Replacement of fossil fuel-based energy generation by renewable technologies offers the potential to reduce greenhouse gas emissions. In most European countries, wind energy is the renewable resource that has undergone the greatest expansion in the last twenty years. However, public concerns about visual intrusion and noise are growing, and it is becoming increasing difficult to obtain planning permission for wind farm development on land. Against this background, offshore wind farm development appears increasingly attractive, despite higher construction and maintenance costs. Offshore wind speeds should be at least as high as at many onshore sites, and public concerns should be fewer.

When planning a wind farm onshore, a developer will typically measure wind speeds for a period of at least one year, in order to evaluate the size of the available resource. Offshore, such an undertaking may be prohibitively expensive. This paper reports results from the three-year POWER research project, ‘Predicting the Offshore Wind Energy Resource’, which set out to evaluate the size and characteristics of the wind energy resource for the offshore waters of European nations.

The project required the development and application of two models: first, to predict wind speeds in the far offshore zone and, second, to develop a coastal discontinuity model, or CDM, to predict wind speeds in the region closer to shore, where thermal and dynamical effects from the coastal discontinuity exist. The first step in the far offshore model was to calculate the geostrophic wind speed and direction from sea level pressure data taken from the NCEP reanalyses. This was done at the daily scale for a ten-year period, and at a spatial resolution of 0.5o by 0.5o. These data were translated to turbine hub height using the simple dynamical boundary layer model WAsP, which applies a wind speed vertical profile to the geostrophic wind speeds, assuming a constant roughness length of 0.0001 m. Then, the CDM modifies these wind speeds to take account of roughness and stability changes between land and sea. A one-second digitised coastline is used, and the CDM calculates distance to coast in twelve compass directions.

To validate the model output, extensive use has been made of observational data from fixed anemometers, SODAR and radiosonde ascents at coastal stations, weather ships and oil and gas platforms in the North Sea. However, for some areas, such as the Mediterranean Sea, very little useful data exists. The context of the ten-year period within which the two prediction models have been applied, in terms of low-frequency climate variability, has been established by constructing and analysing geostrophic wind speeds for the period 1880 to present using a coarse-resolution sea level pressure data set. Overall, it is anticipated that this comprehensive collection of modelled and observed data on wind speeds in the near and far offshore will facilitate wind farm development in European waters.