Variabilidade multiescalar da velocidade do vento no Nordeste do Brasil

Abstract

This study analyzed the multiscale variability of wind speed in Northeast Brazil (NEB) using observed data from INMET meteorological stations (2010–2021) and reanalysis products (ERA-5 and MERRA-2). Multiple imputation techniques were applied to fill gaps in the time series, ensuring data consistency. Statistical analysis included the Weibull distribution and Pearson correlation, while hierarchical clustering identified spatial and temporal patterns. The results revealed a well-defined diurnal cycle, with higher wind speeds during the day (09:00–20:00 UTC) due to Atmospheric Boundary Layer (ABL) dynamics, and four distinct clusters, with the highest wind speeds concentrated along the coast (Ceará, Rio Grande do Norte, and Paraíba), influenced by sea breezes and trade winds. Seasonally, wind speeds were more intense between August and September, associated with the positioning of the Intertropical Convergence Zone (ITCZ), and lower in March and April. The interannual analysis showed limitations in cluster separation and the identification of interannual patterns but indicated influences from phenomena such as El Niño and La Niña. It was concluded that multiple imputation and clustering are effective for climate studies, providing valuable insights for energy planning in NEB. Local validation of reanalysis data and the use of longer time series are recommended to improve future analyses.