Chen Yueqing, Qiao Sijia, Zhang Guangxin, Xu Y Jun, Chen Liwen, Wu Lili
Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun, Jilin, China.
School of Geographic Sciences, Xinyang Normal University, Xinyang, Henan, China.
PeerJ. 2020 Feb 17;8:e8616. doi: 10.7717/peerj.8616. eCollection 2020.
Interferometric Synthetic Aperture Radar (InSAR) has become a promising technique for monitoring wetland water levels. However, its capability in monitoring wetland water level changes with Sentine-1 data has not yet been thoroughly investigated.
In this study, we produced a multitemporal Sentinel-1 C-band VV-polarized SAR backscatter images and generated a total of 28 interferometric coherence maps for marsh wetlands of China's Momoge National Nature Reserve to investigate the interferometric coherence level of Sentinel-1 C-VV data as a function of perpendicular and temporal baseline, water depth, and SAR backscattering intensity. We also selected six interferogram pairs acquired within 24 days for quantitative analysis of the accuracy of water level changes monitored by Sentinel-1 InSAR. The accuracy of water level changes determined through the Sentinel-1 InSAR technique was calibrated by the values of six field water level loggers.
Our study showed that (1) the coherence was mainly dependent on the temporal baseline and was little affected by the perpendicular baseline for Sentinel-1 C-VV data in marsh wetlands; (2) in the early stage of a growing season, a clear negative correlation was found between Sentinel-1 coherence and water depth; (3) there was an almost linear negative correlation between Sentinel-1 C-VV coherence and backscatter for the marsh wetlands; (4) once the coherence exceeds a threshold of 0.3, the stage during the growing season, rather than the coherence, appeared to be the primary factor determining the quality of the interferogram for the marsh wetlands, even though the quality of the interferogram largely depends on the coherence; (5) the results of water level changes from InSAR processing show no agreement with in-situ measurements during most growth stages. Based on the findings, we can conclude that although the interferometric coherence of the Sentinel-1 C-VV data is high enough, the data is generally unsuitable for monitoring water level changes in marsh wetlands of China's Momoge National Nature Reserve.
干涉合成孔径雷达（InSAR）已成为监测湿地水位的一项有前景的技术。然而，其利用哨兵 -1 数据监测湿地水位变化的能力尚未得到充分研究。
在本研究中，我们制作了多时相哨兵 -1 C 波段 VV 极化 SAR 后向散射图像，并为中国莫莫格国家级自然保护区的沼泽湿地生成了总共 28 幅干涉相干图，以研究哨兵 -1 C - VV 数据的干涉相干水平与垂直基线、时间基线、水深及 SAR 后向散射强度的关系。我们还选取了 24 天内获取的六对干涉图进行定量分析，以评估哨兵 -1 InSAR 监测水位变化的准确性。通过六个野外水位记录仪的值对哨兵 -1 InSAR 技术确定的水位变化准确性进行校准。
我们的研究表明：（1）对于沼泽湿地中的哨兵 -1 C - VV 数据，相干性主要取决于时间基线，受垂直基线影响较小；（2）在生长季早期，哨兵 -1 相干性与水深之间存在明显的负相关；（3）沼泽湿地中哨兵 -1 C - VV 相干性与后向散射之间存在几乎线性的负相关；（4）一旦相干性超过 0.3 的阈值，在生长季期间，尽管干涉图质量很大程度上取决于相干性，但生长阶段而非相干性似乎是决定沼泽湿地干涉图质量的主要因素；（5）InSAR 处理得到的水位变化结果在大多数生长阶段与实地测量结果不一致。基于这些发现，我们可以得出结论，尽管哨兵 -1 C - VV 数据的干涉相干性足够高，但该数据总体上不适用于监测中国莫莫格国家级自然保护区沼泽湿地的水位变化。
