Attributing the July 2021 Ahr flood across the hydrological impact chain

We perform a model-based end-to-end attribution of the July 2021 flood in the Ahr catchment to assess how anthropogenic climate change affected the likelihood of the event from extreme precipitation to direct residential building losses. The analysis compares a present-day (factual) and a pre-industrial (counterfactual) climate using an unconditional event attribution approach based on a Regional Flood Model (RFM) that links meteorological forcing, hydrological response, inundation characteristics, and impacts. Synthetic precipitation is generated with a non-stationary weather generator conditioned on large-scale circulation patterns and regional mean temperature and is used to drive the hydrological model mHM. Extreme flood hydrographs are routed with the hydrodynamic model RIM2D to derive flood depths and inundation extent, which are combined with exposure information in a flood loss model to estimate direct residential losses. Event likelihoods are derived consistently along the flood impact chain, and probability ratios between the factual and counterfactual climates are used to quantify climate change influence.

Results show that daily precipitation extremes comparable to the July 2021 event are about 1.2 times more likely in the current climate, while corresponding flood peaks are about 1.6 times more likely. The likelihood of inundation impacts increases more moderately, with probability ratios of 1.3 for average maximum inundation depth and 1.25 for inundation extent, and the likelihood of high residential building losses shows a probability ratio of 1.15. Overall, these synthesis results indicate that climate change has increased the likelihood of an event such as the July 2021 flood, with differing amplification factors reflecting the nonlinear transformation of climate-driven changes by hydrological, hydraulic, and damage processes acting on different spatial scales and response time-scales along the flood impact chain.