Use of entropy and clustering analysis for the evaluation of water resources potential availability in the Northeastern Brazil

Abstract

The aim of this study was to define, based on Shannon entropy theory and on cluster analysis and metrics to represent the randomness of the intensity and occurrence of monthly rainfall and using these metrics to evaluate the potential availability of water resources (PAWR) in northeastern Brazil. Here the entropy is expressed in the context of probability of occurrence of the "intensity of hydrological variables," i.e, within the structure of Entropy Intensity. The criterion to perform the sampling of rainfall data was the period of data collection. Rain gauges with less than 25 years of observations were left out of this study. Data of total monthly precipitation from 874 rain gauges in the Northeast of Brazil were analyzed. This study showed the perfect adaptation of the methodology applied for the studied area. The results showed a perfect configuration with the Köppen climatic types and the weather systems operating in each area.

Keywords

• Information theory of Shannon
• precipitation
• k-means