Calibration of the SMAP rainfall-runoff model in a semi-arid gauging station in the northwest region of Brazil, using PSO, AG and Nealder & Mead algorithms and its effects on the definition of the monthly stream flow series.

Abstract

This study evaluated the effects on estimating the monthly streamflow of a northeast gauging station by employing the PSO (Particle Swarm Optimization) and AG (Genetic Algorithm), associated with the local search method Nelder Mead (1965) for the calibration of the rainfall-runoff model SMAP. It was observed that when generating the flow series for the 1963-2013 period the average monthly flow hydrograph showed monthly values higher than those one observed in the gauging station. It was found that the SMAP model applied on this northeast semi-arid region the parameters CREC, K and EBin proved to be expendable and the model can be calibrated only for SAT, PES and TUin parameters.