A young region, in full process of consolidation and continuously expanding, with ever-changing landscapes, with an inherent complexity, the Danube Delta is a very challenging target for monitoring. The complexity of the site provides an almost ideal environment for testing and evaluating the newest algorithms developed for analyzing and interpreting polarimetric SAR images. The analyzed data consists of PolSAR images acquired at L band by the ALOS / PALSAR system. Firstly, the cross-polarization channels HV and VH are analyzed over wetlands. Secondly, several information extraction, processing and interpretation algorithms are applied: parametric stochastic modeling (SIRV heterogeneous clutter models), hierarchical segmentation, various decompositions (TSVM), etc. The obtained results are consistently compared against ground-truth and both model validation and conclusions are drawn.