Methods for Estimating Wind Speed Frequency Distributions

C. G. Justus School of Aerospace Engineering, Georgia Institute of Technology, Atlanta 30332

Search for other papers by C. G. Justus in
Current site
Google Scholar
PubMed
Close
,
W. R. Hargraves School of Aerospace Engineering, Georgia Institute of Technology, Atlanta 30332

Search for other papers by W. R. Hargraves in
Current site
Google Scholar
PubMed
Close
,
Amir Mikhail School of Aerospace Engineering, Georgia Institute of Technology, Atlanta 30332

Search for other papers by Amir Mikhail in
Current site
Google Scholar
PubMed
Close
, and
Denise Graber School of Aerospace Engineering, Georgia Institute of Technology, Atlanta 30332

Search for other papers by Denise Graber in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

The Weibull function is discussed for representation of the wind speed frequency distribution. Methods are presented for estimating the two Weibull parameters (scale factor c and shape factor k) from simple wind statistics. Comparison is made with a recently proposed method based on the “square-root-normal” distribution with mean wind speed and fastest mile data as input statistics. The Weibull distribution is shown to give smaller root-mean-square errors than the square-root-normal distribution when fitting actual distributions of observed wind speed. Another advantage of the Weibull distribution is the available methodology for projecting to another height the observed Weibull distribution parameters at anemometer height.

Abstract

The Weibull function is discussed for representation of the wind speed frequency distribution. Methods are presented for estimating the two Weibull parameters (scale factor c and shape factor k) from simple wind statistics. Comparison is made with a recently proposed method based on the “square-root-normal” distribution with mean wind speed and fastest mile data as input statistics. The Weibull distribution is shown to give smaller root-mean-square errors than the square-root-normal distribution when fitting actual distributions of observed wind speed. Another advantage of the Weibull distribution is the available methodology for projecting to another height the observed Weibull distribution parameters at anemometer height.

Save