Time series analysis of wind speed with time-varying turbulence

The characterization of the time series properties of wind speed, in terms of the mean and variance, is important and relevant to both engineers and businesses. This research investigates the first and second moments of the Texas Tech WERFL wind speed data utilizing the ARMA-GARCH-in-mean framework. The methodology allows the conditional variance to depend on the size of past shocks (i.e. gusts) in the series. Results have important implications for wind energy production as well as for the operational and financial hedging strategies of companies exposed to wind-related risk. The findings can be summarized as follows: (i) mean wind speeds measured at different heights above ground exhibit persistence and are highly dependent on immediate past wind speed values; (ii) regardless of the height at which the data were collected, wind speed exhibits time-varying variance; (iii) persistence in conditional variance increases with height at which the data were collected; (iv) there is strong evidence that conditional volatility is positively correlated with mean wind speed while the magnitude of this relationship declines with height.