Combination of Persistent Scatterer Interferometry and Single-Baseline Polarimetric Coherence Optimisation to Estimate Deformation Rates with Application to Tehran Basin

Abstract

This study reports on the monitoring of land subsidence in a rural area located in the southwest of the Tehran basin, Iran, by combining a persistent scatterer interferometry (PSI) method with a single-baseline polarimetric coherence optimisation. Owing to vegetation coverage in this rural area, coherence level experiences a decline and the performance and coverage of conventional interferometry to estimate deformation rate reduces concomitantly. Since the launch of satellites with polarimetric information, the polarimetric InSAR (PolInSAR) technique, which is vector interferometry with different polarimetric channels, has been introduced to optimise the coherence level. One of the most common criteria to select PS pixels is coherence and maximising the coherence can lead to an increased number of selected PS pixels and enhanced PSI performance. The single-baseline polarimetric coherence optimisation method assumes equal polarisation states at the end of each baseline. In order to apply this technique in our study, two different multi-look windows for coherence calculation and also two TerraSAR-X data sets with different numbers of images are used to assess their effect on the polarimetric PSI. Combination of the single-baseline coherence optimisation method with PSI shows significant improvements (more than 50%) in terms of the number of selected PS pixels in the case study even using a data set with a small number of images. A 15 times 15 multi-look window selects a greater number of PS pixels compared to a 9 times 9 multi-look window, although this entails reducing spatial resolution. The most effective PSI approach in terms of the density of the selected PS turned out to be polarimetric PSI using a data set with a large number of images and a selection of a 15 times 15 multi-look window. Validation of the PSI methods using a large number of images with 9 times 9 and 15 times 15 multi-look windows via levelling measurements confirms the accuracy and reliability of the results obtained.