Long-Term Continuously Updated Deformation Time Series From Multisensor InSAR in Xi'an, China From 2007 to 2021

Abstract

According to historical records, land subsidence has been occurring in Xi'an, China, since the 1960s, characterized by complex land subsidence patterns. This subsidence has the potential to cause serious societal and economic problems during the process of urbanization. Long-term, large-scale monitoring and dynamic high precision tracking of the evolution of surface deformation associated with geohazards is a prerequisite for effective prevention or advance warning of geological disasters. Synthetic aperture radar interferometry (InSAR), a satellite remote sensing technology, can facilitate such monitoring. Both currently operating and planned SAR satellites would provide extensive amounts of SAR data. In this article, we describe an approach for continuously updating long-term multisensor InSAR deformation time series using robust sequential least squares. It is successfully applied to near-real-time monitoring of the long-term evolution of surface deformation in Xi'an, China, from January 3, 2007 to February 12, 2021, using four SAR satellites: ALOS/PALSAR-1, TerraSAR-X, ALOS/PALSAR-2, and Sentinel-1A. In order to analyze deformation evolution, temporal independent component analysis was used to interpret deformation patterns. We found that land subsidence in Xi'an has slowed and even halted in some areas. However, large areas are uplifting, which presents a potential for geohazards. We conclude that the proposed approach using continuously updated deformation time series from multisensor InSAR can provide nearreal-time deformation measurements, which are necessary for an early warning system.

Land Subsidence Monitoring and Dynamic Prediction of Reclaimed Islands with Multi-Temporal InSAR Techniques in Xiamen and Zhangzhou Cities, China

“…As for the dynamic land subsidence monitoring of Shuangyu Island, the sequential estimation proposed by Wang et al [26][27][28]36] and the correlation model between ground subsidence and time adopted by Zhao et al [30] are considered. As the sequential estimation does not process all SAR data repeatedly when new SAR data added, the deformation parameters can be dynamically updated without loss of accuracy, improving the computational efficiency of InSAR processing.…”

“…With the arrival of SAR big data era, the near real-time processing of SAR data can facilitate the development of disaster mitigation. At present, many algorithms for dynamic processing of SAR data have been developed, for example the sequential least squares algorithm [25][26][27][28], which helps us improve the efficiency of SAR data processing and reduce the consumption of storage space. In this paper, a geotechnical model is used to predict land subsidence caused by reclamation.…”

Near Real-Time InSAR Deformation Time Series Estimation With Modified Kalman Filter and Sequential Least Squares

“…Sequential least squares (SLS) [28,29], complex sequential least squares for pixel offset SBAS [30], and robust sequential least squares for Multi-Sensor InSAR data approaches [31] have been proposed to deal with this problem. However, for SLS processing, we must save the cofactor matrix of the ground-deformation time series with dimensions of N × N for each pixel (where N represents the number of SAR data), which consumes a large amount of storage space as the number of SAR data becomes extremely large in the era of big SAR data [28,30,31].…”

Time-Series InSAR Technology for Ascending and Descending Orbital Images to Monitor Surface Deformation of the Metro Network in Chengdu

“…The deformation of distance direction consists of deformation in the east direction and the north direction. In this regard, the equation for the deformation geometry of the satellite can be expressed as [39]- [41]:…”