Robust velocity estimates, stream functions, and simulated Lagrangian drifters from sequential spacecraft data

The maximum cross-correlation (MCC) method has been used to compute both oceanic and cloud velocity vectors from sequences of satellite data [e.g., advanced very high resolution radiometer (AVHRR), coastal zone color scanner (CZCS), geostationary observing earth satellite (GOES)]. Unfortunately, the two-dimensional cross-correlation functions used in the computation often contain saddlepoints, which can give rise to large magnitude and direction uncertainties in the derived velocity estimates. This paper develops a numerical iterative procedure that combines image analysis methods and dynamical constraints to minimize these difficulties. The resultant velocities are both physically realistic and numerically stable. Thus, it is also possible to compute stream functions and simulated Lagrangian drifters. The validity of these results are confirmed with independent oceanic observations. Finally, the advective-diffusive equation is solved for a few oceanic applications (e.g., prediction of sea-surface temperature, dispersal of anchovy eggs and larvae) using the derived velocities