Clustering of Argo Float Data

Richard Karsten
Associate Professor
Department of Mathematics and Statistics
Acadia University

The large amount of data that the Argo floats supply allows for the use of a variety of statistical tools to examine the physical properties of the worlds ocean. In particular, using principle component and clustering techniques, we can identify water masses that share similar variability. For the first time, the Argo data allows us to examine both geographical and temporal variability of water at intermediate depths. These techniques are used to examine the characteristics of water masses in the Southern Ocean, in particular Antarctic Intermediate Water.