Hi,

interesting question, firstly one has to differentiate between repeat pass and single pass bistatic (srtm, tandem-x). Also the used wavelength, incidence angle, effective baseline and of course the observed terrain play a crucial role. Next is the processing can lead to errors, the phase unwrapping is not trivial, but also the question which multi looking suits your application best is a difficult question. The different spatial resolutions are also related to different issues. Other errors that come to my mind: height ambiguity, temporal decorrelation, sensor/measurements errors, atmospheric effects, speckle, low coherence... did I forget a important one?

Hope my answer gives some further ideas, what is your application?

From a single-pass multi-antenna InSAR/IFSAR susyem, precision is mainly about baseline length (measured in wavelength), and accuracy is mainly about knowledge of baseline orientation...  range and resolution play a role...  precision can be traded somewhat for post-spacing...  larger baselines also yield height ambiguities which affect accuracy...  tie-points in the scene can help, if available...  another factor is shadowing, since you have to see the target before you can measure it (e.g. problems with urban canyons, steep terrain, etc.)...

The accuracy of DEMs derived from InSAR depends on many factors , some of these factors related to the space platform. While other factors relate to terrain characteristics, radar antenna design, signal propagation through atmosphere, physical characteristics, platform dynamic motion and processing algorithm for radar echo's.