Sentinel-1A影像在山区管道地表形变监测中的适用性评价指标构建

Sentinel-1A卫星数据覆盖范围广、重访周期快、获取成本低，通过合成孔径雷达干涉测量（interferometric synthetic aperture radar，InSAR）技术能够高效获取管道沿线大面积地表形变信息。但是山区管道所处地形复杂、起伏大、植被茂盛，监测中易存在叠掩、阴影、失相干等现象，Sentinel-1A卫星数据在管道不同区段的适用性有差异。为了对Sentinel-1A在山区管道地表形变测量中的适用程度进行评价，本文以三段不同山区地形的管道为研究区，结合Sentinel-1A数据、Sentinel-2数据、ALOS DEM数据进行相关性分析，构建适用性评价指标。结果显示：管道沿线Sentinel-1A影像的叠掩阴影占总面积的比例与坡度的Pearson相关性为-0.914，Spearman相关性为-1，呈显著负相关；影像相干性和归一化植被指数的Pearson相关性为-0.972，Spearman相关性为-0.99，呈显著负相关。使用回归分析和归一化的方法建立了山区管道沿线Sentinel-1A数据的坡度适用性指标和植被适用性指标，指标可对山区管道沿线域使用Sentinel-1A数据进行形变监测的适用性进行评价。

Construction of Applicability Evaluation indexes for Sentinel-1A Image in Surface Deformation Monitoring of Mountainous Pipelines

Sentinel-1A satellite data is characterized by its broad coverage, rapid revisit periods, and cost-effectiveness, allowing for the efficient acquisition of large-scale surface deformation information along pipelines through InSAR technology. However, with complex terrain, large topographic relief and lush vegetation, phenomena such as layover, shadow, and incoherence commonly arise during monitoring. The applicability of Sentinel-1A satellite data in different sections of the pipeline is different. To evaluate the applicability of Sentinel-1A data in surface deformation measurements, three pipeline sections in varying mountainous terrains were selected as study areas. Correlation analyses were conducted by combining Sentinel-1A data, Sentinel-2 data, and ALOS DEM data, in order to construct applicability evaluation indexes. The results show that the ratio of layover and shadow regions to the total areas in Sentinel-1A images along the pipelines is significantly negatively correlated with the slope, with Pearson correlation coefficient of -0.914 and Spearman correlation coefficient of -1. Likewise, there was a significant negative correlation between image coherence and normalized vegetation index, with Pearson correlation coefficient of -0.972 and Spearman correlation coefficient of -0.99. Applicability evaluation indexes for slope and vegetation were established for Sentinel-1A data along pipelines in mountainous areas using the method of regression analysis and normalization, which can evaluate the applicability of Sentinel-1A data for deformation monitoring along the mountainous pipelines.