| Summary: | Climate change is likely to influence the frequency of extreme extremes - temperature, precipitation and hydrological extremes, which implies increasing risks for flood and drought events in Europe. In current climate, European countries were often not sufficiently prepared to deal with the great social vulnerability originating from such extreme meteorological events. European extreme events are dependent on atmospheric circulation patterns (synoptic scale systems in our environment), which have a large impact on climate variability and exhibit a strong seasonal variation. Major investments in European climate adaptation measures and infrastructure planning are based on the outcomes of impact studies, with special attention to the changes of extreme events. A powerful tool used in impact studies are climate models but because of their biases, they are unsuitable for direct use; a bias correction (i.e. using an observation record) is essential. Several statistical methods are used in bias correction, but only few of them include information of the relationship between circulation patterns and temperature or precipitation extremes. Such incorporation provides an important contribution in understanding extremes and improves the correction. This thesis examines the connections between temperature and precipitation extremes, and atmospheric circulation patterns and adapts a set of statistical methods to include these connections. Temperature and precipitation extremes are treated separately based on the challenges they pose to the European climate adaptation planning. Extreme temperature was analyzed in a broad European domain, while precipitation extremes were explored in the Rhine basin, specifically to improve estimates of extreme river discharge. The effects of atmospheric circulation patterns on temperature and precipitation extremes are addressed separately to better assess immediate consequences, important for emergency planning. The thesis begins with an analysis of European hot spells i.e. a sequence of days with maximum temperature above a certain threshold. Known circulation patterns like the North Atlantic Oscillation pattern and the atmospheric blocking are incorporated in extreme value analysis in an elaborated way to spatially analyze European hot spells. It is concluded that these circulation patterns influence the magnitude, frequency and duration of the hot spells in Europe, establishing the proposed extreme value analysis as a suitable candidate to analyze connections between hot spells and circulations. Extreme value analysis was used further in the thesis for the analysis of precipitation extremes and the bias correction of climate models. Two main topics are addressed here. Firstly, the precipitation observation record for the Rhine basin and the influence of the observation in the bias correction, and in the estimation of return precipitation and discharge levels. The extrapolation of such levels to can induce large uncertainties. The extended observation record developed for the Rhine significantly improves the discharge estimations. Secondly, the adaptation of a well-used bias correction method is addressed to include circulation patterns. The bias correction is applied on an ensemble of climate models and uses the extended precipitation data set described before. The adjustment in the bias correction allows for improvements in the precipitation changes that are associated with changes in specific circulation patterns and their frequency of occurrence. Finally, it is concluded that the improvements in the precipitation correction leads to improvements in the estimation of extreme discharge return levels in the Rhine basin.
|
|---|