Géoinformatique et géostatistique : un aperçu

The Interest of Coherence from Radar SAR Images in the Follow-Up of Urban Extensions of FocÅŸani City (Romania)

K Hachemi, M Visan, A Ozer, F Grecu, K Anouche, M Jurchescu and M Nouacer

The Interest of Coherence from Radar SAR Images in the Follow-Up of Urban Extensions of Focşani City (Romania)

Usually, the coherence image obtained by processing SAR (Synthetic Aperture Radar) images is indirectly used in the interferometry to identify and delimit different areas that disturb the phase of the radar wave. The phase plays the role of indicator and in the same time separator between areas where the elements within the pixel are unstable and the stable. Considering this principle, we use here these images with the purpose to follow the expansion of the city of Focşani, Romania (urban areas). The results show the very good coherence for the urban area, which appears in white in greyscale visualisation corresponding to maximum values of coherence. The analysis of the coherence images at different time intervals allowed us to characterise the stability of the Focşani city (without considerable extension) for almost 10 years (1995/2005). Instead, we distinguished many buildings that were multiplied towards the periphery and outside of the city. During the same period, we found out that in downtown, there were urban renovations represented by green spaces. We also found that the three pairs, which were processed by the image acquired the 01/31/2005 have a low coherence, which allows us to state that the rainfalls can be the cause of the incoherence. We also observed that the incoherence is more important within the pairs formed by ERS-1 and ERS- 2 than the pairs formed by ERS-2 and ERS-2, even if the time interval between them is 24 hours (the temporal decorrelation is null or identical). This study confirms the interest of the coherence image to identify the spatial and temporal decorrelations of the pairs processed by the interferometry technique (InSAR). The role of coherence images intervenes also to distinguish the urban areas from the green or water covered areas.