A stochastic model has been developed to generate sequences of hourly mean wind speeds for different sites in the United Kingdom. A feature of its formulation is that three distinct wind speed regimes have been identified within a time series and each regime fitted separately. Sixteen stations providing data over a 15 year period have been used to develop and test the model. Results are encouraging, more especially for the higher speeds.

A stochastic model to generate sequences of hourly mean wind speeds for different sites in the United Kingdom

A technique has been developed to predict wind statistics in remote locations where wind data are usually not available. The technique uses historical wind data from a synoptic weather station together with a statistical prediction model to obtain data from which wind statistics can be developed. The form of the statistical model and its parameter estimates were developed from simulations for the region of interest that, in turn, were based on a hydrodynamic model. Predictions and evaluations were made in and around the city of St. Louis. The best predictions of wind statistics were made at locations near the center of the city where the simulations were best. The accuracy of the simulations depended on the parameters used to describe the forcing functions. It was found that a major source of error in the predictions can be attributable to the discrete nature of the wind data from the synoptic weather station which was used as the basis for determining the wind statistics.

C.A. Clayton

Papers

Technique to determine wind statistics in remote locations. Final report