Seasonal Climate Extremes: Mechanisms, Predictability & Responses to Global Warming

Climate extremes are rarely occurring natural phenomena in the climate system. They often pose one of the greatest environmental threats to human and natural systems. Statistical methods are commonly used to investigate characteristics of climate extremes. The fitted statistical properties are often interpolated or extrapolated to give an indication of the likelihood of a certain event within a given period or interval. Under changing climatic conditions, the statistical properties of climate extremes are also changing. It is an important scientific goal to predict how the properties of extreme events change. To achieve this goal, observational and model studies aimed at revealing important features are a necessary prerequisite.

Notable progress has been made in understanding mechanisms that influence climate variability and extremes in many parts of the globe including Europe. However, some of the recently observed unprecedented extremes cannot be fully explained from the already identified forcing factors. A better understanding of why these extreme events occur and their sensitivity to certain reinforcing and/or competing factors is useful. Understanding their basic form as well as their temporal variability is also vital and can contribute to global scientific efforts directed at advancing climate prediction capabilities, particularly making skilful forecasts and realistic projections of extremes.

In this thesis temperature and precipitation extremes in Europe and Africa, respectively, are investigated. Emphasis is placed on the mechanisms underlying the occurrence of the extremes, their predictability and their likely response to global warming. The focus is on some selected seasons when extremes typically occur. An atmospheric energy budget analysis for the record-breaking European Autumn 2006 event has been carried out with the goal to identify the sources of energy for the extreme event. Net radiational heating is compared to surface turbulent fluxes of energy and dynamic horizontal advection of heat. There is clear evidence that the central North Atlantic Ocean was the major source of energy for the Autumn 2006 extreme event. Within Europe, anomalously high atmospheric water-vapor loading played a significant role through its strong greenhouse effect which resulted in an increase of downwelling infrared flux to the surface. Potential influences and connections between boreal snow cover during the melt season (February–April) and near-surface temperature in the spring season are established. Large amounts of snow act as a precursor to cold spring seasons by altering the coupling between the land and the overlying air through a modification of the surface energy and hydrological processes. In operational numerical models, a snow signal is found to provide some seasonal forecast skill for cold spring seasons in Europe.

Changes in the intensity of droughts and floods in Africa in response to global warming are investigated and compared with changes in mean precipitation simulated by an ensemble of climate models selected from the Intergovernmental Panel on Climate Change (IPCC) fourth assessment report (AR4) set. The model simulations are objectively combined using a Bayesian weighting procedure. In southern Africa south of about 15ºS, the most robust climate-change signal is a shortening of the main rainfall season. This arises from a delayed onset of seasonal rainfall associated with a reduction in lower-tropospheric moisture advection from the southwestern Indian Ocean. The semi-arid areas closer to the Kalahari desert are projected to become drier, while the wet areas are projected to become wetter. East Africa is projected to get wet in the future climate, much wetter than other regions within the same latitudinal belt. The zonal asymmetry in tropical precipitation increase is associated with a shift towards positive Indian Ocean Zonal Mode (IOZM)-like events via an alteration in the structure of the Eastern Hemisphere Walker circulation.