Simulação do escoamento com o modelo navmo e regionalização hidrológica no sistema Piranhas/PB - Açu/RN.

Abstract

In this research, regional analyses techniques were applied to long-term simulations of mean and maximum annual discharges, obtained with the model NAVMO, as well as discharges with 95% of permanence, in order to establish equations to estimate hydrologic variables for catchment planning and decision making in the Piranhas/PB-Acu/RN region in the Northeast region of Brazil. Owing to the lack of observed data for comparisons, the homogeneity and consistency of the simulated series were investigated through the double mass test, parametric and non-parametric tests, as well as the test of Kolmogorov-Smirnov (K-S) to investigate some statistic distributions to represent the frequencies of the simulated data at the level of 5% of significance. Simple and multiple regression analyses were utilized to establish equations capable of estimating hydrological variables in terms of climatic and physic characteristics of the region at non-observed sites. The results showed that, in general, the simulated series are homogenous. The results from the parametric and non-parametric tests suggest that the statistical characteristics of some series do not match. The K-S test showed that, for the study region the distribution of Pearson III can be utilized to estimate the discharges in general, and particularly the Gama distribution for the mean annual discharges. Maps for the parameters of the Pearson III distribution, and of the equation established to estimate the discharges with 95% of permanence, were prepared for the region. Finally, the regression analyses showed that the hydrologic variables were well explained by the characteristics of the region