Implementation of Non-Linear Non-Parametric Persistent Scatterer Interferometry and Its Robustness for Displacement Monitoring
-
- Fumitaka Ogushi
- Department of Built Environment, Tokyo Institute of Technology, Yokohama 226-8502, Japan
-
- Masashi Matsuoka
- Department of Architecture and Building Engineering, Tokyo Institute of Technology, Yokohama 226-8502, Japan
-
- Marco Defilippi
- Sarmap S.A., 6987 Caslano, Switzerland
-
- Paolo Pasquali
- Sarmap S.A., 6987 Caslano, Switzerland
説明
<jats:p>To derive surface displacement, interferometric stacking with synthetic aperture radar (SAR) data is commonly used, and this technique is now in the implementation phase in the real world. Persistent scatterer interferometry (PSI) is one of the most universal approaches among in- terferometric stacking techniques, and non-linear non-parametric PSI (NN-PSI) was proposed to overcome the drawbacks of PSI approaches. The estimation of the non-linear displacements was successfully conducted using NN-PSI. However, the estimation of NN-PSI is not always stable with certain displacements because wider range of the velocity spectrum is used in NN-PSI than the conventional approaches; therefore, a calculation procedure and parameter optimization are needed to consider. In this paper, optimized parameters and procedures of NN-PSI are proposed, and real data processing with Sentinel-1 in the Kanto region in Japan was conducted. We confirmed that the displacement estimation was comparable to the measurement of the permanent global positioning system (GPS) stations, and the root mean square error between the GPS measurement and NN-PSI estimation was less than 3 mm in two years. The displacement over 2π ambiguity, which the conventional PSI approach wrongly reconstructed, was also quantitatively validated and successfully estimated by NN-PSI. As a result of the real data processing, periodical displacements were also reconstructed through NN-PSI. We concluded that the NN-PSI approach with the proposed parameters and method enabled the estimation of several types of surface displacements that conventional PSI approaches could not reconstruct.</jats:p>
収録刊行物
-
- Sensors
-
Sensors 21 (3), 1004-, 2021-02-02
MDPI AG