An Evaluation of Tropospheric Correction Models for InSAR in Ground Deformation Monitoring: A Case Study in Zhejiang Province, China

Abstract

Interferometric Synthetic Aperture Radar (InSAR) is a powerful and cost-effective technology to monitor ground deformation. Its accuracy is highly influenced by the atmospheric characteristic of the monitoring area. Separating the true ground deformation from atmospheric signals remains one of the major challenges in the application of InSAR. In this paper, the phase-based linear model, high-spatial resolution weather model (MERRA-2 and GACOS) and combination of the MERRA-2 and phase-based linear model are selected, and their performances in reducing the tropospheric delay are assessed based on the detrending standard deviation (DStd) of all Persistent Scattered (PS) points. A framework for the assessment is proposed and applied to a selected region of Shaoxing, Zhejiang Province, China. A total of 26 Sentinel-1A images are used and processed by the method of PS-InSAR. It is found that the phase-based linear model outperforms the other models by at least 6.6% if the whole monitoring time span of the SAR images in the study area is considered. The proper tropospheric correction model in different seasons is not the same. The phase-based linear model is robust against the variations in the atmospheric characteristics of the four seasons.