Europe has experienced several damaging wind storms in recent years. For example, a series of three severe storm (Anatol, Lothar and Martin) swept across Europe in December 1999. The most costly of these three storms was Lothar - responsible for an estimated insured loss of 5.8 million US dollars. This raises concerns as to whether the risk of such events occurring will increase as climate changes.
Changes in European wind storm risk at the end of the century have been investigated using climate model simulations from the ENSEMBLES project. Analyses focus on changes in cyclone (strong depressions) activity, extreme wind speeds and loss potentials and assume no action is taken to reduce greenhouse gas emissions (the SRES A1B emisssion scenario). While most models agree in showing a largely reduced total number of cyclones in the Northern Hemisphere, in the majority of model simulations there is a hotspot of increased activity of extreme storms over the Eastern North Atlantic. Both signals can also be identified in the ensemble (combined) mean of all simulations. Furthermore, there is also an increase in the mean intensity of cyclone systems over the Eastern North Atlantic / European region. In agreement with the changes in extreme cyclones, extreme wind speeds show increases over northern parts of Central and Western Europe, while over Southern Europe wind speeds are reduced. Assessments of changes in storm loss potentials reveal an increased risk (i.e. more potetntial damage) especially in Central Europe, accompanied by increased variability from year to year, indicating the occurrence of some extremely strong events.
Projected changes in cyclones (left) and extreme cyclones (right) at the end of the century. Blue indicates a decrease, red an increase (Leckebush et al., 2004).
Ensemble mean climate change signal (IPCC SRES A1B) of the cyclone track density. Left a): All cyclones. Right b): Extreme cyclones. Units: systems per winter (ONDJFM). Areas with an altitude above 1500 m are eliminated. Coloured: Statistical significance above the 90/95/99% level according to a student t-test. The ACC ensemble mean signals are weighted by the quality of each model in reproducing observed cyclone climatologies.