Research

The scientific work packages (WP) beam the ‚risk chain – interactions – dynamics‘ spotlights on the ‚Atmosphere-Catchment-System‘ (WP1), ‚River-Dike-Floodplain System‘ (WP2) and the ‚Socio-Economic System‘ (WP3).

WP1: Spotlight on the Atmosphere-Catchment System (ESR1 – ESR5)

The quantification of flood hazard and risk in a pan-European context relies on the sound representation of the interactions of the atmosphere-catchment system. For instance, flood hazard patterns depend on the in-phase or out-of-phase seasonality of precipitation and soil moisture dynamics. Further, extreme floods are typically a consequence of an unfavourable interplay of different factors.

A remarkable example is the June 2013 flood in Central Europe with intensive and persistent precipitation in combination with very wet catchments. Another example is the flooding across the UK over the winter 2013/14, generated by the superposition of continuous heavy rainfall and coastal surges. Understanding the interplay of atmospheric and catchment processes is also critical for the evaluation of climate change impacts, such as changes in spatial patterns of flood types and flood regimes.

WP1 aims at understanding how the interplay between atmospheric and catchment processes influences flood regimes, the occurrence of extreme floods, and temporal flood changes. This knowledge will be used to develop methods for deriving consistent large-scale flood scenarios for present and future climates.

By developing very high resolution realizations of future weather events and by tailoring climate simulations to the flood problem, more realistic future flood scenarios including unprecedented and compound events, such as the superposition of storm surges and heavy rainfall or the clustering of events, will be constructed (ESR1).  This simulation based approach will be complemented by a data based approach, where decadal changes in the flood characteristics of several hundred catchments across Europe will be analysed and attributed to atmospheric and/or catchment processes (ESR2). A large-scale weather generator capable of representing the spatial correlation structure of meteorological variables over large domains and retaining at the same time local-scale variability will be developed (ESR4).  Its output in the form of large-scale meteorological fields will be fed into hydrological models by several ESR projects to obtain a sound representation of joint probability of flood peaks across large areas. Developing simplified conceptualizations of the interplay of atmospheric and catchment processes, methods for flood hazard mapping and for estimating extreme floods at large scales will be developed and tested jointly with river basin authorities (ESR3).  Extending the focus and including socio-economic factors, such as land use change and mitigation measures, in the assessment of flood changes, will help to understand how the dynamics of human systems affects flood-hazard patterns, and how the dynamics of flood hazard affects patterns of vulnerability and exposure (ESR5).


WP2: Spotlight on the River–Dike–Floodplain System (ESR6 – ESR10)

Flood hazard and risk assessments consistent across regions are required by the EU Flood Directive, by national programmes in non-European countries, e.g. USA, Australia, and in the insurance industry for the management of insurance portfolios. The EU Flood Directive appeals to consider the solidarity principle where spatiotemporal risk changes should be considered when implementing risk reduction measures. Downstream communities should not suffer from or should be compensated for increasing risk due to measures upstream.

Most methodologies for large-scale flood hazard and risk mapping do not go beyond a simple mosaicking of local scale inundation extents/depths and damages. This approach of merging local-scale maps into a large-scale picture violates fundamental characteristics of the spatial dependence structure of flooding. For instance, a flood with a uniform return period of 100 years over the entire river basin may be implausible or would have a much larger return period than 100 years. Furthermore, the performance of flood defences, the interdependence of their failures and cascading effects significantly shape the spatial distribution of flood risk.

WP2 develops new methods for spatially consistent, large-scale risk assessment focusing on interactions and the temporal dynamics in the river-dike-floodplain system. This includes the development of large-scale 1D-2D coupled hydraulic modelling approaches which are capable to continuously simulate flooding in the entire river network over long time periods (decades of simulation periods), using smart simplified probabilistic/deterministic techniques with risk importance sampling. Coupled hydrology and hydraulic models taking into account process interactions in catchments and in river networks will make it possible to generate largescale hazard and risk statements which do not violate the space-time characteristics of flood risk systems. Threshold processes patterns such as dike failures which can abruptly change spatial risk will be considered by dike fragility curves, derived from many simulations of dike breach models, giving the probability of dike breach as a function of hydrological loads.

A Europe-wide probabilistic flood hazard chain will be developed by combining models of catchment hydrology and floodplain inundation at 100 m resolution (ESR9). At the scale of large river basins, existing tools will be enhanced to be able to consider flood defence failures and hydrodynamic interactions which shape the spatio-temporal patterns of flood risk (ESR6 with focus on large river deltas, ESR7 with focus on large inland rivers). These risk assessment tools will be applied to evaluate risk mitigation measures by comparing different decision criteria, such as cost-benefit ratio, robustness or flexibility. The results will serve as a basis for policy recommendations, placing emphasis on the implications of uncertainty for decision making (ESR8). Cascading effects of flooding on infrastructure (transport, supply chains etc.) will be analysed, and a policy analysis will evaluate beneficial and inhibitive factors for implementing policy options and climate change adaptation measures (ESR10).


WP3: Spotlight on the Socio-Economic System (ESR11 – ESR15)

Flood risk management in Europe has been broadening and delivery can be through a large suite of measures implemented by both the public and private sector. This diversification requires a profound understanding of how the socio-economic system interacts with the physical processes in the atmosphere, catchments and river systems. There is a need for a more generalised understanding of coupled system dynamics of human and water systems. An example is the question of how different risk reduction measures combine or conflict with each other.

Investigations on the interconnections between different measures are scarce, and developing an approach for analysing how different options work together to tackle the social and economic consequences of flooding will assist in identifying more effective portfolios of measures. The aim is to design portfolios of risk reduction measures – and not separate measures in isolation. The quantitative appraisal of risk reduction, however, is plagued by crude damage estimation methods.

Typically, direct damage to objects or land use units is estimated using only inundation depth as a flood impact variable. Recently, the high-dimensional flood damage database of GFZ has been used to develop multi-parameter models, and it has been demonstrated that those models improve flood damage assessment. What is missing, however, is the inclusion of temporal changes in damage models. Another research direction concerns higher order losses which are often neglected in damage estimation. Large uncertainties reside in assessing the indirect costs of disasters due to the lack of data and inadequate methodologies.

A fundamentally new approach is to address the problem as a complex system by considering the impacts on flows and by better integrating the time dimension. Consideration of long-term and broad scale resilience would permit us to look more widely and to consider indirect, longer-term cascading consequences and impacts, both outside and inside the affected areas. WP3 aims at gaining an improved understanding of the how the socio-economic system interacts with physical processes. Further, it will develop novel methods for estimating the socio-economic consequences of flooding and for designing risk management strategies.

Focusing on private precaution as a main determinant of vulnerability of households, past changes in private precaution and their drivers will be analysed, and time-varying damage models will be developed (ESR13). The hypothesis is that this will not only improve flood damage assessment, but will help to unravel the role of changes in vulnerability to long-term changes in flood risk. This will be an important input to attributing risk changes to the contributing causes via Bayesian fingerprinting methods (ESR14). By exploring the use of agent-based modelling and general systems modelling, a better understanding of the critical determinants of potential ripple effects and indirect losses following an extreme event will be sought (ESR12). It is important to understand how different approaches to risk reduction work in combination, such as the balance between pro-active resilience measures and recovery and how these can be effectively integrated into delivering sustainable and effective flood management policy. The interactions and conflicts between portfolios of approaches will be examined and a framework for understanding how approaches combine to improve resilience to flooding will be developed (ESR11). These analyses will be bracketed by studying the interactions of floods and human dynamics at the centennial time scale via simplified coupled models (ESR15).