Estimation and correction of ionospheric effects on SAR interferometry

Abstract

Interferometric Synthetic Aperture Radar (InSAR) has demonstrated its potential for high-density spatial mapping of ground displacement associated with earthquakes, volcanoes, and other geologic process. However, a challenge to the InSAR technique is the influence of ionosphere, which may result in the distortions of Synthetic Aperture Radar (SAR) images, phases, and polarization. Moreover, this effect has become and is becoming further significant with the increasing interest in low-frequency SAR systems, limiting the further development of InSAR technique. Although some research has been carried out, this topic remains at a still very preliminary stage of development. Based on this background, this thesis focuses on the comprehensive investigation of ionospheric effects on SAR interferometry, and then develops efficient methods to estimate and correct the ionospheric artifacts on SAR interferometry.Theoretical analysis and experimental investigation are conducted to characterize the ionospheric effect on SAR interferometry. The result shows that the small-scale ionospheric irregularities can cause the azimuth pixel shifts and phase advance artifacts on interferograms, both presenting the shape of stripes and invariably extending direction in space. Compared with the Line-of-Sight (LOS) direction, displacements in the along-track direction are more easily contaminated by the ionospheric irregularities when applying InSAR technique to monitor the ground deformation. Moreover, it is found that GPS-derived rate of total electron content change index (ROTI), an index to reflect the level of ionospheric disturbances, can be used to predict the ionospheric effect for SAR images. This finding can help us evaluate the quality of SAR images in the point of ionospheric effect. Based on the fact that the SAR observations in the azimuth direction are more sensitive to the ionosphere than in the LOS direction, an efficient method is proposed to estimate and correct for the ionospheric effects on InSAR measurements. The method uses the estimated azimuth offset fields to determine the Ionospheric Phase Streaks (IPS). After that, the IPS are then subtracted from the original interferometric phase values to correct the ionosphere-contaminated nterferograms. For the performance test of the proposed method, two experiments with the real data are carried out. One experiment is applied to Chongqing City (China) and another is applied to Yushu earthquake. Results from both of the cases show that the standard deviations of the ionosphere-corrected phase values decrease by almost 2 times compared to those before the correction, demonstrating the feasibility of the method.To improve the observation precision of full-polarimetric InSAR technique, an innovative Faraday Rotation (FR)-based method is proposed to estimate and correct ionospheric artifacts. This method exploits ionospheric artifacts on interferograms being extracted from the relation between FR angles and Vertical Total Electron Content (VTEC). Compared with the existing methods, this approach is considered to be more practical since there are no limited conditions, such as coherence and bandwidth. For the performance test of the proposed method, two L-band ALOS-1/PALSAR full-polarimetric SAR images over Alaska are processed. The results show that the standard derivations of the ionosphere-orbit-corrected phase values decrease by almost 8 times compared to those of the original phase values. This experiment verifies that the proposed method can successfully estimate ionospheric phase noise and effectively correct the ionospheric noise and orbital error.