Entropia de Shannon aplicada à precipitação pluvial

Abstract

The theory of information entropy developed by Claude Elwood Shannon in the 1940s was an important milestone in computer science and communications systems. His ideas have been broadened and applied to other fields of science, including environmental science and water resources. The purpose of this paper is to demonstrate part of the entropy theory of information applied to water resources in three chapters: the first chapter demonstrates the general concepts of marginal entropy, joint entropy, conditional entropy, mutual entropy, maximal entropy principle, the main applications in hydrology are analyzed: variability analysis, statistical hydrology, modeling of hydrological phenomena, evaluation of hydrological monitoring networks. The second chapter was to use the concept of marginal entropy to investigate the spatial behavior of total annual precipitation between the period 1990-2015 of the state of Minas Gerais. It was possible to identify an increasing gradient of precipitation variability in the south-north direction of Minas Gerais. It was delineated 4 homogeneous areas with their own characteristics from the values of distribution entropy and annual precipitated volume. The third chapter evaluated the quality of the information from three networks of rainfall stations of 3 UPGRH in the state of Minas Gerais. It was verified that the water resources planning units analyzed presented regions with deficits and regions with excess stations indicating, according to this methodology, the need for reallocation of the stations.