城市地表形变的双极化Sentinel‑1数据极化时序InSAR技术监测

Polarimetric Persistent Scatterer Interferometry for Urban Ground Deformation Monitoring with Sentinel‑1 Dual Polarimetric Data

利用双极化Sentinel⁃1数据有望得到较单极化数据更好的城市地表形变监测结果。以典型受地表沉降灾害影响的两座超大城市(墨西哥的墨西哥城和中国北京市)为研究区,分别基于长时序单、双极化Senti⁃nel⁃1数据,利用时序合成孔径雷达干涉测量(interferometric synthetic aperture radar,InSAR)技术和极化时序InSAR技术对两城市近年地表形变进行监测研究。结果表明,利用双极化Sentinel⁃1数据对干涉图进行极化优化后,干涉图相位质量得到提升,高质量像元比例明显增加。使用双极化数据后,形变监测点密度有显著提升,相比于单极化数据结果,墨西哥城和北京市监测点密度分别提升88%和50%;更高的监测点密度使部分区域双极化数据反演地表形变的准确性更高。与单极化Sentinel⁃1数据相比,利用双极化Sentinel⁃1数据进行城市地表形变监测可得到监测点密度更高、可靠性更好的结果。

Objectives: Thanks to the data open access policy of European Space Agency, it has become easy to acquire long time series dual polarimetric Sentinel-1 data over the most of the world ’s big cities.The use of dual polarimetric Sentinel-1 data is expected to produce better urban ground deformation monitoring results than single polarimetric data, however, there are few studies about this topic. To this end,based on the long time series single and dual polarimetric Sentinel-1 data, this study employs persistent scatterer interferometry(PSI) and polarimetric persistent scatterer interferometry(PolPSI) to monitor the recent ground deformation of the Mexico City in United Mexican States and Beijing City in China, respectively.Methods: For the interferograms’ polarimetric optimization of PolPSI technique, the exhaustive search polarimetric optimization(ESPO) method has been used, and the dispersion of amplitude(DA) is taken as the interferometric phase quality criteria. After the polarimetric optimization, for the PolPSI technique, the optimized interferograms are then employed for ground deformation monitoring through PSI processing over the two study areas. The StaMPS algorithm has been employed for the PSI processing for both single and dual polarimetric Sentinel-1 data sets. Results: The results show that, after the polarimetric optimization with dual polarimetric Sentinel-1 data, the phase quality of the interferograms is improved and the proportion of high quality pixels is significantly increased. There is a significant increase in high quality pixel after using dual polarimetric data with respect to that of single polarimetric data for ground deformation monitoring. Specifically, the qualified pixels’ density has been increased by 88% and 50% for Mexico City and Beijing City, respectively. Moreover, due to the higher pixel density, the obtained ground deformation monitoring results of some subareas are more reliable by using dual polarimetric data. Conclusion: This study demonstrates that higher density and better reliability of the qualified pixels can be obtained by using Sentinel-1 dual polarimetric data for urban ground deformation monitoring.