Clustering of a Runoff Regime in Slovakia Using the PCA Method

Abstract

A new scheme has been created for pooling data on the long-term average monthly runoff in Slovakia using information from 57 gauging stations from 1991 to 2020. A statistical analysis has determined the optimal number of clusters from the normalised data, and the Principal Component Analysis (PCA) method and K-means clustering were used to group basins into five pooling groups. Finally, the critical characteristics of these pooling groups were determined, and a typical regime was determined.