Mitigation of ionospheric artifacts in InSAR data for estimating earthquake deformation

Abstract

Earthquakes are a common natural hazard threatening human lives and property. As a result of the sudden relaxation of stored elastic strain energy, earthquakes permanently deform the Earth's crust around the plane of fault interfaces. Geodetic observations are generally used for estimating the crustal displacement triggered by an earthquake, and for providing insights into the geophysical processes and fault mechanisms of an earthquake. Since the early 1990s, spaceborne interferometric synthetic aperture radar (InSAR) has become one of the most popular geodetic techniques for studying earthquake-induced crustal displacements and has many applications worldwide. Spaceborne synthetic aperture radar (SAR) signals interact with the ionosphere when they travel through it during the synthetic aperture time, and the condition of the ionosphere and its variation can significantly affect the measurements of spaceborne InSAR. Furthermore, the artifacts in InSAR measurements can lead to uncertainties in the source parameters of an earthquake. To explore and mitigate the effects of ionospheric artifacts on InSAR measurements, we begin by extending and improving the azimuth shift method for the case of earthquakes. However, in practice, the effectiveness of this method is affected by the deformation phase contribution when applied to earthquakes. We propose an offset field method to improve the estimation of the ionospheric phase screen (IPS), and successfully applied the proposed method to the 2008 Wenchuan earthquake and the 2010 Darfield earthquake. Reliable separation of the IPS from other interferometric phase contributions, especially the deformation, is always a critical task when estimating coseismic deformation of an earthquake. The azimuth shift-based method is therefore not a theoretically perfect method in some cases. Nevertheless, the range split-spectrum method is based on the natural dispersive properties of the ionosphere and is currently the most operational and effective method for mitigating the ionospheric artifacts in InSAR. However, when mitigating the IPS with the range split-spectrum method, we noticed that the Gibbs ringing artifacts are often amplified when a sharp transition appears in the computed pixel offset. We then propose the use of a Kaiser windowed sinc interpolation kernel to effectively reduce the biases. Additionally, the performance of the symmetric and asymmetric range split-spectrum methods is also evaluated for further SAR missions. After reliably mitigating the ionospheric artifacts in InSAR, we discuss the effect of IPS on the estimation of the earthquake slip model using the InSAR measurement alone and when combined with other geodetic data. First, we apply the range split-spectrum method to the 2009 southern Sumatra earthquake to evaluate the effect of ionospheric artifacts on the fault slip model with single-track of InSAR data. We then provide a comprehensive study of the impacts of ionospheric artifacts on the fault slip models of the 2010 Maule, Chile megathrust earthquake using combined geodetic data (leveling, GPS, and InSAR). The results from these studies show the importance of considering ionospheric effects on InSAR measurements when studying earthquakes. Otherwise, the source parameters of the earthquakes, and also their reliability, determined based on InSAR measurements can be significantly impacted.