Matrizes de probabilidade de transição de um reservatório no semiárido

Abstract

The superficial reservoirs located in the Brazilian Northeast are subject to high rates of evaporation in most of the year, with the aggravation of receiving inflow only in a few months due to the intermittent nature of their rivers. Thus, these reservoirs tend to face problems with their storage, which causes the region to suffer the consequences of long periods of drought. In order to understand the behavior of surface reservoirs located in the semi-arid region, the present work used the principles of the Stochastic Theory of Reservoirs, developed by the statistician Pat Moran, to obtain transition probability probabilities between storage states pre established in the reservoir Rice Pond. A total of 25 transition matrices were obtained. The first array compares the reservoir volumes for a series of 100-year flows without taking into account in what months storage state transitions were occurring. The following twelve matrices show the transition probabilities for the twelve sequences of months of the year. Such matrices show the tendency of the reservoir to remain in the state of storage in which it was in the previous month or to jump to another state of storage. The last twelve transition matrices relate the storage volumes of a month in relation to the same month of the previous year. Thus, the transition probability matrices obtained have the main purpose of obtaining the characteristics of semi-arid reservoirs in order to be able to have a better operational control and a more efficient management of the use of the water stored in the reservoir.