Sie verwenden einen veralteten Browser mit Sicherheitsschwachstellen und können die Funktionen dieser Webseite nicht nutzen.
Timescale: Oct.2015 – Sept.2018
Prof. Dr. Uwe Ulbrich, Freie Universität Berlin
Prof. Dr. Henning Rust, Freie Universität Berlin
The frequency of climate-related natural hazards is expected to alter in a changing climate. The effect of anthropogenic climate change depends on the type of hazard and on the region considered. The IPCC tends to focus on common signals (“trends”) in climate models under increasing greenhouse gas forcing. The known fact that the hazard probability varies slowly in a natural way is treated as a problem of detection and attribution. Recent initiatives aimed at establishing decadal predictions (like the German MiKlip initiative) are designed to forecast trends in the coming years, including the occurrence of extremes (e.g. Kruschke et al., 2014). An investigation of mechanisms associated with long term variability of European windstorms by Nissen et al. (2013), based on an ensemble of climate simulations used a simple method to identify anomalous periods.
Objectives and Methods
The goal of this PhD-project is to develop sophisticated methods for the identification and quantification of decadal anomaly periods with respect to the occurrence and intensities of winter wind storms in Central and Southern Europe, discerning them from the anthropogenic climate change. In particular, non-stationary response patterns from ocean temperature anomalies are considered, using decadal climate prediction runs. Sources of variability in the climate system will be studied by using non-stationary extreme value statistics (e. g., Rust et al., 2009) and the fraction of attributable risk approach (Jaeger et al., 2008) in order to analyse the effects of drivers of long-term climate variability. Climate simulations with different set-ups are used, e.g., control runs without anthropogenic forcing and ocean-induced decadal variability, transient runs including anthropogenic forcing factors (Millennium run) and finally also decadal forecast experiments. Typical response patterns from specified sources of variability (e. g. greenhouse gases) can be used as statistical fingerprints to increase the signal-to-noise ratio in detected and attributed analyses. Non-stationarity is incorporated by using spatio-temporal fingerprints.
Thomas Moran is based at the research team “Climate diagnostics and extreme meteorological events” of Freie Universität Berlin.