Maximum Flow Regionalization in the Hydrographic Basin on Piquiri River - Pr. (PAP014786)

Código

PAP014786

Autores

Fernanda Cristina Araújo, ELOY LEMOS DE MELLO, GISELE MARIA GOLLIN, RENATO DOS SANTOS SANCHES

Tema

Urban Floods

Resumo

The knowledge about maxim flow enables the avaliation of the flood hazard and the flashflood control for the measurement hydraulic building. However, one of the more impediments for the realization in this work is default hydrological data, outstanding inadequacy monitoring mesh or short season in the most historical series. The flow regionalization in a interesting alternative for the supply this deficiency. Among others aspects the regionalization techniques have in view to transfer the information for monitor regions the areas wanting data. Besides, the study enables identification areas to need improvement to the hydrometeorology mesh. The purpose this study was accomplishes the maxim flow regionalization in the hydrographic basin on Piquiri River - PR. As explanatory variables were used the physics characteristics: draining area (A); principal river length (L); medium declivities basin (Dm); medium declivities the principal river (Dr); draining density (Dd) an the climatic characteristic: total annual precipitation (Ptot); medium annual precipitation (Pmed) and the more rainy trimester precipitation. Were tested probability distribution, Log Normal two and three parameters, Gumbel, Pearson type III, Log Pearson type III, to identify the better fitting of the series maxim events series to the return period 2,5,10,25,50,100 years. Subsequently applied the regression analysis in the maxim flow specify associated in the return period worked out, in the base math function Qmax Tr= Qmax Tr (A,L,Dm,Dr, Dd,Ptot, Pmed, Ptri). The regression multiple model were select based in the criteria: i) model simplicity (the least number explanatory variable) and ii) adjustment quality, measured by determination coefficient, standard error, model significance by test F, hypothesis test model parameters and residual analysis. With the choice purpose the representative distribution each station, elected that better adjusted the highest number to the return periods. The distribution Log Normal in three parameters was better adjusted in the majority station, subsequently Log Normal two parameters. Also observed the Log Normal Pearson III distribution adjusted well to short return periods.