Wind speed and groupings of frequency distribution for weibull and rayleigh models

The objective is to compare continuous mathematical models for representing wind speed behavior in Dourados, Mato Grosso do Sul, Brazil. The raw data were obtained from a semiautomatic weather station located on the campus of the Grande Dourados Federal University (UFGD). An appropriate statistical methodology was used to relate the discrete raw data to continuous mathematical functions (probability density functions). The conclusions indicate: a) a predominant northeast (NE) direction; b) a representative local wind speed of 1.6 m.s-1 according to the Rayleigh model and 1.4 -1.8 m.s-1 according to the Weibull model. The Rayleigh function shows a poor fit to the raw data, with a DPower ~1 W.m-2 lower than that obtained with the Weibull model. The Weibull function better represents the behavior and intensity of wind speed, both qualitatively and quantitatively, and is therefore recommended for estimating wind power using local data. The power density is on the order of 10 W·m?², corresponding to the theoretical power per unit rotor area of a wind turbine installed in the municipality.