Enhanced description of the wind climate in Denmark for determination of wind resources

The model used is HIRLAM, the numerical weather prediction model used by DMI. By coupling the HIRLAM model to a sea surface wave model, WAM, a coupled model system with an enhanced ability of describing the sea-atmosphere interactions was established. Different ways of coupling the weather model and the wave model have been investigated, and it is now clear that the coupled model system should be covering a quite large geographical area if the improvement should be significant. The fact that the coupled model system should run over a larger area made it more heavy in terms of calculation and computer needs, than previously expected. We chose to focus on thoroughly calculating a number of conditions relating to the quality of the potential wind power forecasts. It turns out that the wind forecasts' quality is much more dependent of the horizontal resolution than of the vertical resolution. Furthermore the quality of the forecasts is very sensitive to the accuracy of the description of small scale turbulence in the model. We also investigated if information on the actual wind power production could improve the wind forecasts in low height, near the land-/sea-surface, and for that purpose we developed a new method based on conditional probabilities. A conditional wind speed distribution is determined by Baye's rule, using the probability function of the expected wind power production as a function of the wind. This method was tested using data from Abild wind mill farm, and it showed that it is possible to improve a first guess wind forecast further by taking into account the actual power production from the wind mills. The results obtained during the project is documented in the final report and it is to be expected that they will be utilised in future work for improved model system for analysis and forecasting purposes