A Small-Baseline InSAR Inversion Algorithm Combining a Smoothing Constraint and $L_1$ -Norm Minimization

Abstract

Atmospheric artifacts and phase unwrapping errors have unfavorable effects on differential synthetic aperture radar interferometry (DInSAR) deformation monitoring. In this letter, we present an alternative small-baseline DInSAR inversion algorithm, velocity-constraint $L_{1}$ -norm minimization. The proposed algorithm improves the robustness of time series deformation estimation by combining a smoothing constraint and $L_{1}$ -norm minimization. The smoothing constraint can minimize temporal atmospheric artifacts, and the $L_{1}$ -norm minimization outperforms $L_{2}$ -norm minimization in the presence of phase unwrapping errors. The iteratively reweighted least square algorithm is employed to adjust the weights of DInSAR observations and smoothing constraints in $L_{1}$ -norm minimization. The proposed algorithm is validated using simulated data and TerraSAR-X data. The experimental results show that the proposed algorithm is suitable for the inversion of approximately linear deformation processes affected by both atmospheric artifacts and unwrapping errors.