Regionalization of Maximum Flows in the West Region of the State of Santa Catarina, Brazil

Código

I-EBHE0044

Autores

Luca Bonaspetti Caprara, Patrícia Kazue Uda

Tema

WG 1.12: Development & application of river basin simulators

Resumo

Between 1991 and 2012, the western region of Santa Catarina (SC) was the most affected by flooding and was the second most affected by flash floods and inundation. The overflow of streams, rivers and channels, generates considerable economic and social damage to the state. For these and other reasons, the Hydrological Monitoring Department of the Civil Defense of the State of SC encouraged this study, about the regionalization of maximum flows to the western region of SC, with the objective of developing equations that assist flood prevention systems and the sizing of hydraulic constructions in non-instrumentalized basins. After analyzing the historical flow series of the fluviometric stations available on the HidroWeb Portal, a total of 22 stations were selected, making it possible to regionalize the maximum flows for three different base periods, from 1977-1996, 1980-2014 and 2001-2014. For each base period, the statistical distributions of Gumbel, Lognormal2, Lognormal3, Pearson3 and Logpearson3 were applied to estimate the maximum flow rates for the return times of 2, 10, 20, 50 and 100 years. The explanatory variables tested in the regionalization were the drainage area (Ad), the length of the main river (L), the average slope of the main river (S), the average annual precipitation (PanualM) and the average rainfall of the semester, quarter and rainiest month (PSem +, PTri + and PMês +). The regionalization of maximum flow rates was performed using the traditional methods and the dimensionless probability curve, with the aid of the SisCoRV 1.0 software. The western region of SC was divided into two Hydrologically Homogeneous Regions (RHH) for the period 1977-1996, and a single RHH for the period 1980-2014. In these periods, the exponential model equations, when associated with variables L and S, resulted in satisfactory statistical adjustments, with determination coefficients R2 and R2 adjusted above 0.90, t and F tests below 5%, factor error below 0.50 and percentage error between observed and estimated flows lower or close to 30%. Another equation model that had satisfactory results in these base periods was the potential, with the variables Ad and L. The period 2001-2014 obtained unsatisfactory statistical results, where the values of R2 and R2a were close to 0.70. Generally, the main results of this work are satisfactory, turning possible to Civil Defense workers estimate de maximum flows to the non monitoring rivers of the western region of Santa Catarina.