A method for tracking individual planetary waves in remotely sensed data

We describe a methodology for tracking individual planetary waves in longitude-time plots of satellite data, based on fitting an elementary wave shape model to subsets of the data by maximum likelihood, then reconstructing the trajectory and evolution of every single wave (where for "single wave" we mean an individual positive or negative westward propagating anomaly) by joining the elementary waves according to their similarity. We then illustrate the potential of the methodology with an example at 34/spl deg/N in the Atlantic Ocean and its adaptability to different cases with a second example on eastward-propagating Kelvin waves in the equatorial Pacific. Although the examples given use sea surface height anomaly data, the technique lends itself to be applied to any space-time plot of any dataset displaying propagation and, in particular, to sea surface temperature data.