Numerical modeling is an important tool in Meteorology studies to support weather forecasting and mitigation of possible disasters. Therefore, improving better its performance is one of the most discussed challenges within this thematic. In this sense, an investigation was carried out on the performance of the BRAMS model (Brazilian Regional Atmospheric Modeling System) version 5.3, in order to verify which model run options, between parameterization and input data quality, would be able to represent more accurately the winds and precipitation related to the storm that occurred on January 30, 2020, near the Brazilian southeast coast, which caused the displacement of the P-70 platform ship in Guanabara Bay. For this purpose, sensitivity tests were carried out with the BRAMS in which different parameterizations of cloud microphysics and Sea Surface Temperature (SST) datasets were changed and compared. Then, the winds of the simulations made with BRAMS were compared with the METAR (METeorological Aerodrome Report) data from the SBRJ (Santos Dumont Airport) and SBGL (Rio de Janeiro/Galeão International Airport) and also with information extracted from the reanalysis ERA5 (5th Generation of ECMWF ReAnalysis). Results show that the combination of Greg Thompson Double Moment cloud microphysics scheme and aerosol aware added to the weekly SST presented the best correlations and lowest statistical errors. It was also observed that the BRAMS had better performance after the insertion of the weekly SST as an initial condition instead of the climatological one, indicating a marked instability in Guanabara Bay and nearby. Finally, the BRAMS model was able to reproduce the increasing and decreasing trends observed by the SBGL and SBRJ airports, obtaining moderate correlations and the smallest statistical errors compared to the ERA5 reanalysis. However, for both locations, there was an underestimation of the data by the model.
Share this page on your timeline