Authors: Stephen Cusack
Publish Date: 2012/06/22
Volume: 116, Issue: 3-4, Pages: 693-704
Abstract
A 101 year timeseries of storm losses in the Netherlands is developed from the nearsurface wind speed records at five Dutch stations Station metadata combined with results from statistical tests were used to homogenise the data and retain the temporal variability driven solely by changes in climate processes The wind speed data were transformed into storm damage using a model measuring loss impacts upon societyThe resulting windstorm loss timeseries for the Netherlands contains some interesting features Annual losses are stable over the whole period and have a dominant cycle with a period of about 50 years The Netherlands is currently experiencing the minimum aggregate storm damage of the past 100 years though only slightly lower than a quiet period of 50 years ago Both of these minima are driven primarily by lowered rates of occurrence of damaging storms However further analysis reveals the presentday minimum has different characteristics from the previous lull currently the frequency of stronger storms is slightly above the previous minimum whereas the frequency of weaker storms is uniquely low A seasonal analysis provides more information there is a dearth of damaging storms in the earlier half of the storm season in the present day since this period contains generally weaker storms this seasonality is also manifested as a lack of weaker storms These results suggest a different mix of climate forcing mechanisms in modern times compared to 50 years ago in the earlier half of the storm seasonThe author is grateful to Christos Mitas for his advice Robert MuirWood and Steve Jewson for the initial motivation to study temporal variability of storminess in Europe Paul Wilson for his advice on loss indices in the re/insurance sector Arno Hilberts for assistance with KNMI datasets and the constructive criticisms of two reviewers
Keywords: