Stochastic modeling of time series radar interferometry

Quality description and evaluation of InSAR results is hampered by the fact that the model to derive parameters from the observations is usually underdetermined. Only using strong, often rather qualitative, assumptions it is possible to reach unique solutions. One of the most prominent assumptions is that phase ambiguity resolution can be treated as a deterministic problem. In this study, a model formulation is presented that captures the majority of the assumptions in a mathematical sense, allowing for adjustment, testing procedures and formal error propagation. The influence of stochastic ambiguity resolution to the probability distribution of the estimated parameters is shown.