基于ERS和ENVISAT交叉干涉构建高精度冰架高程模型—研究样区：南极埃默里冰架

发布时间：2018-02-28 02:43

本文关键词： 交叉干涉技术高程模型 ERS ENVISAT 埃默里冰架　出处：《北京建筑大学》2013年硕士论文　论文类型：学位论文

【摘要】：合成孔径雷达干涉技术（InSAR）可利用由空间轨道略微差别所产生的回波雷达信号的相位差来探测地表的微小变化，为测量冰盖和冰川以及他们的表面高程等参数开拓了一条道路。但InSAR在冰架表面高程的提取中具有很大难度，原因在于干涉影像对获取时间间隔如果过长，冰架的水平冰流运动和垂直的潮汐运动会导致影像间失去相关性，进而无法准确获取高程信息。欧空局的第三次ERS和ENVISAT编队串联飞行任务（EET）为利用InSAR技术获取冰架表面高程的研究带来了机会，其任务飞跃南极上空时，干涉影像对时间间隔仅为28分钟，且EET的交叉干涉影像对有2公里的空间基线，使得干涉对地形变化十分敏感，高程模糊度仅为5米。本文研究充分利用EET数据的优势，以东南极的埃默里冰架为例，选取合适EET交叉干涉影像对，经过预处理，得知试验数据的空间基线和多普勒质心差满足EET交叉干涉的条件，保证了除了冰流一致性的运动外，在观测间隔期间没有其他变形发生（例如潮汐运动）。并采用200米分辨率的RAMP高程模型作为参考高程，，进行高程信息提取，获取了高空间分辨率（25米）的高程图，结果显示提取的高程结果能准确的表现冰架表面的精细结构特征，垂直精度已经达到米级单位。本文依据实践结果对提取方法进行了改进，主要体现在：（1）根据相干图结果进行分析，表明在平坦的冰架区域有较好的可视结果，而在山地区域的解缠相位较差，因此在处理过程中对影像进行了适当裁剪，有效地避免了山地区域对整体影像解缠的影响。（2）对冰架区域的不同流速和不同干雪类型地表覆盖进行了分析，发现它们对结果的影响较小，基本可以忽略，最终发现所获取的高程误差主要是卫星基线不准确所带来的，因此本文采用GLAS点代替传统选取的控制点来重新估算基线去除残差。应用Cryosat-2和少量GPS数据与修正前后的高程结果进行比较，确认修改方法对结果有明显的改进。在埃默里冰架的实验表明，EET的交叉干涉技术可实现冰架高分辨率高精度高程模型的获取。
[Abstract]:The synthetic Aperture Radar Interferometry (InSAR) can detect small changes in the surface using the phase difference of echo radar signals generated by slight differences in space orbits. It opens a way to measure ice caps and glaciers and their surface elevation. But InSAR is very difficult to extract from ice shelf surface elevation because the interference image is too long for the acquisition time interval. The horizontal ice flow movement and the vertical tidal movement of the ice shelf have led to a loss of correlation between the images, The third ERS and ENVISAT formation series flight mission of ESA offers an opportunity to use InSAR technology to obtain the surface elevation of the ice shelf, whose mission is flying over the South Pole. The interval between interference images is only 28 minutes, and the cross-interference images of EET have a spatial baseline of 2 km, which makes the interference very sensitive to terrain changes, and the height ambiguity is only 5 meters. In this paper, the advantage of EET data is fully utilized. The Emory ice shelf in east Antarctica is taken as an example, and the suitable EET cross interference image pair is selected. After preprocessing, it is found that the spatial baseline and Doppler centroid difference of the experimental data meet the condition of EET cross interference. Besides the consistent motion of ice flow, there are no other deformations (such as tidal motion) during the observation interval. The 200m resolution RAMP elevation model is used as the reference height to extract the elevation information. The height map with high spatial resolution of 25 m) is obtained. The results show that the extracted elevation can accurately represent the fine structure characteristics of the ice shelf surface, and the vertical accuracy has reached the meter level unit. In this paper, the extraction method is improved according to the practical results, mainly reflected in the analysis of the results of the coherence map. The results show that there are better visual results in the flat ice shelf area, but the unwrapping phase is poor in the mountainous area. Therefore, in the process of processing, the image is cut properly, which effectively avoids the influence of the mountainous region on the whole image unwrapping. (2) the surface coverage of different velocity and dry snow types in the ice shelf area is analyzed. It is found that they have little influence on the results and can be neglected. Finally, it is found that the elevation error obtained is mainly caused by the inaccurate satellite baseline. Therefore, in this paper, GLAS points are used to replace the traditional control points to reestimate the baseline to remove the residual errors. The Cryosat-2 and a small amount of GPS data are compared with the corrected elevation results. The experimental results at Emory Ice Shelf show that the cross interference technique of EET can obtain the high resolution and high precision elevation model of the ice shelf.
【学位授予单位】：北京建筑大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P225.1;P224

【参考文献】