Abstract
This thesis investigates the evolution of German Bight (southeastern North Sea) stormi-ness from 1897 to 2018 through analyzing upper quantiles of geostrophic wind speeds,which act as a proxy for past storm activity. Here, geostrophic wind speeds are cal-culated from triplets of mean sea level pressure observations that form triangles overthe German Bight. The data used in the thesis are provided by the International Sur-face Pressure Databank (ISPD) and the national meteorological services, namely theDeutscher Wetterdienst (DWD), the Danmarks Meteorologiske Institut (DMI), andthe Koninklijk Nederlands Meteorologisch Instituut (KNMI).The derivation of storm activity is achieved by enhancing the established triangle proxymethod via combining and merging storminess time series from numerous partiallyoverlapping triangles in an ensemble-like manner. The utilized approach allows for theconstruction of robust, long-term and subdaily German Bight storminess time series.Further, the method provides insights into the underlying uncertainty of the timeseries.The results show that storm activity over the German Bight is subject to multidecadalvariability. The latest decades are characterized by an increase in activity from the1960s to the 1990s, followed by a decline lasting into the 2000s and below-average ac-tivity up until present. The results are backed through a comparison with reanalysisproducts from four datasets, which provide high-resolution wind and pressure datastarting in 1979, and offshore wind speed measurements taken from the FINO-WINDproject. This study also finds that German Bight storminess positively correlates withstorminess in the NE Atlantic in general. In certain years, however, notably differentlevels of storm activity in the two regions can be found, which can be attributed tovariabilities in the occurrence frequency of large-scale circulation patterns. An inves-tigation of the underlying uncertainty reveals that uncertainty is inversely related todata availability and generally reduced compared to previous studies for the NortheastAtlantic, owing to the increased number of triangles and the more robust method.
