Identifying Persistent Scatterers (PSs) is one of the key processing steps of the Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR) technique. The number, density, and reliability of identified PSs directly affect the monitoring accuracy of land subsidence, especially in higher density urban environments. As a result of the side-looking viewing geometry of SAR, the layover effect poses a major challenge to the PS identification. This research proposes joint modeling of the PS-InSAR technique and RELAX algorithm for SAR tomography (PS-InSAR+RELAX) to detect single and double scatterers and to improve the identification and reliability of PSs. It has been demonstrated that RELAX improves separation of the scatterers when compared to two other spectral analysis methods for SAR tomography, Beam-Forming (BF) and Singular Value Decomposition (SVD). RELAX exhibits the least noise when the number of baseline changes from 15 to 30, and it can separate the scatterers at a lower Normal-Slant-Range (NSR) height than the two other methods. As RELAX can better identify, separate, and then filter out layover scatterers, the number and density of PSs identified by PS-InSAR+RELAX is reduced and visually simplified, suggesting that the method can effectively reduce the influence of the layover effect on the PS identification. Also, the PSs identified by PS-InSAR+RELAX are more coherent than those identified by the traditional PS-InSAR technique. The proposed technique has been applied to Sentinel-1A data acquired from 2014 to 2016, to monitor land subsidence in the city of Beijing, China. When evaluated against the leveling measurements, PS-InSAR+RELAX performs better than the traditional PS-InSAR technique, with the correlation coefficients (r) of r = 0.98 and r = 0.95, respectively.
A Critical Analysis to Generate Change Detection Map using SAR Interferometry for Land Subsidence Monitoring of New Orleans City of USA