ValInSAR: A Systematic Approach for the Validation of Differential SAR Interferometry in Land Subsidence Areas

Abstract

Land subsidence is a natural or anthropogenic process triggering the settlement of the Earth's surface. When this phenomenon is induced by groundwater withdrawal, compaction of unconsolidated sediments causes land displacement. Differential interferometric synthetic aperture radar (DInSAR) is widely used nowadays to monitor subsidence over extensive areas. However, validation of DInSAR measurements with in-situ techniques is lacking in many case studies, reducing the reliability of further analyses. The aim of this article is to propose a systematic methodology to validate DInSAR measurements with in-situ techniques to obtain reliable subsidence measurements. The article provides a literature review of the most common approaches to validate DInSAR measurements and a description of the proposed systematic methodology, which is supported by a MATLAB open-source code. The methodology allows the analysis of both DInSAR-based velocity and displacement time series. We propose a set of statistics to assess the accuracy of the DInSAR estimates. For this purpose, RMSE parameters have been normalized with the range and the average of the in-situ deformation values. Moreover, combining these normalized parameters with the Pearson correlation coefficient (R2), a classification scheme is recommended for accepting/rejecting the DInSAR data for further analyses. This methodology has been applied in three study areas characterized by very well-documented subsidence processes: The Alto Guadalentín Valley and Murcia City in Spain, and San Luis Potosí in Mexico.

Keywords

• Accuracy
• differential interferometric synthetic aperture radar (DInSAR)
• land subsidence
• validation