Karin. M. De Bruijn1 & 2 & Frans Klijn2
(1Delft University of Technology, 2 WL|Delft Hydraulics)
WL|Delft Hydraulics, P.O. Box 177, 2600 MH Delft, The Netherlands,
Tel: (31)-15-2858585, Fax: (31)-15-2858582 E-mail:karin.debruijn@wldelft.nl
Abstract: The traditional flood risk management strategy in the Netherlands is based on preventing floods by constructing dikes and other structures. After each flood dikes were raised, people felt safer and investments in the area increased, causing a further need to prevent flooding. Nowadays, safety in the Netherlands is legally based on the exceedence probability of the discharge: all dikes should be able to withstand water levels related to a discharge with a return time of 1250 years. However, flood risk management not only depends on the hazard but also on the vulnerability of the area. The society does not only determine this vulnerability but also the strategy followed to cope with flood risks. The social and economical values indicate which level of risk is accepted, how much can be spent to prevent floods and what other values, such as nature and cultural landscape values should be taken into account. Changed societal views in the Netherlands lead to a discussion on alternative flood management strategies. A resilient flood risk management strategy might be a more appropriate strategy nowadays. This paper defines resilience in the context of flood risk management. A resilient strategy focuses on reducing the impacts of floods by increasing the speed of recovery of the system. The system is defined as the society and ecosystems in the area threatened by river floods.
Keywords: flood risk management, resilience, floods, uncertainties
The Netherlands has a long history of coping with floods. Traditionally, flood risk management focused on preventing floods by river training and embankments. This flood risk management strategy has some disadvantages that justify investigating the possibilities of an alternative strategy. Instead of focusing on preventing floods (the traditional resistant strategy), it might be safer to try to live with floods of various magnitude. This new strategy is called a resilient flood risk management strategy. This paper first describes the history of flood risk management in the Netherlands and the disadvantages of this strategy. Then the resilience concept is explained and defined in the context of flood risk management. Finally, the proposed resilient flood risk management strategy is discussed in relation to the current strategy.
One third of the Netherlands needs and has artificial protection against floods from the sea or the major rivers (see Fig. 1). In this area millions of people live and large industries have settled, therefore flood risk management is an important issue for the Netherlands.
Around 1000 AD people started to build the first dikes around relatively small polders. Already in 1400 AD an almost completely closed dike system existed along the rivers (Commissie Rivierdijken, 1977). Dike breaches occurred regularly, leading to dike improvements after each case. Furthermore, huge changes in the course of the river and the river bed in the form of regulation and canalization took place (Janssen and Jorissen, 1997). The river was confined and controlled more and more. The last great flood of the Rhine River occurred in 1926, after which the dikes were improved again (Commissie Rivierdijken, 1977). In 1953 a major flood from the North See took place, killing more than 1800 people (Janssen and Jorissen, 1997). This flood incurred new regulations for safety not only for the coast but also in the rivers. Previously, dike heights were based on the maximal recorded water level, but after 1953 a more scientific base was used. The optimal level of safety was defined as the accepted probability of flooding for the different areas in the Netherlands. To be able to use this new norm, it was simplified to the demand that dike levels should exceed water levels related to a discharge with a chosen return time (Commissie Rivierdijken, 1977; WL, 1993). After several years of discussion one safety level was chosen for the whole area threatened by river floods: the discharge with a probability of once in 1250 years (the design discharge). Flooding by other causes than overtopping of dikes and uncertainties in nature and in the calculations were considered by adding 0.5 m to the required height and some regulations for the design of the dikes. Every five years the design discharge is calculated anew based on the recorded discharges.
In 1993 and 1995 extremely high discharges occurred. After these events planned dike improvements were carried out much faster than foreseen. Because of these extreme peak discharges, the design discharge with a probability of once in 1250 years will rise. Using the traditional approach of flood risk management this would lead to a further increase of dike heights. However, new solutions are developed nowadays creating room for the rivers (Min.VROM en V&W)
Due to changes in the values of society and available technology a new flood risk management might be needed. Some disadvantages of the current flood risk management strategy are mentioned below. This paper does not pretend to give a complete overview, but will discuss some disadvantages to justify the research on an alternative strategy.
(1) The current strategy is based on one design discharge for the whole area threatened by inundation from the main rivers in the Netherlands. This approach leads to the same level of protection of different kind of areas: cities, agricultural areas and nature reserves all have the same probability of flooding. The question arises if this is economically sensible. More importantly, it means that it is unknown which area will flood whenever the design discharge is exceeded. Because all areas theoretically have the same probability of flooding, a large area has to be evacuated (loss of lifes is unacceptable in the Netherlands).
(2) Due to the fact that this strategy focuses on preventing floods, little attention is paid to the consequences of possible floods. The potential flood damage has increased enormously so probably the once-in-1250-year-discharge is not optimal anymore for the current situation. Furthermore, relatively little attention is paid to plans for emergency situations, evacuation plans, flood mitigation measures, etc. which also might lower flood risks. New technology has increased the possibilities to make good flood forecasts, of communication technology and flood warnings. Also the possibility to anticipate on peak flows by lowering a structure and use a detention pond has grown. These technologies make it easier to allow controlled floods and therefore reduce the need to prevent floods in all circumstances.
(3) Also the coping with flood risk is a weak point of this strategy. By only looking at probabilities of a design discharge, flood risks are not clearly visualized. The uncertainties in for example the design discharge, the translation of the discharge to water levels, the diversion of the peak flow through the different river branches, the strength of dikes and structures, human behaviour etc. are not clearly considered. For inhabitants it is unclear which risk they face and before 1995 it may even have been unclear that there was a flood risk at all. This causes a false sense of safety in the area and thus a quick economic development causing the potential flood damage to rise. It also gives reason to the people to blame the Water Board for possible floods.
(4) The current flood risk strategy causes an endless need for raising and improving the water defense structures. Restricting the natural dynamics of a river system by canalization and embankments will require a continuous attention, otherwise the river dynamics will damage these works and the river will try to return to its own natural behaviour. Because of the restriction of the river to very narrow floodplains, large amounts of sediment have been deposited resulting in a significant increase of their level (Janssen and Jorissen, 1997). The increasing level of the floodplains implies less room for the river and thus higher water levels and an increased flood risk.
(5) Besides, the land surface in the Netherlands is subsiding by the improved drainage of the land and soil composition of the area. When a dike breaches or flows over, water will enter relatively low-lying areas where many people live and work. The protected areas have only two flooding states: not-flooded or catastrophically flooded (see Fig. 2).
(6) The current strategy results in a destruction of the river scenery, damaes the nature values and cultural values. Recently, these values have become more important. A new strategy should consider these values and not only focus on reaching a “safe’ situation by technical solutions.
These examples show that there is reason enough to study other flood risk management strategies.
The concept of resilience is used in many disciplines. In the Netherlands’ water management the concept of resilience is taken from the discipline ecology (Klijn and Marchand, 2000). In ecology Holling (1973) introduced the concept. Other ecologists picked up the concept also (Pimm, 1984; May, 1975; Jøgenson, 1992; Clapham, 1973, Knaapen et al., 1999; Klijn, 1999). They discuss the relation of resilience with resistance and persistence. Resilience and resistance both are ways to reach a persistent (or sustainable) system. A resilient system reacts on a disturbance and then recovers; a resistant system does not show a reaction but just absorbs the disturbance and persists. A tree species could for example survive fires by having a fire-resisting bark (resistant strategy), or alternatively it could burn down and regenerate from seeds with a fire-induced germination (resilience strategy). In very dynamic environments, like coasts and natural floodplains, resilient species dominate, while in stable environments, like rainforests, coral reefs etc. more resistant species will be found. These resistant ecosystems are generally very vulnerable for bigger disturbances. In general “resilience” can be defined as the ability of a system to persist if exposed to a disturbance by recovering after a response. In contradiction, resistance, can be defined as the ability of a system to persist if exposed to a disturbance without showing any response at all (see fig. 3).
Several attempts already have been made to apply this general definition of resilience in water management. Klijn and Marchand (2000), Termes et al. (1999) and Klein et al. (1998) define resilience in the context of coastal management and water systems in general. Also Remmelzwaal and Vroon (2000) used resilience for water systems and Hashimoto et al. (1982) wrote about resilience in relation to the design and operation of reservoirs. However resilience has not been applied to flood risk management of low land rivers.
To define the resilience of a system the disturbance, the response to the disturbance and the recovery should be described (Klijn and Marchand, 2000). Therefore, to define resilience in the context of flood risk management we need to answer the following questions (Klijn and Marchand, 2000):
(1) What system do we consider? Where do we put the boundaries of that system (scale)?
(2) For what disturbances do we want the system to be resilient?
(3) What kind of response do we expect?
(4) How can recovery be defined?
These questions are discussed below.
To limit the research area to a tangible size, system boundaries have to be defined. Because in reality the whole world forms a system it is difficult or even impossible to define boundaries in such a way that no objects fall outside the system that influence its behaviour. The boundaries must be chosen in such a way that the most important aspects for flood risk management are included. Flood risk management consists in the analysis of two elements: the hazard threatening the system and the vulnerability of the system. To assess the vulnerability of the system and the resilience of such a system, the society and ecosystems in that system should be taken into account. Besides, the behaviour of most lowland river systems is heavily influenced by man. To understand and study flood risk management and the reaction to floods, the society should be included in the focal system. The problem to flood risk management of lowland rivers can geographically be restricted to the area that could be flooded if no dikes existed. Society in the protected parts of the floodplains (see Fig. 2) is of course influenced by society outside that area, the national economy and the government. These factors also influence the resilience of the society in the area. In countries with a healthy economy potential damage probably will be higher, but damage to e.g. roads can be repaired faster than in a developing country. The government of a country could do a lot to reduce the impacts of floods or improve the recovery after a flood, e.g. pay damage compensation, construct new houses, invest in a warning system, etc. Also friends, family and other people living outside the river area will help the victims of floods. Therefore, the system limited by these geographical boundaries should not be seen as the research area, but as the focal system. The focal system can thus be defined as: the physical system and society in the whole area that might be flooded if there were no dikes.
The disturbances in the context of flood risk management are peak discharges flowing into the river system. These disturbances might increase in intensity or frequency due to climate change, hereby increasing the need for resilience.
The response to peak discharges consists in the disruption of the normal situation of the focal system. Floods can heavily disturb the social system and the ecosystems in the area. This flood and the disruption of the normal situation to a worse situation, is the reaction seen in the system defined as above. If response is defined as the disruption of the normal situation, recovery can be seen as the return to the normal situation. What exactly is this normal situation, is difficult to define. At least the economical, social and ecological development should continue as before the flooding or in a similar way as in the non-flooded parts. Damage is repaired, companies reach their normal production rate again, etc. As Holling (1973) already pointed out, recovery does not require a return to exactly the same situation as before. Probably people will change their furniture after a flood, companies might use damage compensation not only to repair the damage but to also modernize their company, etc.
Resilience in the concept of flood risk management can thus be defined as:
the
ability of system to return to a normal situation after flooding of a part of
the area caused by a peak discharge.
The current flood risk management of the Rhine River can be regarded a resistant strategy, The differences between a resilience and the current (resistance) strategy for the river Rhine River are discussed below.
(1) Dealing with uncertainty:
Both strategies have to deal with uncertainty, but in the resilience strategy this is more explicitly done. The resistant strategy is based on the chosen design discharge and risks are not quantified explicitly. The resilience strategy is not based on a single norm and uncertainties can be included explicitly in the strategy.
(2) Range of discharges:
In the resilience strategy the whole discharge regime is included in the analysis in stead of one design discharge. Once a flood occurs higher than this resistant system can bare, a sudden collapse will occur. In a resilience system damage would raise gradually and no sudden collapse would be expected (see Fig. 4).
(3) Focus: Reducing the impacts of floods instead of flood prevention:
A resilience strategy does not focus on preventing floods, but aims at accelerating the return to a normal situation after a flood. It focuses on recovery. This means that floods might be allowed but on a controlled, less harmful manner. It also brings along that not only probability of events and hydrologic and hydraulic processes within in the river must be studied, but also the consequences of events.
(4) One safety level for the whole area, or a differentiation of the area:
The current strategy has the same probability of flooding in all areas (theoretically at least). The new strategy might use the spatial planning and the land use to differentiate. The area as a whole will be more resilient if the less valuable parts flood earlier than the more valuable parts. By controlled floods elsewhere these valuable parts will be safer and the recovery after a certain event will be easier.
(5) Location and type of measures:
The current strategy focuses on water defenses,
near the river itself. The resilient flood risk strategy will mainly take
measures in the area threatened by floods. The resilience strategy will not only
use water defenses but also measures to reduce the impacts of floods such as:
warning systems, evacuation plans, spatial planning regulations, compartment
(dikes to divide the area in smaller pieces so the flooded area will be smaller)
etc. and measures to accelerate the recovery after a flood, such as damage
compensation regulations, insurance etc.
This paper shows that the current flood risk management strategy has sufficient disadvantages to look at other strategies for the Netherlands. In this paper the resilience strategy is proposed as the contrary to the current resistance strategy. The resilience strategy focuses not on preventing floods but on avoiding damage and on rapid recovery after a flood. The paper focuses on the Netherlands. However, the resistance strategies are nog only common in the Netherlands. Recent floods in China, Bangladesh, Cambodia, Poland, Germany, the United States, Chili and Mozambique show that floods still cause disasters, parly because of these strategy. A resilient flood risk management strategy might be a useful alternative in some situations. A resilience strategy, or “living with floods” can prevent disasters, caused by failure of dikes and may lead to a more sustainable system. Fighting against natural dynamics, could be senseless in many ways. A resilience strategy might be the way to work with instead of against the natural dynamics, and to deal with uncertainties. More research has to be done to define and operationalize the resilience concept. This paper illustrates that the possibilities of a resilience strategy are worth investigating.
References
Commissie Rivierdijken, 1977. Rapport commissie rivierdijken. (In Dutch) Hoofddirectie van de Waterstaat,’ s-Gravenhage.
Clapham, W.B, 1973. Natural ecosystems. Macmillan, New York.
Hashimoto, R., Stedinger, J.R., Loucks, D.P., 1982. Reliability, Resiliency, and Vulnerability Criteria for water resource system performance evaluation. Water resources research. 18 (1) p. 14-20.
Holling, C.S., 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4: 1-24.
Janssen, J.P.F.M., & Jorissen, R.E., 1996. Flood management in the Netherlands; Recent development and research needs. In: R. Casale, K. Havnø, P. Samuels (eds.) Ribamod, River basin modelling, management and flood mitigation Concerted action, p. 89-104, 1996.
Jøgenson, S.E., 1992. Integration of Ecosystem Theories: a pattern. Kluwer Academic Publishers, Dordrecht/Boston/London.
Klein, R.J.T., C.H. Hulsbergen, H. Goosen and M.J. Smit (1998). Resilience and Vulnerability: Coastal Dynamics or Dutch Dykes? The Geographical 164 (3).
Klijn, F, 1999. Duurzaamheid , biodiversiteit en veerkracht: verkenning rond toxische stoffen in rivier-ecosystemen. WL|Delft Hydraulics, R3408. Delft. (In Dutch)
Klijn, F. & Marchand, M., 2000. Veerkracht een nieuw doel voor het waterbeheer? Landschap 17(1): 31-44 (In Dutch).
Knaapen, J.P., J. Klijn & M. van Eupen, 1999. Veerkracht van zoete en brakke wateren: een benadering vanuit ecologie en ruimte. RIZA werkdocument 99.137X. Lelystad: RIZA / Alterra rapport 688. Wageningen: Alterra (In Dutch).
May, R.M. 1974. Stability and complexity in model ecosystems. Second edition. Princeton University Press, Princeton, New Jersey, USA.
Min.VROM en V&W (Ministeries VROM en V&W), 1997. Beleidslijn Ruimte voor de Rivier. Den Haag (In Dutch) .
Pimm, S.L. 1984. “The Complexity and Stability of Ecosystems.” Nature 307: 321-326.
Remmelzwaal, A. Vroon, J. 2000. Werken met water: Veerkracht als strategie. (In Dutch). RIZA Rapport 2000.021 Lelystad (In Dutch)
Termes A.P.P., Kok M., Klijn, F. , 1999. Flexibiliteit en veerkracht van inrichtingsstrategieën Rijntakken. Effecten van Klimaatveranderingen. HKV rapport PR, 231, Lelystad (In Dutch).