Estimation of mean wind climate and probability of strong winds for wind risk assessment

Assessment of the relative windiness of afforestation sites has enabled the development of site classification for species selection, growth rate and wind hazard. The current development of a quantitative classification of wind risk requires a method of estimating the probability of strong winds. The relationship between wind strength and probability is commonly derived from several years of measurements at the site of interest. This is not practical when assessing wind risk for a land use rather than a single engineering structure. However, the extreme value distribution that represents the relationship between strength and probability can be derived from the mean wind climate, as represented by parameters of the parent Weibull distribution. The relationship between these parameters and previous estimates of relative windiness developed from tatter flags is explored using wind measurements from reference sites in upland Britain. A strong relationship is established between a modified geographic predictor and the Weibull c parameter. Satisfactory predictions of the Weibull c parameter are obtained for a number of validation sites using this regression and the parameters are used to derive estimates of the 1:50 year return period wind speed. A relationship between predicted and measured wind speed is also found using a dataset derived from low elevation Meteorological Office sites, but an offset is found. It is unclear whether this reflects a difference between the wind climates sampled, the effects of local surface roughness, or differences in equipment sensitivity. Further work is required to extend the method to low wind speed and high roughness sites.