InSAR海量数据高精度准动态形变时序分析

The coming of InSAR massive data era can greatly reduce the SAR satellite revisit time, which can significantly improve the InSAR deformation monitoring and forecasting ability to the environment and natural hazard in terms of monitoring accuracy, dimension, and quasi-dynamic. Accordingly, the earth science research can be transformed from postmortem event-based inversion stage to the beforehand process-based forecast stage. This proposal aims to answer three questions facing InSAR massive data processing as follows. How to suppress the interferogram noise and to increase the coherent targets density efficiently? How to increase the reliability of InSAR results automatically? And how to obtain one-dimensional or two-dimensional quasi-dynamic time-series deformation results? Finally, a software suite of InSAR massive data processing will be developed including the following techniques improvements, batch filtering of SAR interferograms, optimized SAR interferometric pairs combination, automatic phase unwrapping error detection and correction based on QUAD algorithm and phase unwrapping error mitigation algorithm based on weight iteration method, sequential one-dimensional time-series InSAR results quasi-dynamic update and two-dimensional time-series InSAR results update based on multidimensional SBAS (MSBAS) method. This research can be expected to increase the InSAR fine monitoring ability to natural hazards and promote InSAR technique to engineering level, which will provide a great support to hazards prevention and mitigation and man-made infrastructures safety.

InSAR海量数据时代的出现极大地缩短了SAR卫星的重访周期，显著地提高了InSAR技术对环境、灾害的监测精度、监测维度、准实时监测能力和预警水平，将地球科学的研究从事后事件式反演阶段推进到事前过程式预警阶段。本项目针对InSAR海量数据处理中面临的三大问题，即如何有效抑制干涉图的噪声并增加有效监测点的密度，如何自动化提高InSAR监测结果的可靠性和如何进行InSAR准动态形变时间序列监测开展干涉图批滤波与干涉对优化组合算法研究，基于拟准检定法的相位解缠误差自动化识别与修正和基于选权迭代法的相位解缠误差减弱算法研究，序贯平差一维时序InSAR结果准动态解算和基于MSBAS二维时序InSAR结果准动态解算，最终实现一套InSAR海量数据处理软件系统。项目预期研究成果可提升InSAR技术对自然灾害的精细化监测能力，有利推动InSAR技术的工程化应用，为防灾减灾以及工程安全提供重要技术支撑。

项目围绕InSAR有效监测点位密度增加、高精度准动态时序解算、多维形变反演、海量数据自动化处理以及科学解译等内容开展系统研究。在InSAR有效监测点密度增加方面，开展了DS相位优化、ISBAS时序形变估计等研究，针对时空去相干因素，提升有效监测点位密度与精度。在InSAR高精度准动态时序形变解算研究方面，开展了InSAR时序形变序贯估计与稳健估计，多传感器融合InSAR时序形变序贯估计，SAR偏移量二维时序形变序贯估计及改进的卡尔曼滤波和序贯最小二乘平差用于InSAR时序动态更新。在InSAR多维形变反演研究方面，开展了InSAR蠕变型滑坡三维时序形变监测方法研究，跨平台SAR偏移量剧滑型滑坡二维与三维时序形变监测方法研究，并基于地表多维形变揭示了多处典型滑坡变形破坏模式。在InSAR海量数据自动化处理研究方面，搭建了InSAR海量数据自动处理云平台。此外，在InSAR结果后处理分析研究方面，基于所获取InSAR地表多维长时序形变，开展了地下水储量反演等相关研究。整体研究内容推动了InSAR技术用于地表形变多维长时序监测，准动态更新等方面的发展，为地质灾害识别监测及后续研究提供了技术基础。