Automatic Clustering of Metocean Conditions on the Brazilian Coast

Abstract

This paper introduces a pipeline that assembles a dataset of metocean conditions consisting of wind, wave, and surface currents, and then clusters these data to find the characteristic environmental conditions of each region on the Brazilian coast and the associated Exclusive Economic Zone. Clustering uses the Partitioning Around Medoids algorithm with the silhouette coefficient. As examples, we first present an analysis of the whole Exclusive Economic Zone and then a focused analysis around the Santos port in Southeastern Brazil.