Mesoscale Events Classifier: an algorithm for the detection and classification of upwelling events using Sea Surface Temperature satellite data

In two previous technical reports we described a tool that produces a so-called spaghetti plot, i.e. a plot that is able to capture the trends of the sea surface temperature (SST) in a chosen time interval and within a target area; and the formalization of spaghetti plots through the definition of two custom Python 3 classes. In this report we outline an algorithm that uses SST data to detect and classify mesoscale upwelling events. In particular, the algorithm (called Mesoscale Events Classifier, MEC) takes as input the SST data organized as a SpaghettiData dictionary and returns a map of the area of interest where the zones in which the algorithm detects an event are highlighted and labelled with an event type.