Coupling the atmospheric model RAMS 6.0/ECHAM 4.1 to hydrologic model SMA/HMS for operating a reservoir in Brazil’s semiarid

Abstract

The planning and the efficiency of water resources are subject to the uncertainties of the input data of climate and hydrological models. Prediction of water inflow to reservoirs that would help decision making for the various water uses, contain uncertainties fundamentally the initial conditions assumed in the modeled processes. This paper evaluates the coupling of a regional atmospheric model with a hydrological model to make streamflow forecast for seasonal operation of Orós reservoir, Ceará State, Brazil. RAMS model, version 6.0, was forced by the ECHAM 4.5 atmospheric general circulation model over Alto Jaguaribe basin to obtain the rainfall data. To remove biases in the simulated precipitation fields was applied the probability density function (PDF) correction on them. Then the corrected precipitation data were inserted in the hydrologic Soil Moisture Account (SMA) model from the Hydrologic Modeling System (HEC‑HMS). For SMA calibration, it was used the Nash-Sutcliffe objective function. Finally, decisions to water release from the Orós were evaluated using the Heidke Skill Score (HSS). The SMA model showed a satisfactory performance with Nash‑Sutcliffe values of 0.92 (0.87) in the calibration (validation) phase, indicating that it is a rainfall runoff model alternative. For decisions in releasing water from the Orós reservoir, using climate predictions, obtained HSS = 0.43. The results show that the simulated rainfall coupled with a hydrological model is able to represent the hydrological operation of Brazilian semiarid reservoir.