has been intensively investigated, but the factors that control this anomaly, such as the glacier velocity, topography, and mass balance, remain poorly understood. To improve our understanding of the velocity, topography, and mass balance of the Karakoram Glacier, in this study, the spatiotemporal variability of four glacier velocities in the Hunza Basin of the Karakoram range were surveyed using co-registration of optically sensed images and correlation (COSI-Corr) on Landsat imagery from 1993-2019. The results show that the velocity of the Gulmit Glacier increases with a rising altitude from the glacier terminal. The three other glaciers initially display high velocity, followed by a decrease from the glacier terminal, with the maximum velocity attained in the middle of the glacier. In addition, the Karakoram glaciers produced a slight mass gain, with all mountain glaciers exhibiting clear regional acceleration from 1993-2019. The ice deformation velocity of the Batura Glacier diminished at an average rate of 8.49 %. However, the topography of the glacier base and physical factors require further analysis to determine their contribution to the observed changes in glacier velocity. In the present work, multi-temporal remote sensing image interpretations were carried out to determine glacier kinematics, which could enhance our understanding of glacier change mechanisms.