1 Meteorology, Wind Energy Division, Risø National Laboratory for Sustainable Energy, Technical University of Denmark2 Wind Energy Division, Risø National Laboratory for Sustainable Energy, Technical University of Denmark3 Risø National Laboratory for Sustainable Energy, Technical University of Denmark4 Department of Wind Energy, Technical University of Denmark
A statistical model for extreme winds in the western North Pacific is developed, the region on the Planet where tropical cyclones are most common. The model is based on best track data derived mostly from satellite images of tropical cyclones. The methodsused to estimate surface wind speeds from satellite images is discussed with emphasis on the empirical basis, which, unfortunately, is not very strong. This is stressed by the fact that Japanese and US agencies arrive at markedly different estimates. Onthe other hand, best track data records cover a long period of time and if not perfect they are at least coherent over time in their imperfections. Applying the the Holland model to the best track data, wind profiles can be assigned along the tracks. Fromthis annual wind speed maxima at any particular point in the region can be derived. The annual maxima, in turn, are fitted to a Gumbel distribution using a generalization Abild’s method that allows for data wind collected from multiple positions. Thechoice of this method is justified by a Monte Carlo simulation comparing it to two other methods. The principle output is a map showing fifty year winds in the region. The method is tested against observed winds from Philippine synoptic stations and fairagreement is found for observed and predicted 48–year maxima. However, the almost bias–free performance of the model could be fortuitous, since precise definitions of ’windspeed’ in terms averaging time, height above ground and assumed surface roughnessare not available, neither for best tracks nor for the synoptic data.