Channelized systems form complex sedimentary reservoirs. Their modeling calls for using conditioning data when possible. A reverse migration simulation fitting to observed data has been proposed. It permits to reconstruct the channelized paleo-system and to better understand reservoir architecture. Ancient oxbow lakes are precious witnesses of the evolution of the system. However, the global chronology of their abandonment can be hard to deduce from subsurface data. Most of the time only local relative chronologies are accessible for close oxbow lakes. The knowledge of oxbow lakes global relative abandonment ages is of first importance for the reconstruction of the paleo-system. We propose a few criteria based on the geometry of the system to facilitate the extraction of local chronologies. From these local analyses, deduction of the global chronology is possible thanks to a stochastic ordering method.