Resposta hidrológica das porções mineiras das bacias hidrográficas dos rios Doce e São Francisco à Modelagem Chuva-Vazão

Abstract

Flow forecasting in extreme events caused by an intense volume of offers is one of today's greatest challenges and can help water resource managers in decisionmaking, minimizing the risks arising from these events. The present work aimed to implement the HEC-HMS model in the Minas Gerais portions of the Doce and São Francisco river basins. The methods used in hydrological modeling were: Deficit and Constant for runoff separation; Clark Unit Hydrograph for hydrograph formation; Recession for base flow; and Muskingum-Cunge for runoff propagation. The input data were obtained from the CHIRPS product and data from the hydrometric stations in Hidroweb. The precipitation input data were obtained by the CHIRPS product and the flow data for calibration and validation from the fluviometric stations in Hidroweb. The calibration period was from January 1, 2010 to December 31, 2017 and the validation period from January 1, 2019 to December 31, 2022, reproducing the main extreme flood events that occurred in recent years. The model results were evaluated based on the Nash-Stucliffe Efficiency Index and PBIAS and were classified as “satisfactory to very good”. Flow forecasting was also performed using the ACCESSCM2 model considering two scenarios SSP245 (optimistic) and SSP585 (pessimistic) for a period of 10 years and the results indicated that the type of scenario had a great influence on the magnitude of the peak discharge. The HEC-HMS software performed well and simulated the values coherently and confirming its capacity in flow prediction.