Análise do comportamento da velocidade do vento na região Nordeste do Brasil utilizando dados da ERA-40

Abstract

Wind speed distribution depends on relief of a certain region, in the case of Brazilian Northeast (NE) there are four subregions with different characteristics of relief: Atlantic Rainforest zone, Agreste, North-Eastern Backlands and Northeast Mid North. Among probability distributions that were proposed for wind data, Weibull, General-ized Gamma and Rayleigh distributions were found to be most appropriate to model wind speed variability in many locations. In this work we analyze ERA-40 data, during the period 1958-2001, to evaluate which probability distribution is most suitable to describe temporal variability of wind speed in NE. ERA-40 is a re-analysis of meteor-ological observations from September 1958 to August 2002 produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) in collaboration with many institutions. The data were produced using different sources such as radiosondes, ocean buoys, satellites and meteorological stations, and have temporal resolution of 6h and spatial resolution of 2.5ox2.5º. The two parameter Weibull distribution is found to be the best for modeling frequency distribution of wind speed data for most of the NE. The coastal zone is characterized by strongest winds and smallest temporal variability. These results can be used to evaluate the wind energy potential at NE and the influence of wind on various environmental phenomena as soil erosion, rainfall, dune formation and dispersion of seeds and poluents.