融合多轨SAR数据的山地冰川三维时序形变估计

The early-warning of disasters related to the motion of mountain glacier has urgent requirement on the three-dimensional glacier displacements with high precision and temporal resolution. At present, retrieving the time series of three-dimensional displacement of the earth's surface based on multi-track spaceborne Synthetic Aperture Radar (SAR) data is one of the few ways that can achieve precise flow velocity of mountain glacier. However, previous studies could only estimate the three-dimensional displacements during the intersection time spans of multi-track SAR image pairs. For supporting the situation judgement, its temporal resolution is not sufficient yet. Aiming at this issue, our project plans to derive the three-dimensional glacier displacement in each interval of two images taken successively by integrating multi-track SAR data. With guaranteed precision, the three-dimensional flow velocity of mountain glacier will be sampled more frequently than before. Three research topics will be highlighted in this project: 1) building a new functional model that integrates the SAR temporal and three-dimensional displacement adjustments, to fully make use of the correlations among multi-track SAR data; 2) building adaptive constraint equations with respect to the feature of glacier motion, to increase the stability and logic of the solution system; 3) design an algorithm for solving the functional model with constraint, to retrieve the three-dimensional displacement with high temporal resolution. The fulfilment of this project can promote the early-warning of disasters related to the motion of mountain glacier, and can also provide better velocity samples for the study of glacier surge/collapse.

山地冰川运动相关灾害分析和预警对高精度高时间分辨率冰川流速有迫切需求。星载SAR三维时序形变估计是获取高精度山地冰川流速的有效手段，但现有研究只能估计不同轨SAR影像对的公共时段内三维形变，其时间分辨率对支持灾害分析来说还不够。对此，本项目拟通过融合多轨SAR数据来估计时间相邻影像之间的三维形变，在保证估计精度的同时大幅提高时间分辨率。主要研究内容包括：1）建立SAR三维时序形变联合平差函数模型，充分利用不同轨SAR数据之间的关联信息；2）建立顾及山地冰川运动特征的自适应约束方程，使解算系统更稳定、更符合逻辑；3）研究带约束的三维形变估计算法，实现山地冰川高时间分辨率三维形变求解。本项目可推动我国冰川运动相关灾害的预警工作，同时可为冰川跃动/冰崩机制研究提供更高质量数据支持。

山地冰川运动相关灾害分析和预警对高精度高时间分辨率冰川流速有迫切需求。星载SAR三维时序形变估计是获取高精度山地冰川流速的有效手段，但现有研究只能估计不同轨SAR影像对的公共时段内三维形变，其时间分辨率对支持灾害分析来说还不够。对此，本项目通过融合多轨SAR数据来估计时间相邻影像之间的三维形变，在保证估计精度的同时大幅提高时间分辨率。主要研究内容包括：1）建立SAR三维时序形变联合平差函数模型，充分利用不同轨SAR数据之间的关联信息；2）建立顾及山地冰川运动特征的自适应约束方程，使解算系统更稳定、更符合逻辑；3）研究带约束的三维形变估计算法，实现山地冰川高时间分辨率三维形变求解。本项目完成所有研究目标，取得了预期成果。提出一套基于升降轨匹配偏移量的SAR山地冰川时序和三维形变联合平差函数模型及严密解算方法，在保证山地冰川三维流速估计精度的同时将估计分辨率提高一倍以上。项目共发表11篇期刊论文，培养1名副教授，2名硕士。项目有效推动我国冰川运动相关灾害的预警工作，同时可为冰川跃动/冰崩机制研究提供高质量数据支持。