Modelo de previsão de vazão aplicado ao nordeste brasileiro utilizando a informação climática para auxiliar a operação de hidrossistemas

Abstract

It is known that water is a resource essential to human survival. Environmental managers around the world have been calling attention to the problem of water scarcity and mismanagement. This study highlighted that in the Northeast of Brazil, particularly in semi-arid, water scarcity is accentuated because of the peculiar characteristics that differentiate it from other places, such as the irregularity and poor distribution of rain in time and space. Mathematical models of seasonal/interannual streamflow forecasting incorporating weather information were developed, as well as the methodology for incorporating these forecasts in the operation, management and study of water systems and flood control. These models were based on the methodology of the K-neighbors developed by Souza Filho and Lall, and on the statistical regression models. The software AcquaNet was also introduced as a mechanism to aid in the operation of reservoir systems. The methodology was applied in reservoirs of the State of Ceará and the Northeast of Brazil. We conclude that the inclusion of streamflow forecasting in the operation of reservoir systems can be an effective tool in water management as this process adds more reliability and quality, and, if properly used, can increase the income generation of the State.