Impacts of climate variability and climate change on renewable power generation

Anthropogenic climate change represents a major risk for human civilization and its mitigation requires reductions of greenhouse gas emissions. To stay consistent with the long-term temperature targets of international climate policy, global greenhouse gas emissions have to reach zero within a few decades. Such a dramatic transition towards sustainability in all sectors of human activity requires the decarbonization of power generation at an early stage. In absence of other viable technology choices and given the significant cost declines, renewable power generation forms the backbone of the decarbonization. In contrast to thermal power plants, most renewables are not dispatchable but their generation dynamics are governed by the weather. This dissertation adds to the quantification of impacts of climate variability on wind power generation on different time scales. In particular, it shows that inter-annual wind power generation variability already today has a strong influence on congestion management costs in Germany. Understanding this variability as a normal system feature helps to prevent short-sighted reactions in legislation and power system design. Moreover, it is shown that relevant multi-decadal wind power generation variability exists. Owing to timescales of up to 50 years, these modes are not sufficiently sampled in any modern reanalysis (e.g., MERRA2 or ERA-Interim), which currently cover around 40 years. Consequently, power system assessments based on modern reanalyses may be flawed and should be complemented by multi-decadal assessments. In this context, I also show that 20th century reanalyses (ERA-20C, CERA20C, 20CRv2c) disagree strongly and systematically with respect to long-term wind speed trends. The discrepancy can be traced back to marine wind speed observations which also feature strong upward wind trends that are likely due to an evolving measurement technique. As a consequence, 20th century reanalyses should be employed with care and cross-validation of results is recommended. Due to their weather dependency, renewables are potentially vulnerable to climate change. Indeed, I show that the benefits of large-scale transmission infrastructure in Europe shrink under strong climate change (RCP8.5). The effect is robust across a five member EUROCORDEX ensemble and can be solidified in a larger CMIP5 ensemble. It is rooted in more homogeneous wind conditions over Europe that lead to less smoothing effects via large scale spatial integration. Lastly, the debate around negative emission technologies to enlarge the carbon budget currently focuses on land-based approaches such as Bioenergy with Carbon Capture and Storage. Based on a schematic integration of Direct Air Capture (DAC), we show that its flexibility complements renewable generation variability and can help to integrate large shares of renewables