Lessons learned from PSIC4: Improving PSI results for a constraint test site

In the PSIC4 experiment, eight different processing approaches for time series analysis of InSAR data were applied to a data set containing an unknown deformation signal. Results of the experiment showed different spatial point density for different processing approaches. The results of most of the groups that participated in the project suffer from a low spatial density of persistent scatterers in deforming areas. This lack of PS points in the area of interest raises the hypothesis of type-I errors (falsely rejected PS) due to imperfections in the mathematical model of PS processing. We investigate the contribution of three different sources of type-I errors: atmospheric phase screen (APS) and orbital errors, non-linear deformation mechanisms, and azimuthal sub-pixel position of the scatterers. We investigate these type-I errors and present significant improvements in their identification by optimization of the mathematical model, leading to an improved density of persistent scatterers.