Authors: Wouter Vanneuvillea, Wim Kellensb, Philippe De Maeyerb, Genserik Reniersc,d & Frank Witloxb
aFlemish Authorities, Flanders Hydraulics Research, Berchemlei 115, 2140 Antwerp, e-mail: Wouter.Vanneuville@mow.vlaanderen.be
bGhent University, Department of Geography, Krijgslaan 281 (S8), 9000 Ghent, e-mail: Wim.Kellens@UGent.be ; Philippe.DeMaeyer@UGent.be ; Frank.Witlox@UGent.be
cUniversity of Antwerp, Faculty of Applied Economic Sciences, ARGoSS, Prinsstraat 13, 2000 Antwerp, e-mail: Genserik.Reniers@ua.ac.be
dHUB, Faculty of Economics and Management, CEDON, Stormstraat 2, 1000 Brussels
Related to natural hazards and, in particular, floods, the term ‘risk’ is defined in numerous ways and is often confused with the term ‘disaster.’ This article describes the Flemish flood risk methodology as an approach to tackle flood risks, which are defined as combinations of hazards and vulnerabilities. Furthermore, the relationship between the Flemish flood methodology and the integrated approach (economic damage, human health, ecology and cultural heritage) from the EU Floods Directive, is given. Finally, the concept of the flood risk management cycle and its components is compared with the management of flood disasters.
Keywords: flood risk management, risk assessment, European Floods Directive, flood disaster management
Flood hazards are the most common and destructive of all natural disasters. Each year, flood disasters cause tremendous losses and social disruption worldwide. In recent years, risk-based approaches have received increasing attention as means to ‘manage’ flood hazards. In day-to-day language, the term ‘risk’ is often used as a synonym for probability or chance. This should not be a surprise since the denotation of the term often varies with the sector in which it is applied (e.g. in legal professions, insurance, natural disaster research communities, cultural heritage researchers, etc.) (Coste, 2001). Yet, even in a flood context alone, numerous definitions for flood risk have been suggested. In their ‘Language of Risk’-Report, Samuels and Gouldby (2009, p. 21) recommend to employ the term flood risk as “the probability multiplied by the consequence in which the multiplication is to be understood as including the combination across all floods.” The definition of a disaster is even more vague. In the glossary of the UN International strategy for Disaster Reduction (UN-ISDR, s.d.) ‘a disaster’ is defined as “a serious disruption of the functioning of a community or a society causing widespread human, material, economic or environmental losses which exceed the ability of the affected community or society to cope using its own resources.”
The definitions of ‘flood risk’ and ‘disaster’ indicate that, although related to each other, the terms are not synonymous. Lumbroso (2007) differentiates risk from disaster in terms of impact. Whereas risk refers to any consequence that can be measured, a disaster denotes a large or even catastrophic event. Hence, both terms ask for a different way of management. This paper presents an overview of the functioning and the requirements of flood risk and flood disaster management, with specific attention to the state-of-the-art in Flanders (Belgium). In Belgium, water management, including flood risk management, is a responsibility of the different regions (Flanders, Wallonia and Brussels) for the larger rivers and a responsibility of the provinces and municipalities for the smaller river courses.
Since they are imperative to flood risk management, flood risk calculations are discussed in the next section.
2. Flood risk calculations
A recent overview of the flood risk calculation methodology for Flanders can be found in Kellens et al. (2008) and Deckers et al. (2009). The scheme (cfr. Figure 1) in Kellens et al. (2008) elucidates the type of data that is needed to calculate flood risks.
It is beyond the scope of this paper to discuss the hydrodynamic software that is needful to create flood maps. However, their importance should not be neglected, since detailed flood maps, indicating the flood extent, the water depth and (where relevant) flow velocity and rising velocity, are essential to produce meaningful calculations on the consequences of flooding.
2.1. Digital Elevation Model
The Digital Elevation Model (DEM) plays a key role in the flood risk methodology. For Flanders, a DEM (“DHM Vlaanderen”) was created on initiative of regional water management authorities together with the Flemish agency for geographic information (AGIV) between 2001 and 2004. It is based on laser scanning in rural areas with basically 1 point per 4 m2 during recording. After filtering out buildings and vegetation, 1 point per 20 m2 remains in the final Triangular Irregular Network (TIN) file.
For urban areas, photogrammetric interpretation based on stereoscopic aerial photos is used, resulting in 1 point per 100 m2. Additional break lines are added to compensate for the lower point density. This high density of elevation points is needed, as the topography is rather flat in a large part of the territory. The local variations have to be in the model to accurately delineate the flooded area. Local embankments or quarries not immediately included in the elevation model create over- or underestimation of the flood zone.
2.2. Land-use information
A detailed and uniform land-use map is an important prerequisite to perform flood risk calculations, since it determines what is damaged in case of flooding (De Maeyer et al., 2003). In the Flemish methodology, various land-use categories are distinguished (e.g., urban area, industrial area, infrastructure, crop land, pastures, etc.) and further subdivided according to their vulnerability level, which is related to the desired level of detail: A high level of thematic detail is needed where the damages are high or highly varying (e.g., built-up area, industry and infrastructure) and less where the (economic) damages are low (e.g., different types of pastures of nature areas). The vulnerability of the land-use categories is determined by various socio-economic data (e.g., replacement values).
With respect to spatial distribution and update frequency of land-use data, airborne and space borne data sets have been found promising in the past. Results from the FAME project showed that Landsat images, as compared to IKONOS (Willems et al., 2003), offered less thematic and spatial detail, but allowed to use interpretation keys for the entire territory of Flanders. Therefore, the CORINE Land Cover (which is based on Landsat imagery) was found most suitable in the early stages of the Flemish risk methodology. Recently, cadastral information is being employed to calculate the damage of individual houses and industrial sites resulting in a higher level of detail (Vanderkimpen et al., 2010).
It is important to note that the detailed methodology used in Flanders is currently not applicable on an international river basin. Each country collects socio-economic information differently and applies interpretation keys of land-use information in a (slightly) different way (e.g. Verwaest et al. (2008) for the countries along the North Sea coast). However, there is improvement in sight. As part of the AMICE project, the international catchment of the Meuse is used as a test case to develop overarching interpretations of the CORINE Land Cover in addition to the national and regional interpretations.
2.3. Flood risk calculation variables
The final stage comprises the combination of the flood maps (i.e., hazard maps) with the land-use maps (i.e., vulnerability maps). So-called stage-damage curves are often used to define a relation between the water depth (or another flood characteristic) and the expected damage for each specific land-use category. Many functions have been proposed in the past, either on theoretical grounds, or on empirical grounds (e.g. Van der Sande 2001, Kok et al. 2002, Penning-Rowsell et al. 2005). After the recent floods in Belgium (in November 2010 and in January 2011), it makes sense to set up an inquiry to calibrate the functions used in Flanders.
A specific issue regarding the applicability of stage-damage curves is the limited validity of the economic information. The same damage can be expressed as replacement values, or depreciated costs. Moreover, the same goods can have a different value due to taxes and insurance systems, making values and stage-damage functions not transferable to other countries automatically. These differences are worked out in the Interreg IIIb North Sea project SAFECoast (Verwaest et al. 2008) for countries in the North Sea Region.
As for land-use, these differences in water depth-damage relations make a comparison between different regions complex and the possibility exists that the uncertainties are dominant in people’s perceptions, even when the conclusions are robust. Additional calculations with common stage-damage functions may be needed and the Joint Research Centre in Ispra (Italy) is elaborating such functions (Barredo & De Roo 2010).
In the final phase of the methodology, the set of damage maps is integrated into a single risk map (cf. Figure 1). Each damage map can be seen as a scenario to which a weight is attached, based on the scenario’s return period. The Flemish methodology employs different corrections for these factors to compensate for the limited number of modelled flood maps. Whereas the damage maps represent the consequences for each scenario, the flood risk map expresses the average damage per year.
3. Flood Risk Management
Flood risk management comprises more than flood risk calculations. Figure 2 presents the Lumbroso framework for flood risk management, which is close to the terminology of the EU Directive on the assessment and management of flood risk (EU COM, 2007, further referred to as the “Floods Directive”). While The Floods Directive states that (Article 7 §3): “Flood risk management plans shall address all aspects of flood risk management focusing on prevention, protection, preparedness, including flood forecasts and early warning systems,” Lumbroso (2007) did not implement “protection” in his framework. However, this phase can be located between the prevention and preparation phases.
As denoted in Article 7, §2, the Floods Directive focuses on the reduction of potential adverse consequences of flooding with regard to human health, the environment, cultural heritage and economic activity. Until now, the majority of the flood risk calculations only considers economic losses. Social cost-benefit analyses are ideal instruments to bring together the tangible (economic losses) and intangible elements (social and ecological aspects). For example, the revised Sigma plan, a master plan for flood safety along the river Scheldt, has implemented such analyses to evaluate different scenarios (e.g., heightening of river embankments, controlled overflow zones and reduced tidal areas, etc.).
It is of paramount importance to work with a common framework (using monetary values or another quantified method) to evaluate the effects of different scenarios. Recently, more information is becoming available to directly or indirectly valuate the human and social aspects, the ecology and the cultural heritage. As part of the European FloodResilienCity project, Flanders tried out an integrated approach in the area of Leuven and the Woluwe catchment. Such exercises are valuable to meet the objectives of the Floods Directive in terms of flood risk management plans, but the number of scenarios worked out in the case studies is still too limited to derive conclusions yet. Moreover, these exercises show that the complete process is time-consuming because of the participatory approach. This approach involves the organization of several workshops in which different aspects of flood risk management planning are explained to the participants. These participants are mainly professionals and experts, but in the future, public citizens will be involved as well.
Preamble 17 of the Floods Directive states that synergies between – and benefits of – the Water Framework Directive (EU COM, 2000) and the Floods Directive have to be found. Hence, adequate coordination of the objectives and of the measures proposed in both plans are required. Flanders is surpassing the requirements by combining the River Basin Management Plans (for the Water Framework Directive) and the Flood Risk Management Plans of 2015 into one plan (one document). The legal framework to do so was already set up by the Decree of Integrated Water Policy of 18 JUL 2003 (Belgisch Staatsblad, 14 NOV 2003) where aspects of flooding were already taken into account. The first generation of the River Basin Management Plans and the Programme of Measures (CIW 2010) already contains a group of flood measures.
4. Flood disaster Management
Disaster management can be regarded as the reaction on an event, which is indicated by the response and recovery phase in the scheme of Lumbroso (2007) (see Figure 2). Whereas flood risk management applies to a whole range of events, flood disaster response tries to minimize the effects from just one specific flood disaster.
We demonstrate the difference between flood risk and disaster management with the following example. A flood risk equal to zero does not exist by definition. Even a null flood risk for mortal victims is fictitious since an exceptional – and unforeseen – event is always possible. Yet, as the probability decreases, the contribution to the overall flood risk drops below a point where (according to a social cost benefit analysis) additional measures are not justifiable any more. This is opposite to the way of thinking in disaster management. In case a flood disaster occurs, every possible means will be used to minimize its impact.
Contrary to flood risk management where statistical information is needed, real-time information becomes vital in this context. Disaster response is defined by the actual situation, which means that the actual number of present people will define how much evacuations are needed and not the average number of inhabitants over a year. Likewise, weather patterns of today and the next days will define the actions and not the statistical weather over a (longer) time frame. As such, discussing the possible effects of demographic evolutions, land-use changes and climate change makes less sense in a context of disaster management than in flood risk management, where changing probabilities due to long-term evolutions are highly relevant in a social cost benefit analysis.
Overall, it makes no sense for flood disaster management to use the same complex maps as for flood risk management. Flood risk maps denote the impacts on human health, economic activity, ecology and cultural heritage, and which take different scenarios into account using detailed land-use data sets and stage-damage curves. On the contrary, unambiguous maps with clear key elements are of much more interest to flood rescue and disaster teams. Since the aim and the end-user differ between both management types, the use of tailor-made products is preferred. Geographic Information Systems and other digital techniques offer the opportunity to reach this goal.
Obviously, the picture is not black and white. The fact that there seems to be strong interconnectedness between flood risk and flood disaster indicates an overlap between both concepts to some extent. Moreover, flood disaster response and recovery needs preparation, which is supported by the outcomes of flood risk management. Nowadays, Flemish water managers are preparing their flood risk management plans to report under the EU Floods Directive. Their contacts during the recent floods with fire brigades and the Civil Defence corps make them aware of the operational requests for disaster managers. Operational flood maps will be made together. And more than in the past disasters teams in the provinces ask for input from the water management authorities when updating evacuation plans and specific flood disaster plans.
Due to the structure of the Belgian federal state, regional and federal governments have different responsibilities – respectively for water management (including flood risk management) and disaster handling (including crisis management). As one of the responsibilities cannot be handled without understanding the role of the other actors, having a ‘common language’ or ‘interpretation key’ is essential. There is no need for a totally shared terminology (as long as the meaning is clear for all stakeholders). In the present situation, the same terminology is only common in name, not in meaning.
‘Flood risk’ and ‘flood disaster’ are closely related concepts and the context (i.e., their definition) is necessary to understand what is meant exactly. A distinction between the two concepts cannot be made solely based on required data. Both concepts can be described in detail, where specific and local information is essential. For flood risk management, knowledge of local damage during past floods to update stage-damage functions for better risk evaluation of scenarios can be seen as an example. For flood disaster management, details for example entail the exact number of people and their possibilities in an elderly house during an evacuation.
Having an overview, as well as different levels of detail in the information, is needed and the comparability of data becomes essential. During the planning phase, flood risk managers want to compare their data with others in the same international river basin district or along the same coastline. Then rougher information can bring added value, not as an absolute truth but as a possibility to focus on the robustness of information and the elements that are known rather than on the unknown information. Space-borne data can be of assistance as they cover larger areas and are not restricted to administrative boundaries. A European land-use map, such as CORINE Land Cover, can be too rough for local analysis but can add big value if the objective is to compare situations in different countries.
The same is true during flood disaster response. Summarized information helps the high-level decision-makers in their strategic choices. While hazard data rainfall radars give an impression of the rainfall over the whole area, however, such radars provide less accurate information than gauging stations. At the vulnerability side, an overview (and selection) of hotspots (e.g. schools, hospitals and elderly houses) is needed.
In recent years in flood research communities, the concept of risk as a function of hazards and vulnerabilities or a combination of probability and consequences is extended. Often the word ‘resilience’ appears in the definition. Without focussing on the definition, the term resilience is integrating a variety of dimensions (e.g. social, socio-cultural-historical, legal-institutional, political and economic characteristics), differing with the flood type and the degree of awareness as well as with the uncertainties and the way they are dealt with (Schwindt & Thieken, s.d.). This concept can bridge flood risk management and flood disaster management. An increase of the resilience (also referred to as ‘adaptive capacity’ or ‘coping capacity’) is a ‘before the event’ action to minimize or decrease the negative consequences after the flood disaster.
AGIV (s.d.) DHM Vlaanderen, http://www.agiv.be/gis/projecten/?catid=35 (latest consultation on March 7, 2011)
AMICE (s.d.) Adaptation of the Meuse to the Impacts of Climate Evolutions, Interreg IV B North-West Europe project, http://www.amice-project.eu (latest consultation on 7 MAR 2011)
Barredo, J. & De Roo, A. (2010) Flood damage functions for EU member states – Flood risk mapping using Corine land cover datasets, Presentation on CIS Working Group F Thematic Workshop Floods and Economics Appraising, prioritising and financing flood risk management measures and instruments, Ghent, 25 – 26 OCT 2010.
Bossard, M. ; Feranec, J. & Otahel, J. (2000) CORINE land cover technical guide – Addendum 2000, Technical Report n°40, European Environment Agency, http://www.eea.europa.eu/publications/tech40add
CIW (2010) Maatregelenprogramma voor Vlaanderen, enacted by the Flemish Government on 8 OCT 2010, Coördinatiecommissie Integraal Waterbeleid, http://www.ciwvlaanderen.be/stroomgebieddistricten/vlaams/sgbpen/maatregelenprogramma_def (latest consultation on 7 MAR 2011).
Coste, L. (2001) Sur plusieurs définitions du risque, et leurs conséquences en matière de prévision, prédiction et prévention, La Houille Blanche, n° 2
Deckers, P.; Kellens, W.; Reyns, J.; Vanneuville, W. & De Maeyer, Ph. (2009). A GIS for flood risk management in Flanders, in: Showalter, P.S. et al. (Ed.) (2009). Geospatial techniques in urban hazard and disaster analysis. Geotechnologies and the environment, 2: pp. 51-69
De Maeyer, Ph.; Vanneuville, W.; Maeghe, K. & Mostaert, F. (2003) Modélisation des effets de crue dans le bassin de la Dendre, basée sur une méthodologie de crue. [S.n.]: Paris
EU COM (2000) Directive 2000/60/EC of the European Parliament and of the council of 23 OCT 2000 establishing a framework for Community action in the field of water policy, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2000:327:0001:0072:EN:PDF
EU COM (2007) Directive 2007/60/EC of the European Parliament and of the Council of 23 OCT 2007 on the assessment and management of flood risks, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2007:288:0027:0034:EN:PDF
Kellens, W.; P. Deckers; H. Saleh; W. Vanneuville; P. De Maeyer; G. Allaert & R. De Sutter (2008). “A GIS tool for flood risk analysis in Flanders (Belgium).” Risk Analysis Vi: Simulation and Hazard Mitigation: 21-27.
Kok, M.; Huizinga, H.J.; Meijerink, T.C.; Vrouwenvelder, A.C.W.M.; Vrisou van Eck, N. (2002) Standaardmethode 2002: schade en slachtoffers als gevolg van overstromingen, Minsterie van Verkeer en Waterstaat. Directoraat-Generaal Rijkswaterstaat. Dienst Weg- en Waterbouwkunde [S.l.]. 34 + appendices pp.
Lumbroso, D. (2007) Review report of operational flood management methods and models, FloodSite Consortium, T17-07-01, http://www.floodsite.net/html/partner_area/project_docs/Task17_report_M17_1review_v1_1.pdf
Penning-Rowsell, E.; Johnson, C.; Tunstall, S.; Tapsell, S.; Morris, J.; Chatterton, J.; Green, C.; Wilson, T.; Koussela, K.; Fernandez-Bilbao, A. (2005) The benefits of flood and coastal risk management: a manual of assessment techniques, Middlesex University Press: London, UK. ISBN 1-904750-52-4. XII, 238 pp.
Samuels, P. & Gouldby, B. (2009) Language of Risk – Project Definitions (Second Edition), FloodSite Consortium, T32-04-01, http://www.floodsite.net/html/partner_area/project_docs/T32_04_01_FLOODsite_Language_of_Risk_D32_2_v5_2_P1.pdf
Schwindt, M. & Thieken, A. (s.d.) Scientific coordination project – Review on Resilience, 2nd ERA-NET CRUE Research Funding Initiative “Flood resilient communities – managing the consequences of flooding” [under preparation]UN-ISDR (s.d.) International Strategy for Disaster Reduction – Terminology of disaster risk reduction, http://www.unisdr.org/eng/library/lib-terminology-eng%20home.htm (latest consult on 7 MAR 2011).
Vanderkimpen, P.; Peeters, P. & Deckers, P. (2010) The impact of Individual buildings on urban flood risk analysis, in: (2010). SIMHYDRO 2010: modèles hydrauliques et incertitudes, Nice, 2-4 juin 2010 [CD-ROM]. pp. 1-9.
Van der Sande, C.J. (2001) River flood damage assessment using IKONOS imagery, European Commission, Joint Research Centre, Natural Hazards Unit: Ispra, Italy. viii, 77 pp.
Verwaest, T.; Vanpoucke, Ph.; Reyns, J.; Van der Biest, K.; Vanderkimpen, P.; Peeters, P.; Kellens, W.; Vanneuville, W. (2008) SAFECoast: Comparison between different flood risk methodologies. Action 3B report – SAFECOAST Interreg IIIb North Sea Project, Flanders Hydraulics Research: Borgerhout, Belgium. 128 pp.
Willems, P.; Timbe, L.; Thompson, S.; Campling, P.; Barbieri, M.; Torriano, L. & Vanneuville, W. (2003), ‘FAME: Flood risk and damage assessment using modelling and earth observation techniques’, Integrated Modelling User Group (IMUG) International Conference, Tilburg, 23-25 APR 2003.
Wouter Vanneuville (master’s degree in geography) drew up a flood risk methodology as a researcher of Ghent University. Since 2006, he has worked for the Department of Mobility and Public Works (division Flanders Hydraulics Research) of the Flemish Authorities. He’s mainly involved in the implementation of the EU Floods Directive, in the translation of research into policy and management for water quantity (including floods) and the impact of climate change. He’s a member of several working groups of the Coordination commission on Integrated Water policy, CIW, the competent authority for the Water Framework Directive and Floods Directive in Flanders. He’s also the Ambassador of CRUE, a European network of flood research funding institutes.
Wim Kellens is research assistant at the Department of Geography of Ghent University. He holds a Master’s degree in Geography with specialization in cartography and GIS. His PhD research is funded by the Flemish Fund for Scientific Research (FWO) and focuses on coastal flood risk management, with specific attention to the effects of seasonal tourism on flood risk calculations and the analysis of public risk perceptions. By mapping the concerns, preferences and needs of the coastal population, the research aims at unveiling the necessary information to develop better and more efficient flood risk communication to the public.
Philippe De Maeyer (°Gent, 1955) is professor in cartography and GIS at Ghent University and head of the Department of Geography. He is the chair of the Scientific Board of the Belgian National Mapping Agency (NGI) and chair of the Commission WT8 Earth and Space Sciences of Research Foundation – Flanders (FWO). He is also full member of the GDI-council (Flemish governmental advising council on geographical data infrastructure) and chair of the Belgian sub-commission on cartography (Belgian National Committee of Geography). He is also chair of the ICA Commission on Map Production and Economics. Since 2011, he has been chair of the Belgian National Committee of Geography.
Genserik Reniers lectures in general chemistry, organic chemistry, and chemical process technology at the University of Antwerp, Belgium. He is also visiting professor of Risk Management at the Institute of Transport and Maritime Management in Antwerp. At the Hogeschool-Universiteit Brussel in Brussels, Professor Reniers lectures, amongst others, on prevention management and advanced occupational health and safety management. He has extensive experience in leading research projects funded by the Belgian government and the chemical industry. He is a Fellow of the International Congress of Disaster Management, Member of the Society for Risk Analysis and the Royal Flemish Society for Engineers, and serves as an Associate Editor for the Journal of Loss Prevention in the Process Industries.
Frank Witlox holds a PhD in Urban Planning (Eindhoven University of Technology), a Master’s Degree in Applied Economics and a Master’s Degree in Maritime Sciences (both University of Antwerp). Currently, he is Professor of Economic Geography at the Department of Geography of the Ghent University. He also is a visiting professor at ITMMA (Institute of Transport and Maritime Management Antwerp), and an Associate Director of GaWC (Globalization and World Cities, Loughborough University). Since 2010, he has been Director of the Doctoral School of Natural Sciences (UGent).