Task Force Eifel Floods 2021
Germany, the Netherlands, and Belgium were hit by an extreme rainfall event in July 2021 leading to record-breaking peak flows at many gauges with estimated damage in the order of EUR 30 billion for Germany alone. In addition to the meteorological and hydrological dimensions of the flood, the concurrent coupling of landscape elements and the wood, sediment and debris carried by the fast-flowing water made this flood so devastating and difficult to predict. The flood event of July 14 caused flooding, flash floods and landslides in large parts of North Rhine-Westphalia and Rhineland-Palatinate. In these two states, 185 people lost their lives, two people are still missing, and infrastructure and thousands of buildings were destroyed or damaged.
In October 2021, NatRiskChange doctoral researchers and PIs from the University of Potsdam, the GFZ and the Free University Berlin initiated a Task Force on the flood event in July 2021. They worked on five different topics in different working groups, including hydrology, meteorology, geomorphology, impacts and flood warning.
Output by the Task Force Working Groups
Hydrology
Publications:
Brief communication: Impact forecasting could substantially improve the emergency management of deadly floods: case study July 2021 floods in Germany
Floods affect more people than any other natural hazard; thus flood warning and disaster management are of utmost importance. However, the operational hydrological forecasts do not provide information about affected areas and impact but only discharge and water levels at gauges. We show that a simple hydrodynamic model operating with readily available data is able to provide highly localized information on the expected flood extent and impacts, with simulation times enabling operational flood warning. We demonstrate that such an impact forecast would have indicated the deadly potential of the 2021 flood in western Germany with sufficient lead time.
Apel, H., Vorogushyn, S., and Merz, B. (2022): Brief communication: Impact forecasting could substantially improve the emergency management of deadly floods: case study July 2021 floods in Germany, Nat. Hazards Earth Syst. Sci., 22, 3005–3014. https://doi.org/10.5194/nhess-22-3005-2022
Analyse der Hochwassergefährdung im Ahrtal unter Berücksichtigung historischer Hochwasser / Analysis of flood hazard in the Ahr Valley considering historical floods
The flood disaster in July 2021 in western Germany calls for a critical discussion on flood hazard assessment, revision of flood hazard maps and communication of extreme flood scenarios. In the presented work, extreme value analysis was carried out for annual maximum peak flow series at the Altenahr gauge on the river Ahr. We compared flood statistics with and without considering historical flood events. An estimate for the return period of the recent flood based on the Generalized Extreme Value (GEV) distribution considering historical floods ranges between about 2600 and above 58700 years (90% confidence interval) with a median of approximately 8600 years, whereas an estimate based on the 74-year long systematically recorded flow series would theoretically exceed 100 million years. Consideration of historical floods dramatically changes the flood quantiles that are used for the generation of official flood hazard maps. The fitting of the GEV to the time series with historical floods reveals, however, that the model potentially inadequately reflects the flood population. In this case, we might face a mixed sample, in which extreme floods result from very different processes compared to smaller floods. Hence, the probabilities of extreme floods could be much larger than those resulting from a single GEV model. The application of a process-based mixed flood distribution should be explored in future work.
The comparison of the official HQextrem flood maps for the Ahr Valley with the inundation areas from July 2021 shows a striking discrepancy in the affected areas and calls for revision of design values used to define extreme flood scenarios. The hydrodynamic simulations of a 1000-year return period flood considering historical events and of the 1804 flood scenario compare much better to the flooded areas from July 2021, though both scenarios still underestimated the flood extent.
Particular effects such as clogging of bridges and geomorphological changes of the river channel led to considerably larger flooded areas in July 2021 compared to the simulation results. Based on this analysis, we call for a consistent definition of HQextrem for flood hazard mapping in Germany, and suggest using high flood quantiles in the range of a 1,000-year flood. Flood maps should additionally include model-based reconstructions of the largest, reliably documented historical floods and/or synthetic worst-case scenarios. This would be an important step towards protecting potentially affected population and disaster management from surprises due to very rare and extreme flood events in future.
Vorogushyn, S., Apel, H., Kemter, M., Thieken, A.H. (2022): Analyse der Hochwassergefahrdung im Ahrtal unter Berücksichtigung historischer Hochwasser – Hydrologie & Wasserbewirtschaftung, 66, (5), 244-254. https://doi.org/10.5675/HyWa_2021.5_2
Meteorology
Report on the Extreme Rainfall Event 14.07.2021 Rainfall Intensity-Duration-Frequency Analysis
by Jana Ulrich, Felix Fauer , Alexander Pasternack and Henning Rust
Institute of Meteorology, Free University, Berlin
Aim of the report:
We intend to estimate the annual exceedance probabilities (or average return periods) of the mean precipitation intensities over different time intervals that occurred during the event. For this purpose, we employ a duration-dependent generalized extreme value (d-GEV) distribution, to model the distribution of annual precipitation intensity maxima for different accumulation durations. Since in a previous study we have already applied these methods to the Wupper Catchment, which was affected by the event, we initially focus on this area in this report. However, the methods used can be extended to a larger area. We estimate the Intensity-Duration-Frequency (IDF) relationships for different stations in the study area using both a station-based and a spatial model. From this, we can subsequently classify the recorded precipitation intensities in the event. We aim to answer the following question: Based on the previous analyses, what are the estimated annual exceedance probabilities of the maximum precipitation intensities during the event?
Geomorphology
Short Report on the Field Visit to the Ahr Valley (11th-12th October of 2021) by Melanie Fischer
On the 11th and 12th of October 2021, three NatRiskChange core members, Heather Murdock and Seth Bryant of the third as well as Melanie Fischer of the second PhD-student cohort, accompanied Dr. Sigrid Roessner of the GFZ on a reconnaissance field visit to the Ahr Valley. The scientists from Potsdam joint a team of sedimentologists from the University of Gießen, who collected flood sediment samples for luminescence dating at several points along the entire course of the Ahr. Starting from its confluence with the Rhine near Sinzig, the team followed the Ahr upstream towards its source at Blankenheim in the Eifel mountains. Along the way, floodmarkers, piles of wood debris, and the extent of destruction of houses as well as infrastructure attested to the intensity of the flood and left lasting impressions. On the second day in the field, the team visited a number of landslide locations along the upper reach of the Ahr and smaller tributary streams in the Eifel catchment. The group eventually split in the afternoon and while Heather Murdock and Seth Bryant remained with Sigrid Roessner, Melanie Fischer helped Dr. Michael Dietze (GFZ) and Dr. Rainer Bell (University of Bonn) in installing geophones in two potentially instable hillslopes near Antweiler and Müsch. The two-day visit to the Ahr Valley gave the NatRiskChange-Phd students the opportunity to gain a comprehensive overview of the impacts of the July 2021 event, considering both geomorphic processes in the upper reach and catchment as well as the alarming scale of disruption of settlements and vital infrastructure along the river’s middle and lower reaches.
Related publications:
More than heavy rain turning into fast-flowing water – a landscape perspective on the 2021 Eifel floods
Rapidly evolving floods are rare but powerful drivers of landscape reorganisation that have severe and long lasting impacts on both the functions of a landscape’s subsystems and the affected society. The July 2021 flood that particularly hit several river catchments of the Eifel region in West Germany and Belgium was a drastic example. While media and scientists highlighted the meteorological and hydrological aspects of this flood, it was not just the rising water levels in the main valleys that posed a hazard, caused damage, and drove environmental reorganisation. Instead, the concurrent coupling of landscape elements and the wood, sediment and debris carried by the fast-flowing water made this flood so devastating and difficult to predict. Because more intense floods are able to interact with more landscape components, they at times reveal rare non-linear feedbacks, which may be hidden during smaller events due to their high thresholds of initiation.
Dietze, M., Bell, R., Ozturk, U., Cook, K. L., Andermann, C., Beer, A. R., Damm, B., Lucia, A., Fauer, F. S., Nissen, K. M., Sieg, T., and Thieken, A. H. (2022): More than heavy rain turning into fast-flowing water – a landscape perspective on the 2021 Eifel floods, Nat. Hazards Earth Syst. Sci., 22, 1845–1856, https://doi.org/10.5194/nhess-22-1845-2022
Dietze, M. and Ozturk, U. (2021): A flood of disaster response challenges, Science, 373, 1317–1318, https://doi.org/10.1126/science.abm0617
Impacts
Report by Dr. Tobias Sieg
The flood event of July 14, 2021 caused flooding, flash floods and landslides in large parts of North Rhine-Westphalia and Rhineland-Palatinate. In these two states, 185 people lost their lives, two people are still missing, and infrastructure and thousands of buildings were destroyed or damaged.
This task force group has estimated flood damage to homes and businesses in six hard-hit counties. Based on flood masks from the Copernicus Emergency Management Service (emergency.copernicus.eu) and the German Aerospace Center (zki.dlr.de), we applied two different models to estimate damage to residential buildings and contents, as well as company buildings and their furniture or equipment. The models applied are M. DELENAH (Natho and Thieken, 2018) and Random Forests (Sieg et al., 2019). M. DELENAH is a simple model for calculating direct monetary losses from natural hazards and is based on the UNISDR method (UNISDR, 2015, now UNDRR, United Nations Office for Disaster Risk Reduction). It gives a rough initial estimate using as little data as possible. The Random Forest models were trained with survey data from affected people on flood damage, preparedness, water levels, etc., collected after past flood events in 2002 and 2013 (Kreibich et al., 2005, 2007; Thieken et al., 2016).
We ultimately decided not to publish the estimates due to high uncertainties in the input datasets. The damage estimates in this report are for buildings identified as impacted by flood masks. These flood masks were obtained from two different sources and had not been separately validated at this time. Consequently, the actual areas and buildings affected may differ from those identified by the flood masks. However, the number of affected buildings has a large impact on model results, so the most reliable estimate of affectedness is necessary to obtain reliable damage estimates (Sieg & Thieken, 2022). This was not given at the time of the task force. Other reasons for uncertainty are underestimations of damage models not explicitly designed for the mix of different natural hazards that occurred (river floods, flash floods, landslides), and uncertainties in setting monetary values in affected areas. However, a comparison with data from the German Insurance Association (GDV) showed that the inaccurate water masks accounted for the largest part of the underestimations in this case.
Nevertheless, the activities have led to some insights and possible next steps. The RF models are more elaborate to apply but give a more detailed look at uncertainties, while M. DELENAH is relatively quick to apply due to its structure with limited data. Copernicus Rapid Mapping provides data on affected areas and buildings quickly after an event. However, analysis to date shows that there are remarkable gaps in these inventories. A quick solution may be to combine water masks from different sources to summarize as much available information as possible. Coupling different aerial systems (e.g., by drone flights) could merge recordings directly during the event with later recordings and add content. In the long term, it would be useful to develop methods that reliably summarize and represent different hazards (floods, flash floods, landslides). However, damage models must keep pace with this development and also be able to reliably estimate damage from different hazards. The added value for affected citizens and counties would be faster clarity on the extent of damage shortly after an event and, accordingly, the approximate amount of emergency aid. Additionally, more accurate models can also significantly facilitate the planning of possible protective measures.
Work on the possibilities of remote sensing data in flood damage estimation will be continued as part of a project for the Federal Office of Civil Protection and Disaster Assistance (BBK) starting in October 2022.
Flood warning
Preprint:
Performance of the flood warning system in Germany in July 2021 – insights from affected residents
In July 2021 intense rainfall caused devastating floods in Western Europe and 184 fatalities in the German federal states of North Rhine-Westphalia (NW) and Rhineland-Palatinate (RP) questioning their flood forecasting, warning and response system (FFWRS). Data from an online survey (n = 1315) reveal that 35 % of the respondents from NW and 29 % from RP did not receive any warning. Of those who were warned 85 % did not expect a very severe flooding and 46 % did not know what to do. Regression analysis reveals that this knowledge is influenced by gender and flood experience, but also by the contents and the source of the warning message. The results are complemented by analyses of media reports and official warnings that show shortcomings in providing adequate recommendations to people at risk. Dissemination of warnings, communication of the expected flood magnitude and adequate responses are seen as entry points for improving the FFWRS in Germany.
Thieken, A. H., Bubeck, P., Heidenreich, A., von Keyserlingk, J., Dillenardt, L., and Otto, A. (2022): Performance of the flood warning system in Germany in July 2021 – insights from affected residents, Nat. Hazards Earth Syst. Sci., 23, 973–990, https://doi.org/10.5194/nhess-23-973-2023