Classification of watersheds in Occidental Mindoro and Oriental Mindoro using principal component analysis and k-means clustering

Abstract

Classification of watersheds into similar groups before implementing interventions is essential since it is more systematic and sustainable than treating them individually. This study classified the watersheds within Occidental and Oriental Mindoro by applying principal component analysis (PCA) to selected characteristics of the 47 delineated watersheds and k-means clustering. The results of the PCA reduced 15 variables (area, circularity ratio, population, and 12 land cover classes) into four principal components using a threshold of 75% accounted variance of the original data. K-means clustering classified the watersheds into four clusters based on the principal components. Multivariate analysis of variance (MANOVA) was used to identify if there is a maximum between-cluster variation. Results showed that there was no association between the clusters and that the clustering of the watersheds is significant at a 5% level of significance. The results of the study may be used in choosing the most appropriate management models for each cluster.