Extensive analysis of potentialities and limitations of a maximum cross-correlation technique for surface circulation by using realistic ocean model simulations

As shown in the literature, ocean surface circulation can be estimated from sequential satellite imagery by using the maximum cross-correlation (MCC) technique. This approach is very promising since it offers the potential to acquire synoptic-scale coverage of the surface currents on a quasi-continuous temporal basis. However, MCC has also many limits due, for example, to cloud cover or the assumption that Sea Surface Temperature (SST) or other surface parameters from satellite imagery are considered as conservative passive tracers. Also, since MCC can detect only advective flows, it might not work properly in shallow water, where local heating and cooling, upwelling and other small-scale processes have a strong influence. Another limitation of the MCC technique is the impossibility of detecting currents moving along surface temperature fronts. The accuracy and reliability of MCC can be analysed by comparing the estimated velocities with those measured by in situ instrumentation, but the low number of experimental measurements does not allow a systematic statistical study of the potentials and limitations of the method. Instead, an extensive analysis of these features can be done by applying the MCC to synthetic imagery obtained from a realistic numerical ocean model that takes into account most physical phenomena. In this paper a multi-window (MW-) MCC technique is proposed, and its application to synthetic imagery obtained by a regional high-resolution implementation of the Regional Ocean Modeling System (ROMS) is discussed. An application of the MW-MCC algorithm to a real case and a comparison with experimental measurements are then shown.