Flood resilience

Abstract

Three different conceptual frameworks of resilience, including engineering, ecological and social–ecological have been presented and framed within the context of flood risk management. Engineering resilience has demonstrated its value in the design and operation of technological systems in general and in flood resilient technologies in particular. Although limited to the technical domain, it has broadened the objectives of flood resilient technologies and provided guidance in improving their effectiveness. Socio-ecological resilience is conceived as a broader system characteristic that involves the interaction between human and natural systems. It acknowledges that these systems change over time and that these interactions are of complex nature and associated with uncertainties. Building (socio-ecological) resilience in flood risk management strategies calls for an adaptive approach with short-term measures and a set of monitoring criteria for keeping track of developments that might require adaptation in the long-term (adaptation pathways) and thus built-in adaptive capacity as opposed to building engineering resilience which involves a static approach with a fixed time horizon a set of robust measures designed for specific future conditions or scenarios. The two case studies, from a developing and a developed country, indicate that the concepts of ecological and socio-ecological resilience provide guidance for building more resilient flood risk management systems resulting in an approach that embraces flood protection, prevention and preparedness. The case studies also reveal that the translation of resilience concepts into practice remains a challenge. One plausible explanation for this is our inability to arrive at a quantification of socio-ecological resilience taking into account the various attributes of the concept.

This article is part of the theme issue ‘Urban flood resilience’.

1. Introduction

The emergence of resilience in multiple disciplines presents a challenge and opportunity in flood risk management. Resilience is widely used in flood risk management policies, but is still largely conceptual [1]. The concept of resilience, despite its multi-interpretations, has had a profound influence on the evolution of flood risk management approaches over the past two decades. The transition from structural protection towards a risk-based approach followed by a more holistic whole systems approach has been driven by the emergence of this concept as of the beginning of this century. However, the number of empirical and quantitative case studies to demonstrate the practical relevance in flood risk management is still limited. The concept of resilience, as opposed to resistance (often defined as the ability of a system to remain unchanged when being subjected to disturbances, see also §2), represents a new way of thinking about flood risk management expanding its objective to go beyond the ability to ‘resist’ when exposed to high water levels which have been foreseen in the design, towards the ability to ‘recover’ from a flood event (and/or to reduce the impacts that arise when flows occur that exceed the design standard) and to ‘adapt’ or to ‘transform’ the existing approach based on the recognition that the conditions have been or will change in the future. In this context, the terms reactive and proactive resilience are also used [2]. The latter refers to activities that occur before the disturbance and therefore is close to the ability to adapt and transform, whereas reactive resilience is more aligned with the ability to resist and recover.

To this end, the concept of resilience in flood risk management has contributed to the notion that societies should learn to live with floods and should mitigate disastrous consequences and not seek to avoid them entirely. It follows from the above that a resilient flood risk management strategy should embrace the deployment of measures reducing flood risk through a combination of protection, prevention and preparedness spanning a wide range of flood probabilities (from regular to rare flood events) to reduce the flood hazard and associated consequences. Indeed, there is a growing awareness that resilience should be mainstreamed into the planning and design of flood risk management strategies. However, as developing countries are affected by frequent small events (as opposed to developed countries where the few big events are causing the largest impacts), there is a tendency to prioritize investments in flood protection in the developing countries [3].

In this chapter, different frameworks of resilience will be briefly compared and their application in influence on flood risk management will be discussed. Two case studies are presented to illustrate their practical relevance in the developing and developed world.

2. Frameworks of resilience

Resilience is applied in at least three different ways. This chapter explains and compares the basic features of these frameworks within the context of flood risk management. It should be noted here that the concept of resilience has been adopted by and adapted to a range of disciplines such as psychology, ecology, engineering and sociology, over the past 30 years. Consequently, there are many, sometimes opposing definitions and interpretations. But often it is merely a matter of semantics.

(a). Engineering resilience

In the first, more restricted definition and applied in engineering, resilience is conceptualized as an outcome [1]. Engineering resilience is increasingly being applied in planning, architecture and building technology focusing on flood hazard mitigation and involving the deployment of flood resilient design and technologies to adapt or construct buildings to reduce the probability of failure, reduce the consequence during failure and/or to reduce recovery time after failure by flood water (e.g. [4,5]). A comprehensive example of the dimensions of engineering resilience relevant to architecture, but also useful in flood risk management, is given by Laboy & Fannon [6] (figure 1). The authors expanded the existing 4R resilience model (robustness, redundancy, resourcefulness and rapidity) of Bruneau et al. [7] beyond the design and operation phase to a 6R model that includes ‘risk avoidance’ and ‘recovery’. Risk avoidance can be achieved through proper site selection in the planning phase. With this modification, the authors added qualities to engineering resilience which are strictly not confined to the technical (engineering) domain only, but also lie in the social domain. The dimensions of the 6R model are typically being used in the domain of disaster reduction aiming at prevention (including planning), preparedness and recovering from shocks to preserving the status quo [8].

Figure 1.

Dimensions of engineering resilience based on the 6R model of Laboy & Fannon [6].

(b). Ecological resilience

The concept of engineering resilience assumes that the system remains constant over time. Although this may hold true for situations of moderate shocks and predicable events, disruptive (unexpected) extreme events may have a profound impact on the system's functioning triggering a profound system change (transformation). There are numerous examples of disruptive flood events which have resulted in a shift of the flood risk management strategy from deployment of small, local scale flood mitigation measures to large-scale flood defence systems. Also the principle of ‘building back better’, which is defined as ‘The use of the recovery, rehabilitation and reconstruction phases after a disaster to increase the resilience of nations and communities …’ [9] aligns with the characteristics of ecological resilience. A similar behaviour of system change is observed in natural systems. Building on the paradigm of multi-equilibria in ecology [10], in the second definition, resilience has evolved into a broader concept of ecological resilience which is typically defined from a holistic system's perspective. Holling [10] was one of the first to distinguish two contrasting aspects of ‘stability’, namely the ability of a natural system to return to an equilibrium state (return to its original state) after disturbance from (ecological) resilience, as the ability to absorb change and disturbances and still maintain the same relationships (‘persistence’ or maintaining existence of function). Where engineering resilience is associated with the functional stability of technological systems (maintaining efficiency of function), ecological resilience is being used as an approach for understanding the dynamics of complex and dynamic natural systems. In this emerging concept, (ecological) resilience is observed as a process of which the post-disruption state can be different than the pre-disruption state, but the whole recovery process is resilient [11–13].

In the context of flood risk management, ecological resilience refers to the ability of a system to resist or absorb disturbances, such as storm surges and cloudbursts, and to remain functioning under a wide range of flood wave or rainfall intensities. In this definition, continued functioning implies either withstanding the flood wave (resistance) and/or quick recovery with limited impact after being exposed to flood water (e.g. due to failure of the flood defence system; e.g. [14,15]) with the ultimate aim to avoid impacts from which recovery is extremely difficult (e.g. [16]). Here resilience depends on both the technical and the social properties of the flood risk management system, such as risk avoidance, robustness or the capacity to withstand a disturbance without functional degradation, redundancy or the extent to which system components are substitutable, resourcefulness as the organizational capacity to monitor, detect and respond, and rapidity or the capacity to recover or restore the system in a timely manner [7,17].

(c). Socio-ecological or adaptive resilience

The realization that man-made and ecological systems are exposed to both slow changing drivers and shocks and thus that the stability domain itself is changing has called for the consideration of the temporal dimension of resilience (e.g. [18]). The acknowledgement of this temporal dimension has resulted in the emergence of the framework of socio-ecological or adaptive resilience, which recognizes nonlinear dynamics, thresholds, how periods of gradual change interplay with periods of rapid change and also how to address uncertainty in projections of slow changing drivers such as climate change, population growth and resource depletion (e.g. [11,15]). The concept of socio-ecological resilience has been defined as ‘the capacity of linked social–ecological systems to absorb recurrent disturbances such as floods so as to retain essential structures, processes and feedbacks' [11]. In addition, socio-ecological resilience also reflects the degree to which complex adaptive systems are capable of self-organization and to which these systems can build capacity for learning and adaptation (e.g. [11,19]). It goes without saying that ‘adaptive capacity’ is confined to the social domain with the actors that will anticipate and organize adaptive responses. This broader concept of resilience has been adopted in the domain of climate change adaptation as a way to deal with both gradual, disturbing changes and shocks resulting from climate change and variability [12,13,20]. In this context, an adaptive, resilient strategy is not based on static conditions, but explicitly takes unpredictable changes (trends)—like climate change—into consideration. The term adaptation used in the climate change discourse most often refers to responses that aim to prepare for the consequences of a changing climate. A typical example would be the heightening of a sea dyke as a response to increasing sea-level rise. The term adaptive refers to an adjustment of a plan or strategy following new insights [21].

(d). Comparison of the three frameworks

Three resilience frameworks have been presented. Engineering resilience is used to express the functional stability of technological systems, whereas ecological resilience is associated with complex and dynamic living systems acknowledging multiple equilibrium states enabling conditions for persistence [6]. Finally, socio-ecological or adaptive resilience aims to overcome the limitations of both approaches in the sense that these approaches are static and do not address the adaptive capacity of systems needed to adjust to slowly changing conditions (transformation). Figure 2 depicts an example of the ball and cup model used in resilience theory to illustrate the different features of the three frameworks described above [6].

Figure 2.

Ball and cup model of system stability in the three resilience frameworks (Laboy & Fannon [6]). The ball represents the system and arrows represent disturbances. The landscape on the right represents one stability domain (one valley). Engineering resilience is defined by the slope of the sides of an individual valley. There is only one equilibrium (status quo). The landscape in the middle consists of two valleys. A loss of the intervening hill in the middle represents a loss of ecological resilience. At this point, the system can potentially re-organize around another stability domain. Movement of a ball in the horizontal direction is a measure of the change in ecological resilience. The landscape on the right represents the adaptive nature of socio-ecological resilience as the stability domain itself is shifting. (Online version in colour.)

A summary and comparison of the main features of the frameworks discussed above are given in table 1.

Table 1.

Frameworks of resilience and features of resilience used in flood risk management (based on [11,15]).

system attribute	characteristics	stress	aim/strategy
resistance	ability to withstand disturbance without responding	shock	stability (preserve status quo)flood protection
engineering resilience	ability to bounce back and recover from disturbance recovermaintaining efficiency of function	shock	constancy (efficiency of function, preserve status quo)robustnessfail-safe designappropriate for engineering components and systems
ecological resilience	capacity to absorb disturbance, recovermaintenance existence of function	shock	persistence, redundancymultiple equilibrium states
socio-ecological or adaptive resilience	capacity to absorb disturbance, recover, re-organize, and anticipate and adapt while undergoing change	gradual/shock	persistence (existence of function)learning, adaptive capacity, transformation

3. Shifting from a traditional to a resilient approach

For many decades, the traditional paradigm has favoured large-scale infrastructure systems, such as flood embankments and channelization. These systems have not been designed for failure, and, as a consequence, impacts of extreme flood events may be catastrophic. In addition, poor design and maintenance of these infrastructure systems often further aggravate these impacts. Contemporary thinking about the behaviour of these systems has reinforced the need to change the paradigm of managing those systems (table 2). Indeed, in many parts of the world, major flood disasters have acted as catalysts for changing the problem perception underlying the flood risk management approaches: the assumption that systems are predictable and thus controllable has been replaced by the acknowledgement of uncertainty. Hence, flood risk management systems are increasingly being conceived as inherently imperfect systems. They need to become (more) resilient requiring to build in both robustness (such as in the budgeting and infrastructure design) and flexibility (such as in the implementation of measures and infrastructure design) in both the process (governance) and design (hardware) of these systems [21]. These systems require an integrative strategy which brings together human, ecological and technical components.

Table 2.

Features of the traditional flood risk-based approach and the flood resilient approach.

	traditional approach	flood resilient approach
problem perception	changes in system are predictable	changes in system are uncertain
key objective	control changes, stability(problem-solving)	persistence, enhance capacity to adapt to uncertainties (anticipation)
governance perspective	sequential process of planning(focus on flood probability reduction (protection))systems of static norms and standards	continuous alignment of content and process with context(balance between protection, prevention and preparedness)system of strategic alternatives (e.g. adaptation pathways)

The main features of a traditional approach and a flood resilient approach are summarized in table 2 (e.g. [15,18,22]).

It follows from the above that the concept of resilience emerged and challenged scientists, policy makers and practitioners in flood risk management to move away from or expand the traditional paradigm of flood prediction and control, leading to primarily the implementation of structural engineering solutions, towards one of system and whole life thinking fostering an integrative and adaptive, long-term approach, referred to as a flood resilient approach [23–25]. The broader concept of socio-ecological resilience has provided guidance for building more resilient flood risk management systems as they are founded on the following features (e.g. [26–29]):

• (i)accepting that knowledge will never be perfect and that changes are uncertain and hence that there is no ‘optimal’ or ‘best’ solution;
• (ii)taking a long-term view, while nurturing the capacity to monitor and learn from intermediate outcomes and to adapt (short-term incremental changes) and keep options open to transform (long-term system changes);
• (iii)taking into account all of the potential interventions that may alter flood risks (ranging from flood preparedness to prevention); and
• (iv)facilitating participation and collective action and learning.

These resilient approaches aim to establish a balance between flood protection, prevention and preparedness, both now and in the future (e.g. [15,18,22]). Whether a balance between types of interventions is preferred and which interventions are prioritized varies significantly between countries and regions depending on a variety of aspects that can be highly time-dependent [30]. An important aspect of the shift from a traditional to a resilient approach is the relative costs (and benefits) of the two approaches. Vrijling [31] demonstrated that combining flood protection with flood damage reduction is typically not cost-effective for a flood risk management system resembling The Netherlands (with high protection levels). Klijn et al. [32] found similar results, but showed that flood damage reduction by applying unbreachable embankments has equal total societal costs compared with flood protection. Kolen & Kok [33] demonstrated for The Netherlands that it can be cost-effective to invest a small amount in emergency management (next to flood protection). Next to the relative costs (and benefits) a number of other aspects determine the preferences regarding the flood risk management approach. These aspects include [30] the degree of public awareness of flood risk (including social attitudes toward flood risk management), the occurrence and severity of flood events in the recent past, the (potential) flood consequences in terms of human life, economic assets and natural environment, and the degree of flexibility in policy-making to make financial resources available for flood risk management.

4. Quantifying flood resilience

Measuring flood resilience is not an easy task. Partly because there is no shared definition (yet) and, hence, the selection of resilient indicators is a highly subject matter. Partly also because in the flood domain, resilience is always about the interaction between people (e.g. past experience, income level, health status) and the physical environment (e.g. the flood protection level, material selection of flood barriers, buildings). Some variables that represent these indicators are (potentially) available such as flood characteristics. Others are difficult to assess and associated with large uncertainties, such as flood impacts. Most of the frameworks to measure flood resilience focus on the relationship between probability and (direct) impacts of flooding. Flood (damage) models are being used to assess probabilities and consequences of flooding and the effectiveness of management interventions [34].

Attempts to quantify flood resilience are often based on indicators which relate system response to the magnitude of flood waves or rainfall intensity (figure 1; e.g. [14,35,36]). For example, [14] provided an analysis of what makes river basins flood resilient and how flood resilience can be enhanced. In this analysis, flood resilience is quantified using indicators that reflect the different aspects of the reaction, comprising: the reaction threshold, amplitude, graduality and recovery rate. The reaction threshold involves the recurrence time of the maximum load the system can withstand such as the maximum river discharge or rainfall intensity which is not expected to cause floods. The amplitude of the reaction indicates the severity of the expected (direct) damage resulting from a certain peak discharge or extreme rainfall event. The graduality reflects the extent to which the damage increases with increasing disturbances caused by flood waves. The recovery rate describes how fast a system will recover from the reaction to a disturbance. Although these resilient indicators reveal relevant information on the system's performance, they cannot be aggregated and expressed in one numerical value, as it requires to assign a weight to the indicators, which constrains its wider application [37].

Klijn et al. [32] used the term robustness as a proxy for ecological resilience to provide an additional metric in flood risk analyses. To assess robustness a ‘response curve’ was used which depicts the flood impacts as a function of a range of disturbances, such as river peak discharges similar to figure 3. This response curve resembles a risk curve, in which the probabilities are representing the river peak discharges and the impact the system's response. Social–ecological resilience can be analysed in a response curve as the sum of the resistance range and the resilience range [38]. The resistance range is quantified by those discharges that cause no impact to the system. It ends where the impacts become greater than zero. The resilience range is quantified by those discharges that cause limited impact from which the system is able to recover. This range ends where the impacts exceed the recovery threshold, which is the maximum impact from which the system can still recover.

Figure 3.

Theoretic response curve, showing system response as a function of disturbance magnitude (e.g. magnitude of flood wave), indicating resistance and resilience (adapted from [16]).

Other scholars have assessed factors that attribute to socio-ecological resilience such as economic resources, assets and skills, information and knowledge, support and supportive networks, and access to services. These factors are being used to select resilience surrogates as they relate to a particular component or notion of flood resilience.

5. Case studies

In this chapter, two case studies are presented to illustrate the development and assessment of the flood resilient approach for flood risk management. One case study is in Dordrecht, The Netherlands, and the other is in Can Tho, Vietnam.

(a). Resilient flood risk strategy for the Island of Dordrecht, The Netherlands

The Island of Dordrecht is located in the transitional area of the Rhine--Meuse delta (The Netherlands). In this area, the water stages are influenced by the river runoff as well as the sea level. The Island covers about 9000 ha and comprises one municipality (the city of Dordrecht) and one dyke ring area (dyke ring 22). It was largely reclaimed in the seventeenth century after the St. Elizabeth Flood of 1421. This disastrous flood event destroyed 72 villages around the city of Dordrecht, and caused between 2000 and 10 000 casualties. Only the city itself was spared. Between 1700 and 1930, the polders in the south were dyked, and these dykes now function as regional flood defences. To the north of these regional flood defences lies the urban area, and to the south the agricultural and natural area. The city has approximately 119 000 inhabitants.

The Island of Dordrecht was a pilot study of the Dutch Delta Programme on resilient flood risk management, with a specific focus on ‘smart combinations’ of measures. The policy concept of a ‘smart combination’ provides for the possibility, in specific cases, to replace flood protection measures with measures involving prevention and preparedness. In these cases, prevention can be realized through sustainable spatial planning and (flood-proofed) adaptation of buildings, while preparedness can be improved by developing evacuation and rescue plans and preparing emergency stocks. The regional authorities of the Island of Dordrecht and central government have jointly commissioned this pilot study to gain experience with the application of a ‘smart combination’ in an actual context.

In the (first phase of) pilot study for Dordrecht, a smart combination of measures was found to be potentially promising for implementation. The proposal for this smart combination applied to both the northern and southern section of the dyke ring. It involved a reduction of the safety standard for the northern section from 1/3000 to 1/1000 per year and for the southern section from 1/100 to 1/300 per year, in combination with a compartmentalization strategy [39]. With this compartmentalization strategy, the at-risk area is reduced by splitting up the Island into smaller portions (figure 4). This is achieved by strengthening, and partially removing, the regional flood defences to keep water away from the urban area in the case of a river-dominated flood event (figure 5). This would contribute to the reduction of the number of casualties and damage, because the urban area is the most vulnerable area to flooding. Given these positive effects in terms of casualties and damage reduction, a lower safety standard is required for the primary flood defences. This is because the safety to flooding will be achieved through a combination of protection by the primary flood defences and prevention using the compartmentalization strategy. If this smart combination is to be implemented, then the legal instrument of a smart combination allows for the adjustment (i.e. lowering) of the legal standard pertaining to the primary flood defences. It also makes it possible to use the resulting savings for the measures on the regional flood defences—if this achieves the same level of protection.

Figure 4.

Flood resilient risk strategy for the Island of Dordrecht (b), together with the conceptual model of protection, prevention and preparedness (a) (courtesy of DeUrbanisten). (Online version in colour.)

Figure 5.

Effect of the compartmentalization strategy on the flood extent and depths in the case of a river-dominated flood event. (Online version in colour.)

In addition to the smart combination of measures, other measures with a focus on preparedness have been explored. The perspective for a preventive, organized evacuation to other dyke ring areas is quite pessimistic. The casualty risk can be considerably reduced if people can evacuate safely on the island itself. In this regard, vertical evacuation has been chosen as the starting point for evacuation. This implies that people will look for a shelter place in the (potentially) flooded area, either in their own home or in a public shelter. In the pilot study, the evacuation strategy has been operationalized by elaborating the preferred method of evacuation as well as the decision to evacuate in the event of a flooding threat. Moreover, the feasibility of the rescue and evacuation actions after a flood has been examined.

The aforementioned flood resilient risk strategy for the Island of Dordrecht has been compared with the traditional (resistant) strategy. The latter strategy is a continuation of the current (reference) flood risk strategy. It involves managing flood risk through protection by primary flood defences only. No measures are being taken to improve prevention and preparedness. As such, the spatial layout of the regional flood defences remains the same. The evacuation strategy for the island also remains the same, which is directed toward a preventive, organized evacuation.

The analysis and comparison of the two strategies showed that the adoption of a flood resilient risk strategy has several positive effects for the island.

It fulfils the requirements of the flood risk standards until 2050, which implies that the local individual risk should not exceed the value of 10^–5 per year. The local individual risk describes the probability of dying of an individual, present or not, at a certain location per year. Besides that, it decreases the probability of a large number of fatalities per year (that is, the societal risk) more effectively than the resistant strategy.

The level of ecological resilience is enhanced considerably by the resilient flood risk strategy. This was analysed by quantifying the system's ability to recover from flood impacts, with or without financial aid (figure 6). For the analysis of robustness, the recovery threshold was set at 5% of the regional gross domestic product (GDP) and the national GDP. These thresholds indicate whether financial aid of other Dutch regions and whether financial aid from other countries is needed [38]. It was found that the resilient strategy considerably reduces the economic damage for the more extreme flood scenarios compared with the resistant strategy. For this strategy, the response curve remains below the recovery threshold. This implies that the strategy is not dependent of financial support from other countries to recover the Island.

Figure 6.

Economic risk for the resilient flood risk strategy (solid line) and the resistant strategy (dotted line), with the recovery threshold of 5% of regional GDP (small dashed line) and national GDP (long dashed line) (source: [40]).

The resilient strategy also creates added value within the social and ecological domain. With this strategy, the preservation of monumental buildings on the primary flood defence, particularly along the Voorstraat, would be guaranteed. It also contributes to the further restoration of the Biesbosch ecosystem, which is Europe's largest fresh water tidal area. This co-benefit is linked to the realization of the compartmentalization strategy.

Given its positive effects, the smart combination whereby measures for compartmentalization lead to a lower standard for the primary flood defences, has been assessed as promising in the first phase of the pilot study. However, in a follow-up study, the current strength of the regional flood defences was analysed in more detail, showing that (i) this is very likely to be insufficient for a smart combination and that (ii) the necessary reinforcement works lead to higher investment costs than the potential savings on the primary flood defences. With that, the smart combination of measures has not been considered further for implementation, nor has it been subject to open public consultation.

Yet, the pilot study has resulted in an additional strategy for preparedness to flooding that has been translated into a research and implementation agenda. This strategy has been elaborated in a proposal for a preferred evacuation strategy (vertical evacuation) and a prioritization, if more time and resources are available before the onset of the flood. Here, a distinction is made between three aspects: evacuation zoning, healthcare institutions and vital infrastructure. For each of these aspects, building blocks for the most optimal prioritization of actions and resources have been worked out. The proposed evacuation strategy will lead to a significant reduction in the number of fatalities compared with a preventive, organized evacuation. Therefore, the regional authorities have decided to further develop this strategy into a fully-fledged evacuation strategy. This process will be informed through a process of public consultation and engagement, building upon a recent survey on the social attitude toward vertical evacuation. The results of this survey indicated that about three-quarters of the inhabitations are supportive of vertical evacuation. However, more than a quarter gave this strategy an inadequate score. The average score being 6.3 out of 10. This support base is mainly determined by the degree to which the respondents think the strategy is safe (positive effect) and the degree to which it evokes feelings of fear and unease (negative effect). Two-thirds of the respondents consider vertical evacuation safe, but the strategy also evokes feelings of fear and uncertainty with about half the respondents. The limited support for vertical evacuation contrasted strongly with the finding that 88% of the inhabitants indicated (very) likely to stay at home in case of a flood threat. This finding must therefore be seen as a ‘forced choice’, where staying at home prevails strongly over other options such as hiding in a shelter or a preventive evacuation to other dyke ring areas.

(b). Bottom-up flood resilience in Can Tho, Vietnam

Can Tho is the biggest city in the Mekong Delta in Vietnam. Can Tho has been affected by floods and the risk of flooding is exacerbated by demographic changes and human interventions in the Mekong River basin; changing rainfall pattern; and increasing sea levels [41]. In the year 2011, the average direct damage of individual household due to flooding was about USD 822, which is substantial as the monthly household income of 90% of flood affected houses is less then USD 400 [42]. Can Tho has been adapting to flooding through (i) top-down measures such as building of dykes along the Mekong River and improvements to drainage systems and (ii) bottom-up measures such as flood proofing of houses by elevating the floor levels and temporary dykes around properties [43,44] (figure 7).

Figure 7.

Flood risk reduction measures as blocking the entrance of houses using sand bags (a) and elevation of floor level (b) in Can Tho [45]. (Online version in colour.)

In a survey conducted aftermath, the 2011 Can Tho floods 90% of respondents stated that flooding up to 20 cm with in the houses for an hour is acceptable [44]. This is in alignment with the ‘living with water’ way of life that has been in prevalent in the Mekong delta over the past centuries [46]. This is an indication of the importance of the coping and updating capacity of those who are affected by the flood. Flood resilience is no longer limited to the ‘engineering’ purview in Can Tho, but has social evolving into a socio-technical resilience system [47]. Although the socio-technical nature of flood resilience increases the complexity to manage flood resilience, it also provides opportunities to address or increase flood resilience in the context of Can Tho through interventions which are either social or technical (engineering) nature. For example, if all the houses in Ninh Kieu district of Can Tho resort to adaptation measures then big infrastructure measures such as dykes or drainage systems can be deferred by increasing the floor levels of houses by 50 cm in order to minimize the flood damages [48]. Under such circumstances building of dykes around Ninh Kieu district can be deferred till sea level increases by 38 cm, which is likely to occur between the years 2042 and 2055 according to IPCC scenarios [48].

Adapting to flooding at household levels has been a common practice in Can Tho in the past. There is evidence of households resorting to the practice of elevating the floor levels since 1960s as a response to flooding [49]. According to Garschagen [49], the floor elevation ranges between 20 cm and 50 cm on average and in certain cases up to 150 cm. The elevation of floor levels and the height of elevation were not only driven by the increasing water levels, but was also based on the household's affordability to elevate the floor levels. Elevating the floor levels more than 50 cm was five times more expensive than elevating up to 50 cm, as major structural changes such as modifying doors, windows and roof were involved. This resulted in poor households not adapting to floods in spite of facing damage during flooding [45]. Hence instead of investing in large-scale risk reduction measures, such as dykes, poor households can be given a subsidy or grant for elevating the floor level to reduce the vulnerability due to flooding.

Understanding the bottom-up nature of flood resilience measures such as adaptations at household level—both engineering and social in nature—in Can Tho can lead to alternative adaptation strategies, such as adaptation pathways. The realization that flood resilience encompasses both engineering and social elements in Can Tho has resulted in the generation of multiple adaptation pathways for a range of future scenarios (figure 8). Top-down measures such as dykes and bottom-up measures such as floor elevation of houses can be combined using adaptation pathways and the tipping of the pathways for the four different IPCC AR5 climate scenarios are shown in figure 8. This can help decision makers in the selection and implementation of long-term flood resilience measures. Comparison of costs and benefits along a path in different scenarios—such as seal level rise or a combination of drivers such as urbanization, GDP growth—can lead to a well-informed decision based on context-based relationship between adaptation measures and drivers [50]. Radhakrishnan et al. [50] have compared the costs and benefits of several adaptation pathways for building dykes, where the new action is triggered by increasing water levels and the evaluation of the costs and benefits which are based on urbanization and GDP growth in Can Tho. However, decision making should not only be on the basis of economic cost and benefits, where objectives such as eco-systems benefits and sustainable development are not taken into consideration.

Figure 8.

Adaptation pathways comprising top-bottom and bottom-up flood resilience measures in Can Tho for sea-level rise based on IPCC climate scenarios [48]. (Online version in colour.)

Representation of direct flood damage in monetary terms is not the real representation of overall flood damages as observed in the case of Can Tho, where there are direct and indirect losses. During the 2011 flooding in Can Tho, the direct damage per households was about USD 333, the direct damage to the individual business was about USD 152 and the losses due to business disruption were about USD 207. Although these numbers might appear small, they should be valued in the context of an average monthly income of a household of USD 185 [45]. Multi-variate analysis of flood damage in Can Tho reveals that factors such as precautionary measures, socio-economic status, incomes, duration of flooding and velocity of flooding have an influence on the magnitude of damage in households [42]. This is also evident from the residents' willingness to tolerate certain depth and duration of floods (figure 9), which can be considered as a measure of adaptive resilience. However, this social limit of resilience is not likely to remain constant for a longer period and can reduce in the future as there is an aspiration among the household and policy-level decision makers to become like the cities in developed countries [51,52]. Hence this conundrum of decreasing resilience against the backdrop of increasing prosperity should be considered in long-term planning of resilience building in communities to retain the social resilience capital. Reduction in limits of resilience is also true for engineering measures, such as the dykes in Can Tho, though designed for a water level of certain return period, the time up to which the dykes can be functionally effective are likely to reduce due to the uncertainty in sea level due to climate change, as shown in figure 8 [48]. This reduction in limits will have to considered or incorporated when aggregating the impact of engineering and adaptive resilience measures.

Figure 9.

Willingness of residents to accept flooding inside houses in Can Tho [44]. (Online version in colour.)

One of the ways to retain the level of social resilience is to actively engage with the residents and resolve the current flood water quality problem in Can Tho [53]. Nguyen et al. [53] have debunked the ‘living with water’ concept in Can Tho by confirming the presence of Escherichia coli and rotavirus in the flood waters in Can Tho. They recommend to raise awareness of the public about this danger and to put in place sewerage and drainage systems to mitigate associated potential health risks. However, openly discussing about the flood water quality might create a sense of discomfort and panic among the residents of Can Tho in the beginning and might lead to unacceptance of temporary inundation of houses and neighbourhoods. Instead, a genuine engagement with the residents, implementing sewerage systems and encouraging citizen science initiatives, such as flood level measurement and water quality measurement using mobile applications, will likely retain or even increase the flood resilience of the residents.

Can Tho can be seen as a ‘living laboratory’ where potential interventions across sectors—such as risk reduction and poverty alleviation—can increase flood resilience, facilitate participation and collective action. Analysing the flood risk and evolution of flood resilience in Can Tho, it can be stated that (i) flood resilience is adaptive in nature, where the shifting of stability domain has moved from ‘no flooding within or outside the house’ to ‘limited flooding for a limited duration within and around the house’ during the flood seasons; (ii) the flood risk management system in Can Tho is a social–technical system, which enables the adoption of strategic alternatives such as adaptation pathways that comprise bottom-up as well as top-down approaches; (iii) flood resilience is influenced by drivers beyond physical factors such as rainfall and sea-level rise and can be improved by addressing the vulnerability.

6. Summary and concluding remarks

In this chapter, different conceptual frameworks of resilience including engineering, ecological and social–ecological have been presented and framed within the context of flood risk management. A review of the literature revealed that the boundaries between those frameworks are partly overlapping and in some aspects competing. Socio-ecological resilience is conceived as a broader system characteristic that involves the interaction between human and natural systems. It acknowledges that these systems change over time and that these interactions are of complex nature and associated with uncertainties. Social–ecological resilience, therefore, is closely linked to the ability to build capacity to enhance long-term resilience through learning. Building (socio-ecological) resilience in flood risk management strategies calls for an adaptive approach with short-term measures and a set of monitoring criteria for keeping track of developments that might require adaptation in the long-term (adaptation pathways) and thus built-in of adaptive capacity as opposed to building engineering resilience which involves a static approach with a fixed time horizon a set of robust measures designed for specific future conditions or scenario's.

The two case studies presented in this chapter indicate that the concepts of ecological and socio-ecological resilience provide guidance for building more resilient flood risk management systems resulting in an approach embracing flood protection, prevention and preparedness. The case studies also reveal that the translation of resilience concepts into practice remains a challenge. One plausible explanation for this is our inability to arrive at a quantification of socio-ecological resilience taken into account the various attributes of the concept. For example, valuing the benefits of flood protection measures to enhance the flood resilience is constraint by the long timeframes and low frequency of extreme events. Other explanations can be derived from the two case studies. For The Netherlands and countries with similar flood risk strategies (focused on protection), the pilot study in Dordrecht learns to be realistic about the feasibility of a shift from a traditional to a resilient approach. This was evident for the policy concept of a ‘smart combination’, which makes it possible to replace flood protection with measures involving prevention and preparedness. The limitations on costs and the weighting of flood consequences with respect to the (high) safety standard reduce the chances of successfully finding a cost-efficient smart combination. By contrast, the pilot study also learns that the opportunities for resilient flood risk management in a broad sense, which is by managing flood consequences through additional measures are ample. Vertical evacuation is one such measure, which—when applied in addition to current protection measures—will lead to a significant reduction in the number of fatalities. This justifies to recommend the resilience approach for implementation in Dordrecht and elsewhere. Yet, implementation will only be successful if public attitudes toward managing flood consequences are taken into account into the development of measures like evacuation strategies. For example, it was observed in Dordrecht that vertical evacuation evokes feelings of fear and uncertainty among the inhabitants, which should be addressed through communication (among other).

Data accessibility

This article has no additional data.

Authors' contributions

C.Z., B.G., M.R. and C.Z. wrote chapters 1–4, and co-wrote conclusions; B.G. wrote chapter 5a and co-wrote conclusions, M.R. wrote chapter 5b and co-wrote conclusions. C.Z. revision of the article and final approval of the version to be published.

Competing interests

We declare we have no competing interests.

Funding

We received no funding for this study.