Análise de conforto ambiental outdoor no Nordeste do Brasil

Abstract

The aim of this study was to evaluate the data calculated by the Universal Thermal Comfort Index (UTCI) for different locations in northeastern Brazil (NEB) and then analyze how outdoor comfort behaves in each state in the region. To this end, the application was a comparison over the year 2020, a year that was chosen due to the Covid-19 pandemic, which led to the shutdown of observational weather stations. This data was selected based on the amount of missing values, the selected stations had a maximum of 30% missing data, and were then filled in using the Bootstrap Expectation Maximization method (BootEM). (Boot-EM) method. To evaluate the ERA 5-HEAT global model, (Human thErmAl comforT) was calculated based on the year 2020 between January and December, through bilinear interpolations, for the selected stations of the National Institute of Meteorology (INMET), originally with output on an absolute scale, later transformed to a Celsius scale (°C). The error metrics between the observational value and that of the model were compared by Root Mean Square Error (RMSE), bias (BIAS), Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE). The results showed that the ERA 5-HEAT product has little bias due to the BIAS metric, reaching minimum values of 0,03°C. In Jaguaribe-CE, the maximum values for the MAE and RMSE metrics were 4,5°C to the west of Bahia, in the town of Luís Eduardo Magalhães. As for the MAPE metric, the highest value was observed in Oeiras-PB, reaching 17,5%. Furthermore, after calculating the monthly and hourly averages by state, the bioclimatic index identified the points in the NEB with the greatest thermal discomfort.