Compound pluvial and fluvial flooding (CPFF) represents a major and growing threat to urban areas, yet its risk remains systematically underestimated by conventional flood management practices. This is largely due to a critical methodological gap: the prevailing approaches treat pluvial (surface water) and fluvial (riverine) floods in isolation, failing to capture their complex interactions and the resulting amplification of hazard. This dissertation aims to address this gap by advancing the methodological foundation for CPFF risk assessment. Its primary goal is to develop an integrated modelling framework capable of probabilistically quantifying CPFF hazard and damage amplification, moving beyond simplistic scenario-based assumptions. To achieve this, the thesis begins with a systematic literature review, which identifies a poor understanding of dependence structures between flood drivers and a near-total absence of studies quantifying CPFF damage as the core research gaps. In response, a comprehensive model chain is developed and applied to the Ahr Valley, Germany, a region recently devastated by a catastrophic flood event. The framework couples a non-stationary Regional Weather Generator (nsRWG) with a Circulation Pattern-conditioned disaggregation model to generate long-term, high-resolution synthetic precipitation data that intrinsically captures the spatial and temporal dependence. This weather forcing subsequently drives a hydrological model (mHM), a 2D hydrodynamic model (RIM2D), and finally a damage model (FLEMOflash) to complete the pathway from extreme weather to consequence. This research provides a process- and simulation-based foundation for understanding the amplification effect of CPFF hazard and damage. The weather generator proves highly capable of reproducing the cross-scale extremity of observed heavy precipitation across Germany. The disaggregation model significantly improved the simulation of sub-daily rainfall extremes, which is crucial for triggering pluvial floods. The application of the full integrated chain revealed that CPFF hazard maps differ drastically from official single-driver maps, showing synergistic interactions that expand inundation extents and amplify water depths. This translated into a severe amplification of flood damage, with impacts for frequent events proving to be an order of magnitude greater than those of a constituent fluvial-only or pluvial-only scenario. The results also show that catastrophic damage, equivalent to a rare 100-year fluvial flood, can be caused by a CPFF resulting from a combination of moderate events, such as a 40-year discharge and an 8-year rainfall. In conclusion, this thesis makes an original contribution to the field by providing an integrated framework for probabilistic CPFF risk assessment from weather forcing to economic impact. It delivers quantitative evidence of severe damage amplification in a German case study, challenging conventional flood risk management. The finding that extreme damage is driven by the co-occurrence of moderate events, not just rare extremes, has profound implications for risk communication, infrastructure design, and adaptation planning. The developed framework not only provides a tool for reassessing current risk but also opens the pathway for analyzing past trends and future projections of CPFF in a changing climate, thereby enabling a more resilient and proactive approach to flood risk management.
Xiaoxiang Guan (Thu,) studied this question.