A Non-Stationary Periodic Temporal Decorrelation Model for Insar Stacks Over Pasture Areas

Abstract

Temporal decorrelation is one of the main error sources in satellite radar interferometry. As the range of physical mechanisms causing temporal decorrelation is wide, there is no single analytical method to model this effect. Recent studies report seasonally varying coherence behavior over pasture areas, which cannot be described by the current analytical models of temporal decorrelation. To acknowledge this periodicity, we introduce a new analytical model. Here, the hypothetical movements of elementary scatterers within resolution cells are modeled as a periodic stochastic process with non-stationary increments. The proposed model is a function of the temporal baseline and the date of the master image of each interferogram. The parameters of the proposed decorrelation model have been estimated and validated for a case study in the Netherlands