基于Sentinel-1的典型地物时间失相干分析

Temporal Decorrelation Analysis of Different Land Coverage Types Based on Sentinel-1 Images

为了探究影响Sentinel-1时间失相干的因素,实验选取美国南加州柯汶纳市及其周边区域进行研究,使用50景Senti⁃nel-1数据,分析了不同地物的时间失相干现象,并研究了风雨等天气因素及不同极化方式对相干性的影响。主要得出以下结论:Sentinel-1数据典型地物的失相干特征区别较大,可以较好地指导地物识别、辅助土地分类;小于7.5 m/s的风速对相干性几乎无影响,而降水是导致地物失相干的重要因素,对农作物相干性的影响更大且持续性长;VH极化方式抗干扰能力更强,VV极化方式对变化的感知能力更强。实验研究表明:选取相干目标应考虑土地覆盖类型的影响,也需要利用气象数据对干涉图失相干情况进行评估,同时应充分利用不同极化方式的数据。

Temporal decorrelation is the most difficult part in InSAR deformation monitoring.To explore the factors affecting temporal decorrela-tion of Sentinel-1 data,we selected Covina in Southern California and its surrounding areas for the study.We used 50 scenes of sentinel-1 data to analyze the temporal decorrelation phenomenon of different land cover types.Then,we analyzed the relationships of weather conditions(i.e.,wind and precipitation)and different polarization modes with the coherence.The main conclusions are as follows:①The decorrelation character-istics of typical land cover types in sentinel-1 data are quite different,so sentinel-1 data can be used to guide feature recognition and assist land classification.②Wind speed less than 7.5 m/s has little effect on coherence.Precipitation is an important factor leading to decorrelation of land cover types,and its impact on crop coherence is greater and longer.③VH polarization mode has stronger anti-interference capability,and VV po-larization mode has stronger capability to perceive changes.The experiment results show that the influence of land cover types should be consid-ered in InSAR applications based on coherence,meteorological data should be used to evaluate interferogram decorrelation,and data of different polarization modes should be fully utilized.