|Paper title||A new Approach to Hazard Analysis of Heavy Rainfall Events based on the Catchment Area of the Ahr River|
|Form of presentation||Poster|
In this summer a long time stationary rain event struck parts of western Germany leading to massive floodings – especially in the valley of the Ahr approximately 20 km south of Bonn. Such long-term stationary weather conditions get actually more and more frequent and can lead to long extreme heat or massive continous rainfall as shown in a study of the Potsdam-Institut für Klimafolgenforschung (PIK) this year.
The flood of the Ahr revealed that the existing modelling for flood probabilities is not sufficient. Possible causes may be the comparatively short observation period of the underlying measurements, missing historical data or the dynamics of climate change are not taken into account. For this reason, our approach is based on simulations of individually adapted worst case scenarios to derive possible effects of heavy rainfall more generally and over a wide area.
In the last years we developed a methodology for classification of strong rain dangers depending only on the terrain. We calculated strong rain danger maps covering hole Germany and Austria estimating a worst case scenario by not taking into account local drains since those are mostly blocked by leaves and branches at such sudden events. But these maps are only based on the influence of the direct surrounding in strong rain events and do not consider water coming from other areas. So we developed an additional component for including water-run-off from up-stream areas.
In the presented study we calculate the maximum run-off for a whole water catchment area assuming a massive strong rain event and the following flash flood. For each position in the run-off-map a local height profile perpendicular to the flow direction is calculated and filled up with the maximum estimated water volume at this position. So cross sections along a river in a valley giving a maximum water level for the maximum possible run-off for a given strong rain event are derived.
Since some part of the rain will drain away and not contribute to the run-off this is also a worst case estimation. The results are compared to aerial imagery acquired on 2021-07-16 – two days after the flooding struck the Ahr valley –, flood-masks derived from Sentinel-1 imagery and Copernicus damage assessment maps. Based on this imagery and measurements and estimations of water gauge levels we calculate the effective rain-height of the catchment and the simulation is calibrated and adapted to the observed water levels. Based on these results we can derive also an estimation of the flooding situation in the whole catchment area including tributary valleys.