A integração dos modelos WRF e MGB-IPH como suporte à prevenção de inundações: um estudo de caso para a bacia do Paraíba do Sul

Abstract

The Paraíba do Sul River Basin (PSRB) stands out in the Brazilian economic scenario, covering a region responsible for approximately 14% of the national GDP and encompassing the states with the highest population density and economic development in the country. However, the PSRB is frequently impacted by extreme precipitation events, resulting in floods and landslides. This study evaluates the integration of the WRF atmospheric and MGB-IPH hydrological models to predict extreme precipitation and flow events, with the aim of optimizing warnings of extreme flood events in the region. To understand the hydrological behavior of the PSRB, historical precipitation and flow series were analyzed, identifying average patterns, trends and extreme events, as well as selecting a critical period for model validation. The atmospheric simulation with WRF was started with data from GFS/NCEP and validated using data from MERGE and rainfall stations, while the MGB-IPH model was adjusted and calibrated to represent the basin. Hydrological simulations were carried out with precipitation data from WRF and GFS, and evaluated with data from fluviometric stations. The historical analysis revealed that the higher-altitude regions of the PSRB record higher rainfall accumulations, higher flows and a greater frequency of extreme events. Along the Paraíba do Sul River, a negative flow trend was observed, possibly associated with the growing demand for water resources. However, the central region of the PSRB showed an increase in the occurrence of extreme flow events, reinforcing the need for monitoring and flood warnings. The results indicate that WRF simulated the spatial distribution of precipitation with greater consistency, but underestimated extreme events in some areas. The hydrological simulations showed that integrating the WRF with the MGB-IPH improved the flow forecast, reducing errors and better capturing flood peaks compared to the GFS, standing out as a promising tool for hydrological monitoring and forecasting with a focus on mitigating the damage caused by floods.