Implementation of Non-Linear Non-Parametric Persistent Scatterer Interferometry and Its Robustness for Displacement...

Area of Science:

• Geodesy
• Remote Sensing
• Earth Science

Background:

• Surface displacement monitoring is crucial for hazard assessment and infrastructure management.
• Synthetic Aperture Radar (SAR) interferometry, particularly Persistent Scatterer Interferometry (PSI), is a key technique for deriving surface displacement.
• Existing PSI methods face limitations in estimating non-linear and complex displacement patterns.

Purpose of the Study:

• To propose optimized parameters and procedures for Non-linear Non-parametric PSI (NN-PSI).
• To validate the enhanced NN-PSI method using real-world Sentinel-1 SAR data.
• To demonstrate NN-PSI's capability in accurately estimating various surface displacement types.

Main Methods:

• Implementation of optimized NN-PSI algorithms.
• Processing of Sentinel-1 SAR data over the Kanto region, Japan.
• Comparison of NN-PSI results with permanent Global Positioning System (GPS) station measurements.

Main Results:

• NN-PSI achieved high accuracy in surface displacement estimation, with a root mean square error of less than 3 mm compared to GPS data over two years.
• The method successfully reconstructed displacements exceeding 2π ambiguity, which conventional PSI approaches misinterpret.
• Periodical displacements were accurately captured, showcasing NN-PSI's enhanced capability.

Conclusions:

• Optimized NN-PSI provides a robust and accurate method for surface displacement estimation.
• NN-PSI overcomes limitations of conventional PSI, enabling the reconstruction of complex and periodical displacements.
• The validated NN-PSI approach enhances SAR interferometry applications in geosciences.