A growing interest in extreme precipitation has spread through the scientific community due to the effects of global climate change on the hydrological cycle and their threat on natural systems more than averaged climatic values. Understanding the variability of precipitation indices and their association to atmospheric processes could help to project the frequency and severity of extremes. This paper evaluates the trend of three precipitation extremes: the number of consecutive dry/wet days (CDD/CWD) and the quotient of the precipitation in days where daily precipitation exceeds the 95th percentile of the reference period and the total amount of precipitation (or contribution of very wet days, R95pTOT). The aim of this study is twofold. First, extreme indicators are compared against accumulated precipitation (RR) over Europe in terms of trends using non-parametric approaches. Second, we analyse the geographic opposite trends found over different parts of Europe by considering their relationships with large-scale processes, using different teleconnection patterns. The study is accomplished for the four seasons using the gridded E-OBS dataset developed within the EU ENSEMBLES project.
Different patterns of variability were found for CWD and CDD in winter and summer, with north-south and east-west configurations, respectively. We consider physical factors to understand the extremes variability by linking large-scale processes and precipitation extremes. Opposite association with the North Atlantic Oscillation in winter and summer, and the relationships with the Scandinavian, East Atlantic patterns and El Niño/Southern Oscillation events in spring and autumn gave insight into the trend differences. Significant relationships were found between the Atlantic Multidecadal Oscillation and very extreme precipitation (R95pTOT) during the whole year. The largest extreme anomalies were analysed by composite maps using atmospheric variables and sea surface temperature. The association of extreme precipitation indices and large-scale variables found in this work could pave the way of new possibilities for the projection of extremes in downscaling techniques.