Filtering Drifter Trajectories Sampled at Submesoscale Resolution

In this paper, a variational method for removing positioning errors (PEs) from drifter trajectories is proposed. The technique is based on the assumption of statistical independence of the PEs and drifter accelerations. The method provides a realistic approximation to the probability density function of the accelerations while keeping the difference between the filtered and observed trajectories within the error bars of the positioning noise. Performance of the method is demonstrated in application to real data acquired during the Grand Lagrangian Deployment (GLAD) experiment in the Northern Gulf of Mexico in 2012.