We present a motion computation and interpretation framework for oceanographic satellite images. This framework is based on the use of a non quadratic regularization technique in optical flow computation which preserves flow discontinuities. We also show that using an appropriate tessellation of the image according to an estimate of the motion field can improve optical flow accuracy and yields more reliable flows. A new formulation for the approximation of an oriented flow field is described. This allows us to consider an arbitrary polynomial phase portrait model for characterizing flow salient features. This new framework is used for processing oceanographic image sequences and presents an alternative to the very complex physical modelling techniques for ocean surface circulation computation.