Large-Scale ENVISAT ASAR Persistent Scatterer Interferometry Using GNSS ZTD Products

Abstract

<p>Interferometric Synthetic Aperture Radar (InSAR) provides essential information dealing with different natural hazards caused by hydrogeological processes turned into subsidence (Lu et al., 2010). Among all geodetic techniques for monitoring land surface deformation, InSAR has demonstrated the most precise observations in both high spatial and temporal resolution (Dehghani et al., 2009). Different cities of Belgium are subject to land subsidence mostly induced by groundwater over-exploitation. In this study, we present Persistent Scatterer (PS-InSAR) over Belgium to highlight land surface deformation between 2003 to 2010. To this purpose, large-scale interferometry is accomplished by applying the PS-InSAR algorithm on three tracks of ENVISAT ASAR radar images covering the entire country. To provide an integrated land surface velocity map from several single acquisitions, eight stacks of ENVISAT ASAR C-band frames of Single Look Complex (SLC) data acquired from track 380, 423, and 466 spanning from 2003 to 2010 are used. Time series of deformation and mean velocity map are two products of applying PS-InSAR on the dataset. Joining single interferograms generated from successive frames along their acquisition tracks makes a randomly directed phase ramp in the mean velocity map due to atmospheric effects. The atmospheric effect composed of vertical and turbulent components is one of the most important error sources in using space-based deformation monitoring approaches. Monitoring land surface deformations, the atmospheric artifacts could mask real values of deformation phase calculated by different InSAR methodologies (Yu et al., 2018). To address this problem, different methods having different degrees of accuracy are widely suggested in the literature (Xiao et al., 2021). The patchy active deforming areas over Belgium along with a variety of atmospheric conditions through time and very slow ground deformation velocities make it a perfect case study for applying different proposed approaches for reducing the atmospheric effects. These approaches could be categorized into two main groups: i) those which are method-dependent and reduce all stochastic residual noise through processing, and ii) those which are based on external data such as meteorological features. The main advantage of the first group is that using them will not require any external data. However, the drawback is that the high spatial variability of the atmospheric artifact is disregarded by applying them. Overcoming this drawback, using professional GPS data is the most accurate method of the second group. Zenith Total Delay (ZTD) measurements are provided with the highest temporal resolution by GPS; therefore, additional uncertainties are inhibited (Yu et al., 2018). Using this method is an ongoing part of this study, but before that the Generic Atmospheric Correction Online Service for InSAR (GACOS) has been applied. The GACOS procedure is a developed package in the TRAIN software. Applying this procedure to the interferograms showed no tangible changes in them. In this study which is being carried out as part of the BRAIN BESLSPO project: "monitoring LAnd SUbsidence caused by Groundwater exploitation through gEOdetic measurements (LASUGEO)", we benefit from ZTD files which are by-products of GNSS processing to mitigate atmospheric effects.</p>