Simulação da disponibilidade hídrica na bacia hidrográfica do rio Piquiri- Pr

Abstract

This study aims to estimate hydric availability in Piquiri River basin based on the historical flow series and precipitation from 1980 to 2010 in order to associate them to the rainfall estimates provided by Eta, a regional weather model, until 2098. The series of annual minimum flows (Q7) and minimum of seven-day flows with a registering period of ten years (Q7,10) were submitted to statistical analysis to identify the probabilistic model that best fitted the data for each station. Thus, in order to obtain the retention curve, a procedure was carried out based on the obtained frequency classes. The average values of annual rainfall, precipitation of the driest quarter, precipitation of the wettest quarter, precipitation of the driest semester and precipitation of the wettest quarter of the basin were determined using the IDW interpolation method using a geographic information system (SIG). From all precipitation and flow obtained values, regressions were performed among (Q7), (Q7,10), (Q 90%), (Q 95%) flows and rainfalls (total annual, the driest quarter, the wettest quarter, the driest semester and the wettest semester) associated with the drainage area. Standard statistical tests were carried out to assure the quality of the generated model.For estimates of future flows, it was necessary to obtain data of future annual rainfall, simulated by the weather model Eta, from 2010 to 2098, which were provided by the Weather Forecasting and Climate Studies Center (CPTEC/INPE). The independent variable that best corresponds to the proposed standards during the study was the total annual rainfall in order to define the rainfall-runoff basin model. According to the future data of annual rainfall and drainage area, through the rainfall-runoff model, it was possible to obtain estimates of future minimum flows and carry out the trend analysis of time series with the series of future flow estimates. Therefore, a simple linear filter was used for smoothing the data series. It was not possible to find a pattern of rainfall related to the driest and wettest months in Piquiri River basin. Multiple linear model fitted best to the data, whose coefficient of determination (R)2was 0.67. The charts concerning generated trend analysis showed little increase in total annual rainfall index, simulated for the future in the studied region, when compared to the historical series. It was also observed that both members of the Eta model showed similar results. The values of minimum flows estimated for the future have provided to the basin an area without water deficit under natural conditions