United Kingdom windspeed: Measurement, climatology, predictability and link to tropical Atlantic variability

Windspeed impacts the business performance of many industries yet has received relatively little attention compared with other meteorological fields. The present study addresses this inconsistency. A UK seasonal windspeed climatology is de veloped using a new dataset of United Kingdom hourly windspeed measurements comprising 30 years of observations from 52 geographically dispersed sites. The data are shown to contain significant errors associated with non-ideal measurement conditions. A correction algorithm is described and on application the adjusted site-records exhibit improved homogeneity. Seasonal climatological windspeed char acteristics are modelled using the Weibull distribution: results indicate that central southern England can expect 1-6 near-gale events each winter, compared with 22-27 near-gale events in southwest England. Rare (strong) event return periods are modelled using Gumbell extreme-value theory. Seasonal predictability of winter storminess is investigated using an index defined by the 95th percentile of winter daily maximum windspeed (SI). The interannual variability of SI over Europe is dominated by the North Atlantic Oscillation (NAO, 34% Percentage Variance Ex plained (PVE)). Conversely, a secondary SI mode of variability (SI2, 19% PVE) is seen to have a significant impact over the UK. Multi-field correlation analysis is employed to assess potential SI2 predictability, with statistical forecast models built from the results: the models show mixed skill performance. Tropical North Atlantic (TNA) windspeed is shown to co-vary with winter NAO: surface tradewinds for Dec- Jan-Feb are 19% higher in strong-NAO composite years compared to weak-NAO composite years. In turn this impacts the subsequent distribution of Caribbean rainfall: wet-season precipitation is significantly reduced following a strong winter NAO. It is hypothesised that changes in the TNA trade winds create long lasting SST anomalies, which in turn feedback onto wet season convective activity. Results from a long integration coupled climate model (HadCM3) support the observed results: model TNA tradewinds co-vary with winter NAO, and a reduction is seen in Caribbean wet-season rainfall following a strong NAO