Assessing the impact of early warning and evacuation on human losses during the 2021 Ahr Valley flood in Germany using agent-based modelling

André Felipe Rocha Silva, Julian Cardoso Eleutério, Heiko Apel, and Heidi Kreibich

Supplement

https://doi.org/10.5194/nhess-25-1501-2025-supplement

Data sets

Germany 2020 -- Land cover based on Sentinel-2 data Mundialis https://www.mundialis.de/en/germany-2020-land-cover-based-on-sentinel-2-data/

Pan-European exposure maps and uncertainty estimates from HANZE v2.0 model, 1870--2020 D. Paprotny https://doi.org/10.5281/zenodo.7885990

Population in brief (territory on 9 May 2011) Statistische Ämter des Bundes und der Länder https://ergebnisse.zensus2022.de/datenbank/online/statistic/1000X/details

Model code and software

Institute for Water Resources US Army Corps of Engineers https://www.rmc.usace.army.mil/Software/LifeSim/

Short summary

This work uses agent-based modelling to evaluate the impact of flood warning and evacuation systems on human losses during the 2021 Ahr Valley flood in Germany. While the first flood warning with evacuation instructions is identified as timely, its lack of detail and effectiveness resulted in low public risk awareness. Better dissemination of warnings and improved risk perception and preparedness among the population could reduce casualties by up to 80 %.