A crucial statistical method for identifying subgroups within a dataset is clustering. Finding hidden correlations between the observational variables in the data is made possible by clustering. Many researchers have successfully used clustering approaches in a variety of knowledge domains. Here, our attention is on applications to ocean science, with a concentration on the Arabian Sea. This article uses SKATER, a graph-based clustering method, to regionalise the Arabian Sea, and these clustered regions are helpful in locating suitable places to deploy the buoys.