Assessment and Adaptation to Climate Change-Related Flood Risks
Summary and Keywords
The flooding of rivers and coastlines is the most frequent and damaging of all natural hazards. Between 1980 and 2016, total direct damages exceeded $1.6 trillion, and at least 225,000 people lost their lives. Recent events causing major economic losses include the 2011 river flooding in Thailand ($40 billion) and the 2013 coastal floods in the United States caused by Hurricane Sandy (over $50 billion). Flooding also triggers great humanitarian challenges. The 2015 Malawi floods were the worst in the country’s history and were followed by food shortage across large parts of the country.
Flood losses are increasing rapidly in some world regions, driven by economic development in floodplains and increases in the frequency of extreme precipitation events and global sea level due to climate change. The largest increase in flood losses is seen in low-income countries, where population growth is rapid and many cities are expanding quickly. At the same time, evidence shows that adaptation to flood risk is already happening, and a large proportion of losses can be contained successfully by effective risk management strategies. Such risk management strategies may include floodplain zoning, construction and maintenance of flood defenses, reforestation of land draining into rivers, and use of early warning systems.
To reduce risk effectively, it is important to know the location and impact of potential floods under current and future social and environmental conditions. In a risk assessment, models can be used to map the flow of water over land after an intense rainfall event or storm surge (the hazard). Modeled for many different potential events, this provides estimates of potential inundation depth in flood-prone areas. Such maps can be constructed for various scenarios of climate change based on specific changes in rainfall, temperature, and sea level.
To assess the impact of the modeled hazard (e.g., cost of damage or lives lost), the potential exposure (including buildings, population, and infrastructure) must be mapped using land-use and population density data and construction information. Population growth and urban expansion can be simulated by increasing the density or extent of the urban area in the model. The effects of floods on people and different types of buildings and infrastructure are determined using a vulnerability function. This indicates the damage expected to occur to a structure or group of people as a function of flood intensity (e.g., inundation depth and flow velocity).
Potential adaptation measures such as land-use change or new flood defenses can be included in the model in order to understand how effective they may be in reducing flood risk. This way, risk assessments can demonstrate the possible approaches available to policymakers to build a less risky future.
People are naturally attracted to living close to rivers and seas in order to benefit from opportunities offered by the water, such as transport, irrigation, and drinking water. Ever since ancient times, regular flooding of these river and coastal systems has been an inevitable aspect of life in low-lying areas. Floods, especially the smaller and more frequent ones, bring a range of benefits themselves, such as increasing the fertility and nutrient levels of soils, recharging groundwater levels, and maintaining unique floodplain biodiversity (Reice, 2003). However, strong floods also bring substantial risk to lives and livelihoods. Flooding is the most frequent and damaging of all natural hazards. Since 1980, the reported direct economic losses are well in excess of $1.6 trillion in the period 1980–2015, and more than 225,000 people have lost their lives (Munich Re, 2016) (Figure 1).
The main causes of increasing flood losses are population growth, urbanization, and economic growth in flood-prone areas. Around the world, cities located on river deltas are hotspots of expansion and economic development, especially in developing countries. The global population exposed to river and coastal flooding doubled, increasing from around 520 million people in 1970 to almost 1 billion in 2010 (Jongman, Ward, & Aerts, 2012). Population growth is expected to continue this trend into the future, with a world population that is expected to increase to between 9.0 and 13.2 billion by 2100 (Gerland et al., 2014). More people living in flood-prone areas means that the same floods that may have had a minor impact decades ago can have devastating consequences today. Damages from river flooding could be 20 times greater by the end of the century (Winsemius et al., 2015).
The floods themselves are also subject to change, as climate change is altering weather patterns and leading to rising sea levels. As the atmosphere warms, it contains more energy and moisture, which is already leading to more intense rainfall and river flooding in some regions. Other areas may become drier and see a decrease in river flooding. Sea level rise and a possible intensification of storms are increasing the risk of coastal flooding in all coastal areas.
When a flood occurs in a populated area, the impact of that flood is highly dependent on the vulnerability of those affected. Their vulnerability describes the ability of the population to withstand the impact and effectively respond to the event. Community vulnerability depends on things like the strength and quality of buildings and infrastructure and whether early flood warnings are provided. Levels of vulnerability differ around the world and change over time—it is well documented that people in low-income countries are much more vulnerable to negative flood impacts than those living in high-income countries (Jongman et al., 2015).
As the climate changes and flood events become more frequent and damaging, adaptation is essential to keep impacts at acceptable levels and to secure the livability of low-lying areas. Society has a range of options for adaptation, which can be implemented by individuals (e.g., floodproofing of buildings), governments (e.g., levees and land-use planning), the private sector (e.g., insurance products), and the international community (e.g., international risk financing strategies). The poorest people are especially vulnerable to the shocks inflicted by floods. Without increased adaptation, it is estimated that an additional 100 million people will fall into poverty due to floods and droughts (Hallegatte et al., 2015).
The accurate assessment of flood risk and its driving factors is essential as a basis for deciding on efficient adaptation pathways. This article aims to describe the factors contributing to global flood risk, the methodologies for assessing coastal and riverine flood risk, and the ongoing efforts to adapt to floods.
The Evolution of Flood Risk
Flooding, defined as the inundation of predominantly dry land, has occurred ever since water bodies existed on this planet. As societies around the world developed, human presence increasingly intersected with this phenomenon. Currently, the most densely populated and rapidly developing areas are located near rivers or on deltas at the coast. Combined with increasing sea levels and possible changes in rainfall intensity due to climate change, flooding has evolved into the most frequent and damaging natural hazard that people face today.
Flood risk is driven by the three components of risk: hazard, exposure, and vulnerability (IPCC, 2012) (Figure 2). The hazard represents the intensity of the flood event, such as the inundation depth and flow velocity. Flood hazard alone is not necessarily problematic, so long as it occurs in uninhabited areas. The exposure part of the equation refers to the people and the economic value of assets located in flood-prone areas, which are potentially affected during a flood. Finally, vulnerability refers to the susceptibility of the exposed elements to the hazard.
The drivers of flood risk are influenced by human and economic development, climate change, and disaster risk management. As the climate is warming and sea levels rising, the intensity and frequency of river and coastal floods may change. With growing population numbers and expanding cities, the exposure to floods increases around the world, and with economic development, changing building quality, and planned risk management interventions, the vulnerability of the exposed elements may change. In this section, these drivers of flood risk and their contribution to global losses are discussed.
Hazard: Climate and Flooding
To assess flood hazard, the relevant flood processes and the drivers that lead or contribute to these flood processes should be known and the behavior of the drivers in space and time understood. Typically, flood processes are subdivided into three categories: coastal flooding, river flooding, and pluvial flooding. In this section, we discuss these three flood processes and the climate drivers leading to them. It should be noted that in many regions (especially deltas), these processes may strongly interact. For example, Wahl, Jain, Bender, Meyers, and Luther (2015) showed that this interaction exists across a major part of the U.S. coastline, and that these interactions have been increasing over the past decades.
River and Flash Flooding
River and flash flooding are among the world's dominant flood processes. River flooding mainly originates from heavy rainfall within the contributing catchment area of the river. Flash floods can generally be classified as floods in smaller headwater rivers with relatively small upstream catchment areas, which occur in hilly or mountainous areas within a short period of time (i.e., a few hours). The most important drivers of these floods are heavy and relatively local convective thunderstorms that lead to a large amount of rainfall within a river’s catchment area. The rainfall intensities accompanied with such rainstorms are strongly related to the amount of moisture the air can contain (also called atmospheric moisture content). As this is nonlinearly dependent with temperature, the heaviest thunderstorms occur with high air temperatures. With a rising temperature and more extremely hot days, therefore, we can expect that on average, the frequency and intensity of heavy thunderstorms will increase globally. This is shown, for instance, by Trapp et al. (2007) over the United States. Very locally, very intense rainfall events even scale more nonlinearly potential atmospheric moisture content, and this is particularly relevant for local pluvial flood hazard, described in the section “Pluvial Flooding” (Lenderink, Mok, Lee, & van Oldenborgh, 2011; Lenderink & van Meijgaard, 2008).
River flooding represents a much larger scale of flooding that occurs due to rainfall events with a generally much longer duration and falling over a much larger upstream catchment area. Hence, the main driver of such floods is mostly advective storms or seasonal rainfall. For instance, to assess flood hazard in the Netherlands, a 10-day rainfall accumulation is often used as a proxy (see, e.g., Kew, Selten, Lenderink, & Hazeleger, 2013). Changes in river flood hazard due to climate change, therefore, are ambiguous and difficult to model. Although in many areas, flood models forced by climate model outputs show increases in flood hazard, other areas may see decreases in flooding under climate change (Hirabayashi et al., 2013; Winsemius et al., 2015).
It should be noted that river flood hazard also may strongly depend on human interventions in the basin. For instance, reservoir operations may either lead to a reduction in flood hazard, when a reservoir is operated with the target to store floodwater, or an increase, when storage is used for other purposes such as hydropower. The preconditions of heavy rainfall events are very important in these cases, as they may cause a higher or lower probability that a reservoir is capable of capturing the additional floodwaters. For instance, reservoir conditions and operations are believed to have played a significant role in the Brisbane floods in 2011 (see, e.g., van den Honert & McAneney, 2011).
Pluvial flooding is flooding that occurs due to very local rainfall. Most cases of pluvial flooding can be observed in low-lying, flat polder areas and urban centers. Within cities, the occurrence of pluvial flooding heavily depends on two main things. The first element includes the amount of urbanization and the capacity of the soil to absorb rainfall (the degree of accommodation to handle local rainfall by either conveying the floodwater downstream with drainage and sewerage infrastructure or temporarily storing it in storage ponds or underground sewerage basins). As in the case of flash flooding, heavy rainfall events that are local both in space and time are the most relevant phenomena causing pluvial flooding. These, again, typically occur during local thunderstorms, and due to the higher convective potential, the most extreme rainfall events correlate with days with very high temperatures and potential atmospheric water vapor content. Lenderink et al. (2011) and Lenderink and van Meijgaard (2008) show that the increase in potential rainfall intensity with temperature can be as high as between 7%–14% per degree Celsius. Hence, as climate change progresses, more intense rainfall intensities and more frequently occurring pluvial floods may be expected. This has implications for climate-robust design of urban and polder infrastructure.
Extreme water levels and flooding along the coast occur due to high tides, storm surges, high waves, or a combination of these. A storm surge is a rise in the sea surface caused by storms with low-pressure and strong winds, like extratropical cyclones or tropical cyclones, such as hurricanes and typhoons. Tropical cyclones have lower interior pressures and higher wind speeds than extratropical cyclones, and as such, they typically produce much higher surges than extratropical storms. In addition to the storm characteristics (wind speed, pressure, angle, and size), the height of the storm surge is strongly influenced by the local characteristics of a coastal areas, like the nearshore bathymetry and the geometry of the coastline (Resio & Westerink, 2008). Essentially, shallow areas with a wide continental shelf, such as the east coast of the United States or Germany, will experience much higher storm surges than areas with steep offshore slopes, such as Caribbean islands.
Increasing Flood Exposure
Increasing exposure due to rising population and urbanization in flood-prone areas is considered the main driver of increasing flood losses over the past decades (Bouwer, Crompton, Faust, Hoppe, & Pielke, 2007). The world’s population rose from 3 billion in 1960 to over 7.5 billion in 2017. By the year 2100, the global population is expected to reach between 9 and 13 billion (Gerland et al., 2014). A large part of the population growth takes place in flood-prone areas. As a result, the global population exposed to river and coastal flooding is estimated to have increased from 520 million in 1970 to 1 billion in 2010 (Jongman et al., 2012), and this trend is likely to continue into the future (Figure 3). Due to the attraction of water bodies, the growth of population and cities within flood-prone areas is estimated to be higher than outside of flood-prone areas in many world regions.
In most regions of the world, and most profoundly in developing countries in Africa and Asia, population growth is combined with high levels of urbanization as people move to cities for jobs and markets. Cities are dense, highly concentrated locations of exposure and are traditionally often located near rivers or the coast. Hallegatte, Green, Nicholls, and Corfee-Morlot (2013) estimate that the growth of exposure in coastal cities, even without climate change, may lead to a tenfold increase in economic losses by 2050 compared to 2005.
The rapid development of socioeconomic activities in flood-prone areas, including industrial, service, and trading, drives large increases in economic exposure. During the 2011 Thailand floods, for example, over 7,500 industrial facilities were affected across 40 provinces, which disrupted global trade flows and caused an estimated $40 billion in damages.
Another factor that contributes further to increasing exposure is the so-called levee effect (Di Baldassarre, Kooy, Kemerink, & Brandimarte, 2013). As developed areas are being protected by flood defenses, the perceived increased level of security may cause even more people to move to that specific area. Population, urbanization, and flood protection, therefore, are observed to reinforce one another and lead to continuously increasing exposure in flood-prone areas.
Vulnerability and Adaptation
Whereas the exposure of people, buildings, and infrastructure to flooding is increasing steadily as societies develop and the world’s population grows, this does not mean that economic losses increase at the same pace. The impact that floods have on the exposed elements is determined by their vulnerability. Vulnerability refers to the susceptibility of the exposed assets and people to the impact of the hazard, and it has multiple facets (Fraser et al., 2016). Physical or structural vulnerability determines the physical and economic damage to buildings and infrastructure, often depicted as a percentage of their total value (see the section “Exposure and Vulnerability”). Physical vulnerability largely depends on the building material and quality of the asset, as well as any specific measures that have been taken to prevent flood damage (Smith, 1994). Social vulnerability refers to the ability of a community to cope with the impact of asset losses on their livelihood. Social vulnerability, therefore, largely depends on people’s characteristics such as education, age, wealth, degree of access to resources, and political power (see, e.g., Cutter, Boruff, & Shirley, 2003; Koks, Jongman, Husby, & Botzen, 2014).
Vulnerability is dynamic over space and time. For example, it is found to be greater in low-income countries than in high-income countries (Figure 4) due to differences in both social development and asset engineering (Jongman et al., 2015; Mechler & Bouwer, 2014). On a global scale, the total level of vulnerability is observed to be declining due to increasing levels of development and disaster risk management, which is reflected in decreasing loss of life in developed countries (Jongman et al., 2015; UNISDR, 2011). However, without adaptation and infrastructure maintenance, physical and social vulnerability may increase over time. This is especially true in rapidly urbanizing areas, where buildings and infrastructure are often being constructed at low cost and without a clear maintenance plan.
Not only are the world’s poor more exposed to flooding, they are also among the most vulnerable to the impact of shocks caused by floods (Hallegatte et al., 2015). Poor households spend a relatively high percentage of their income on basic needs such as food and housing, and they are less able to rebound from shocks caused by floods. In addition, they often have less access to information (e.g., regarding flood forecasts) and live in houses that are less resistant to the physical impact of flooding.
Various actors are involved in adaptation and the reduction of vulnerability. Governments at both the local and national levels have an influence on building codes and socioeconomic development. They can ensure that new construction in flood-prone areas takes into account potential flood hazards, and they may be able to address social vulnerability by improving access to healthcare, education, and information. International development organizations are increasingly involved in vulnerability reduction in developing countries by providing financing and knowledge.
Recent research on the field of socio-hydrology shows this strong relationship between humans and flooding (Di Baldassarre et al., 2015). By building flood protection, societies can create highly technological systems that have lower losses overall; however, once a flood does happen, the impacts can be catastrophic (Ciullo, Viglione, Castellarin, Crisci, & Di Baldassarre, 2017).
Therefore, whereas it is clear that a large share of flood risk can be reduced, it is equally clear that there are limits to adaptation. These limits are endogenous to society and depend on knowledge, attitude, risk, and culture, as well as our limited ability to have and appreciate foresight of risk (Adger et al., 2009; Dow et al., 2013). However, whereas these limits are complicated and inherent to human sociology, they are understood to be mutable by effective governance.
Assessing Risk and Adaptation Options
The general understanding of the underlying drivers of flood risk (as discussed in the previous section) allows us to consider the possible sources of and changes in flood risk. However, sometimes a more detailed understanding and quantification are needed, in the form of risk assessment. Flood risk assessments entail the mapping and quantification of the economic and/or human damage caused by potential river or coastal flooding. As such, risk assessments can provide exact details of the geographical reach and intensity of potential floods, which can be combined with exposure and vulnerability details to compute financial losses.
A well-executed flood risk assessment, therefore, can effectively inform risk identification, risk reduction, preparedness, financial protection, and resilient reconstruction. Assessments provide the basis for resilient decision-making and disaster risk management across sectors. There is a wide array of risk management policies and solutions available to policymakers, including structural defenses, early warning systems, and nature-based solutions (as discussed later in this article). The decision on each of these possible options should be based on an accurate assessment of risk; therefore, risk assessment is included as a key priority in the post-2015 Sendai Framework devised by the United Nations (UNISDR, 2015).
Flood risk assessment needs to take into account the dynamic nature of risk, which is constantly changing due to socioeconomic growth, climate change–induced changes in rainfall and sea levels, and disaster risk management. As such, interdisciplinary modeling approaches need to be employed to combine the modeling of hazard, exposure, and vulnerability and incorporate scenarios of future changes in the underlying components. In this section, we discuss the common approaches for coastal and riverine flood risk assessment.
Flood Hazard Mapping
The section “The Evolution of Flood Risk” described how various flood processes are caused by different drivers, and so establishing the level of flood hazard for different frequencies of occurrence starts with identifying the relevant drivers, and possibly their interactions. Here, several approaches to mapping flood hazards from these processes are described.
Mapping of Flash and River Flooding Hazard
Flash and river flooding hazard ideally is mapped with very long series of inundation observations from floods that have occurred in the past. Flood mapping can be done with remote-sensing imagery at low resolution for large-scale flooding, and at high resolution for local flooding. The main problem with this approach is that satellite images are not necessarily available during the flood, and even if enough images can be collected, the time series available may be too short to assess very rare but destructive events. Furthermore, additional information often needed to define the severity of a flood, such as water levels and flow velocities, cannot be derived from satellite data.
Hence, computer models are often used to assess flood risk. For a given flood-prone area, a typical method to do this is to estimate several discharge events, also called flood waves and with given return periods, and to use a computer model that simulates the movement of water through the river and its surrounding floodplains, in order to assess how the downstream flood-prone area floods. Critical information for such a flood model includes the river profile, which determines the conveyance capacity of the river; the roughness of the river bed and floodplains (strongly dependent on the river bed material and the presence of vegetation or other obstructions on the banks and floodplains); and the elevation of the floodplains surrounding the river. This approach is often used on a local scale, but recently, global-scale flood maps also have been derived with such an approach, created by stitching together many local flood simulations with upstream flood waves (Sampson et al., 2015). This poses a challenge to deriving global coverage of representative and reliable flood waves.
In the approach described here, any local rainfall that may aggravate flooding within the flood-prone area of interest is disregarded. This method, therefore, is appropriate when the contributing area upstream is far greater than any local contributing area within the flood-prone area. For river floods, which are much larger in scale, this may be the case, but for local flash floods, local rainfall often is key to assessing flood hazard accurately.
An important question is how to derive flood waves that are representative of the flood-prone area investigated. Although they may be derived from extreme value statistics of observed river flows, such observations often are not available or not long enough to derive reliable flood waves. Therefore, hydrological models, forced by time series of rainfall, temperature, and potential evaporation, are used to establish flood-wave scenarios. If one wishes to search for these events in an empirical distribution of events, such series may have to be very long. In the case of the Netherlands, dikes are dimensioned to withstand a very rare event (i.e., occurring once in 1,250 years), and here, a method called Generator of Rainfall and Discharge Extremes is being developed to establish time series of many tens of thousands of years of discharge, which are consequently used to derive representative flood waves.
An alternative approach is to start the statistical analysis within the meteorological domain and derive a set of rainfall events that are representative of certain return periods. These can then be fed into a hydrological model to derive accompanying discharge events. An additional challenge with this approach is that the expected runoff and flood wave from any given meteorological rainfall event may be strongly dependent on the amount of moisture in the catchment area. Therefore, with this approach, combinations of a moisture state and rainfall events often are computed to come up with representative flood waves.
How climate change enters these methods strongly depends on the scale that is investigated. Global-scale studies generally make long simulations within the present-day climate and future climate (using climate model outputs) with a cascade of a hydrological model and a river flood model (Hirabayashi et al., 2013; Winsemius et al., 2015). A large-scale river flood study also may rely on the generation of a rainfall event set, fed into a hydrological and hydraulic model. These rainfall event sets are then scaled using rainfall from climate model outputs to generate future event sets. But if very local flooding is concerned and local intense rainfall from thunderstorms is the predominant mechanism causing floods, an alternative approach is required, as such local rainfall cannot be accurately resolved by a climate model. A method of working at this scale is to find relationships between large-scale atmospheric conditions (which are more accurately described by climate models) or apply model output statistics (MOS) on climate model outputs. MOS means that statistical corrections and disaggregation in space and time are derived for simulated precipitation based upon ground observations.
Coastal Flood Hazard Mapping
The traditional method to estimate the frequency of coastal flooding is based on the analysis of sea level observations (Coles, 2001). A major limitation of this approach is that you typically need 30 years or more to reliably estimate return periods, and unfortunately, observation stations can fail during an extreme event. Furthermore, no observations are available for many areas in the world, and as many local factors strongly influence the sea level, it is difficult to interpolate along the coast. Hence, computational modeling is nowadays the state-of-the-art method to assess coastal flood hazards. By forcing a hydrodynamic model with atmospheric pressure and wind speed from climate models or reanalysis data, it is possible to generate time series of historic sea levels. Recently, studies have applied hydrodynamic models on the continental to global scale to provide multidecadal hindcasts of sea levels (Cid, Castanedo, Abascal, Menéndez, & Medina, 2014; Haigh et al., 2014; Muis, Verlaan, Winsemius, Aerts, & Ward, 2016). On the local scale, the models typically have a higher resolution, and to improve their accuracy, they are coupled with other models to simulate combined effects of tides, storm surges, waves, and river discharge (Dietrich et al., 2010).
By 2100, the global mean sea level is projected to be 25–123 cm higher than the 1985–2005 reference period (Hinkel et al., 2014). Sea-level rise in the tropics can be tens of centimeters higher than at high latitudes due to gravitational and rotational effects from changes in ice masses and changes in ocean circulation. The simplest method to estimate the impact of changes in sea level is by adding the sea level rise to estimated return levels for the current climate. However, there may be nonlinear effects in some areas (Zhang et al., 2013). The effects of changes in the weather system on the frequency of coastal flooding are less well researched. However, recent studies have demonstrated that it is possible to assess the effects of these changes on coastal flood hazard by forcing a hydrodynamic model with projected wind and pressure fields from climate models for various emission scenarios (Vousdoukas, Voukouvalas, Annunziato, Giardino, & Feyen, 2016). A physically based approach is also used to assess the effects of changes in the frequency, intensity, and tracks of tropical cyclones on coastal flooding.
Once extreme sea levels are available, flood hazard maps can be produced that show the extent of a coastal flood for a specific return period (Figure 5). Two main methods are used to map the flood extent: a hydrodynamic approach and a static approach. The static approach, the so-called bathtub approach, is often applied in a geographic information system (GIS). It assumes that any land located below the extreme sea level will be inundated. It generally overestimates the inundation extent, but it provides a good indication of the area potentially at risk. Furthermore, due to the simplicity of the method, it is computationally cheap and easy to apply to large areas.
The hydrodynamic approach includes the physical processes involved in coastal flooding, like the temporal dynamics of a flood wave or the attenuation of the flood wave due to the roughness of the land or vegetation (Bates et al., 2005). These physics are particularly important in flat, low-lying, and vegetated areas. The results of both the static and the dynamic methods are strongly influenced by the quality of the elevation data that are available. For instance, the globally available elevation map from the Shuttle Radar Topography Mission (SRTM) contains bias, striping, and offsets due to buildings and trees and random noise. In general, particularly for coastal flooding, the existence of offsets due to trees and other structures that obstructed a clear view of the ground of a space-based or aerial vehicle instrument, and differences between the mean sea level and the geoid used to reference the elevation data often pose challenges to adequately conveying an extreme water level onto the land.
Exposure and Vulnerability
In flood risk assessment, hazard maps are combined with information on exposure and vulnerability to compute the economic and human impacts. Exposure information can come in many different forms. Maps of land use and population density provide a view of the geographic distribution of buildings, crops, and people: vital inputs to a flood risk assessment. The scale of a flood risk assessment determines the level of exposure information required. On the global or regional scale, it may be sufficient to estimate the area of cropland or number of buildings or population affected. Global maps showing density of population, different land cover types (e.g., crops of forest), or buildings have been generated from satellite images and remote sensing, with enough detail for use in flood risk assessment. These generally show the density of these exposure types within individual grid squares, which are generally tens to hundreds of square meters in size (e.g., CIESIN, 2016). On the local scale, risk assessments are more likely to be focused on the design of flood protection strategies, which require more specific information on the location and characteristics of infrastructure networks and individual buildings at risk. OpenStreetMap is one such source of detailed location information, although additional surveys are often required to provide sufficient characteristic information.
To assess risk, knowledge about the vulnerability of the affected people, buildings and infrastructure is required. Some types of buildings suffer more damage than others with the same depth and duration of flooding. The same is true of crops, infrastructure, and people (whether individuals or whole communities). Generally, vulnerability curves are used to convert flood depth into a level of damage, with each type of building or other item having a different curve (Figure 6). A building vulnerability curve is influenced by things like construction material and quality and height of the ground floor. Vulnerability of people is more affected by wealth and gender and may be represented in terms of vulnerability indexes instead of curves.
It is important to include changes in exposure and vulnerability in estimates of future flood risk: these are constantly changing and will influence future flood losses. To make estimates of flood risk in, say, 30 years’ time as reliable as possible, we need to have an idea of how big a city and its population will be and what type of buildings will be built (and where) between now and then. This information will influence the whole risk assessment chain: how likely heavy rainfall will translate into flooding; the number of affected people, buildings, and other elements; and the resulting level of damage or loss. A number of models today can project such changes (e.g., Linard et al., 2017), based on previous patterns of urban expansion (Angel, Parent, Civco, & Blei, 2013).
In some countries, planning laws and government regulation have resulted in detailed, up-to-date population records, maps of buildings and infrastructure, and building regulations providing details on how various structures perform during floods. These provide a good basis for risk modeling. However, there are many data-scarce regions of the world, where exposure and vulnerability information are incomplete, out of date, or entirely absent because data collection is not standard practice and construction is commonly unregulated. In such cases, it is difficult to assess accurately what might be affected by floods and what the resulting level of damage might be. In such areas, exposure data can be estimated by interpreting features on the ground shown in satellite images and aerial photography, or learned about through ground-based surveys, although to cover large areas in ground-based surveys requires a lot of time and personnel.
Flood Risk Assessment
Using the methods and data described previously in this article, it is possible to develop maps and data sets to represent hazard, exposure, and vulnerability. By combining these elements in a flood risk model, it is then possible to estimate flood risk. A conceptual example of how these different data sets can be brought together is shown in Figure 7. In this case, hydrological data (like measured or observed discharge) are used to develop a flood hazard map showing the flood extent and the depth of flooding for each pixel on a map. The flood hazard map is combined with an exposure map; in the case shown here, it is a land-use map, where each land-use class is assigned a level of economic damage. This shows the elements that would actually be exposed to the flood hazard. The vulnerability curves discussed previously are then used to assess how much of the economic value per land-use class would be damaged for different flood depths. By combining all of these data sets, the flood impacts for this particular hazard can be calculated (e.g., flood damage in euros, or fatalities).
If flood hazard maps are available for different return periods, it is possible to calculate the flood risk in terms of the expected annual impacts (e.g., expected annual damage in euros). This is achieved by integrating the area under a so-called exceedance probability-damage curve, as shown in Figure 7. As mentioned previously, flood risk is expected to change in the future due to changes in hazard, exposure, and vulnerability. For flood risk assessment and planning, therefore, it is also important to know how flood risk may change in the future. To model flood risk under future conditions, it is necessary to have projections of changes in the hazard, exposure, and vulnerability. Developing these projections is no easy task, however, as both climate change and socioeconomic developments involve large uncertainties (IPCC, 2012).
A possible way of dealing with these uncertainties is through the use of scenarios, which can be defined as sets of assumptions reflecting alternative (and preferably contrasting) developments that are coherent, internally consistent, and plausible (Kuik et al., 2008). Many scenarios are available, but in recent years, a new set of global scenarios has been developed to serve research needs across various disciplines in the climate change research community (van Vuuren et al., 2014). This has led to the development of Representative Concentration Pathways (RCPs) and Shared Socioeconomic Pathways (SSPs). RCPs represent changes in radiative forcing due to different levels of greenhouse gas concentrations in the atmosphere, and they can be used to explore future climate change. SSPs are pathways consisting of quantitative and qualitative indicators of how the future might unfold in terms of population growth, governance efficiency, inequality, socioeconomic developments, institutional factors, technology change, and environmental conditions (O’Neill et al., 2014; van Vuuren et al., 2014). By forcing global climate models with the scenarios of radiative forcing from the RCPs, climate modelers can model future meteorological conditions (e.g., rainfall and temperature) and sea level rise. These can then be used as input for hydrological and hydraulic models to simulate flood hazard under future climate conditions (Hinkel et al., 2014; Hirabayashi et al., 2013; Winsemius et al., 2015). Similarly, SSPs can be used as input for land-use and population models to simulate the change in exposure in the future, and these data can be used in the risk models. So far, it has proved very challenging to develop future projections of changes in vulnerability, which remains a key challenge for the flood risk community.
In addition to direct physical risk, floods have significant indirect effects on business and trade flows. This was demonstrated by the 2011 floods in Thailand, which disrupted global sales of electronics and cars (Petherick, 2011). On the other hand, floods in neighboring regions sometimes can even have temporary positive effects on the economy, as demand for construction work soars. There are increasingly reliable models available that can model the effect of direct physical losses on wider economic activity, providing a more robust picture of the total economic impact of floods (e.g., Koks, Bočkarjova, de Moel, & Aerts, 2015).
As noted in the first part of this article, flooding in itself is not necessarily a problem. It is the combination of hazard, exposure, and vulnerability that makes floods affect society, causing losses, including fatalities. The risk of flooding, therefore, is largely driven by human activity. On the flip side of that coin, society also has opportunities to substantially reduce the risk of flooding.
Adaptation to flooding has been practiced for centuries. In the Netherlands, a low-lying country with continuous threats from river and coastal flooding, dikes to protect communities from flooding have been built for over 2,000 years. Currently, the Netherlands has some of the lowest probability of flooding, in spite of having over a third of its land located below sea level. As the exposure to flooding increases globally, and with climate change increasing the frequency and intensity of flood events, the need to increase adaptation efforts in order to reduce their impact is greater than ever.
Effective flood risk management may follow a multilayer safety strategy (Figure 8). Under such a strategy, interventions are implemented that address the various underlying drivers of risk. Structural flood protection can be used to reduce hazard frequency; effective land-use planning can reduce exposure and vulnerability; and disaster risk management practices such as early warning systems and evacuation can improve communities’ preparedness and response to floods.
Structural flood protection (i.e., the construction of levees, dams, and dikes) remains the primary means of adaptation to floods. The construction and maintenance of levees and dikes require intensive engineering and significant investments. Therefore, it is often seen that the level of protection to floods is higher in high-income countries than low-income countries (Scussolini et al., 2016). When properly constructed and maintained, structural measures can protect against floods. However, when not maintained, levees are prone to failure and may give a false sense of security. This was demonstrated, for example, by the widespread levee breaches in New Orleans caused by Hurricane Katrina in 2005.
Nature-based measures, such as mangroves, wetlands, and forests, are increasingly used as alternatives or complements to structural protection (WAVES Partnership, 2016). For example, floodplain vegetation along rivers or coasts can be effective in reducing wave height and inundation, which then requires lower and less costly dikes to protect the hinterland (Figure 9). Nature-based solutions for flood risk management can be used not only in coastal or riverine settings, but also in urban areas. Green roofs, parks, and wetlands can help store storm waters and reduce urban flooding (Soz, Kryspin-Watson, & Stanton-Geddes, 2016).
In areas not protected by embankments, nonstructural measures may be effective at reducing the impact of flood events (Poussin, Bubeck, Aerts, & Ward, 2012). Such measures could include dry-proofing the lower level of the house using stilts, sandbags, or movable barriers; or wet-proofing the structure to remain largely undamaged from the flood (e.g., by using water-resistant building materials and floors). A more effective strategy than reducing damage to buildings in flood hazard areas can be to prevent these buildings from being constructed there in the first place. Spatial zoning can be an effective tool to gain control over land-use changes and development of new and existing urban areas (Burby, Deyle, Godschalk, & Olshansky, 2000). Spatial zoning may limit new development in flood-prone areas or may require such development to take into account the potential flood hazards in the construction.
When potentially damaging floods threaten to materialize and affect settlements or cities, early warning systems can provide signals and possibly enable evacuation or other risk management efforts. Increasingly, social media is used both to detect ongoing floods and to warn the population. Evidence suggests that in data-scarce areas, analysis of Twitter data can be used effectively alongside satellite information or traditional weather modeling to detect floods at an early stage (Jongman, Wagemaker, Romero, & de Perez, 2015).
Even if societies would significantly step up their efforts to reduce flood risk, investing in flood protection only makes sense so long as the benefits outweigh the costs (Kind, 2014). Economically effective flood protection standards will vary based on the potential damage and potential fatalities in a certain area. There will always remain a residual economic risk that cannot be prevented in an economically efficient way. For those risks, financial instruments need to be in place to make sure that individuals and businesses rapidly rebound from shocks. Uninsured losses are a growing concern, as a lack of financing for recovery, reconstruction, and relief may negatively affect peoples’ well-being and national budgets (Bouwer et al., 2007; Mechler, Hochrainer-Stigler, Aaheim, Salen, & Wreford, 2010). Disaster risk financing can be a combination of government-funded compensation schemes, private flood insurance, and international funding mechanisms such as the European Union’s Solidarity Fund. It is important that both changes in climatic extremes and increased exposure are incorporated in the design of such instruments in order to prevent increasing flood losses from depleting the financing (Jongman et al., 2014; Jongman, Koks, Husby, & Ward, 2014). Growing flood losses are already forcing insurance companies to increase their capital base and may reduce the industry’s profits substantially (Mills, 2005).
Adaptation strategies should be based on explicit cost-benefit analysis of risk reduction options, including future projections of climatic and socioeconomic changes (Fraser et al., 2016). Science helps bring analytical insights that may support the understanding of risk and decision-making on adaptation interventions. For example, Aerts, Botzen, Emanuel, Lin, de Moel, and Michel-Kerjan (2014) conducted a probabilistic analysis of storm surge scenarios for New York City and computed the costs and benefits of a range of structural and nonstructural measures, including physical protection and building codes. This specific analysis showed that storm surge barriers are not economically attractive under the current situation or a low-climate-change scenario, but they could become feasible under high-climate-change outcomes. Local-scale solutions such as elevating new buildings and protecting critical infrastructures proved to be the most cost effective. Such an analysis allows the prioritization of limited investment funds in protecting the city against rising sea levels.
The accurate assessment of climate-related flood risk alone is not enough to enable decision-makers to integrate the information into their decisions. The effective communication of the information and the presentation of key indicators in an understandable format are equally important.
Over the last several years, several advances have been made in the translation of scientific information into actionable formats. For example, the World Bank and Global Facility for Disaster Reduction and Recovery (GFDRR) are producing national-level disaster risk profiles for a range of developing countries. These profiles, available at https://www.gfdrr.org/disaster-risk-profiles, illustrate the results of probabilistic hazard and risk assessments and describe the main impacts on a national and subnational level (Figure 10). Such products can act as conversation starters with governments when defining strategies for risk reduction.
In addition, a wide range of online tools have been developed to assist decision-makers in utilizing flood risk assessments for practical application. The World Resources Institute’s Aqueduct Global Flood Analyzer (http://floods.wri.org) is an example of such a platform. This tool visualizes hazard and risk data and enables the user to change protection levels in the country or basin of interest dynamically in order to investigate the potential effect of structural protection on flood risk. This platform is currently being extended to include cost-benefit analysis functionality.
The Aqueduct Global Flood Analyzer is just one of many tools that enable the visualization of spatial flood hazard and risk information, cost-benefit analysis, and mitigation options. GeoNodes have been set up to display and share flood risk information in various countries, such as Malawi (http://www.masdap.mw/), Sri Lanka (http://riskinfo.lk/), and Afghanistan (http://disasterrisk.af). These platforms are widely used by governments and development partners to inform adaptation strategies. Furthermore, a number of specific tools are available to investigate the feasibility of green infrastructure solutions for reducing flood risk. The Coastal Resilience web platform (http://coastalresilience.org), for example, highlights toolkits for the assessment of natural solutions, mainly focusing on the United States. These include mapping tools and economic assessment, and they can support and inform nature-based adaptations.
This article presented the current state of science related to the understanding and assessment of, and adaptation to, climate-related river and coastal flood risk. To pursue adaptation to flooding, it is essential to understand the driving forces behind risk (hazard, vulnerability, and exposure) and how they can be addressed.
On the hazard side, significant scientific advances have been made over the years in improving flood hazard modeling. Flood inundation models have become increasingly able to realistically simulate historical and possible flood events via better inundation routines using high-resolution elevation data. Models are developed that can consistently simulate current flood hazards for the entire world, as well as possible changes in hazards under all possible climate model outputs (Winsemius et al., 2015).
However, challenges remain. Freely available digital elevation models (DEMs), especially in developing countries, remain limited and often have a high degree of vertical and horizontal uncertainty (Schumann, Bates, Neal, & Andreadis, 2014). Furthermore, the reliable integration of flood protection measures in large-scale flood hazard models is a challenge; information on the presence of protection measures is often missing. Recently, the first global database on flood protection standards was presented (Scussolini et al., 2016), although this is still missing reliable data in many areas.
Regarding exposure, recent advances have shown that today, we can dynamically assess the influence of changing populations and urbanization on risk and can use that data to model the effects of various land planning policies on a national scale (e.g., Muis, Güneralp, Jongman, Aerts, & Ward, 2015). This way, we can assess the current contributions to population growth and urban expansion on flood risk and produce possible future outlooks.
However, challenges remain in this field as well. First, there are enormous uncertainties around the valuation of assets and potential damages for various types of land use—from urban to agriculture. Second, there is a growing understanding that actual exposure goes beyond economic values. Whereas the economic value of a poor household is low, the potential impact on their lives is much greater than for nonpoor households (Hallegatte, Vogt-Schilb, Bangalore, & Rozenberg, 2016). Assessing and addressing the nuances of exposure, therefore, are highly complicated.
Finally, we have shown advances in the assessment of vulnerability. We are increasingly able to specify the levels of physical vulnerability of assets, social vulnerability of various societal groups, and changes in the overall global level of vulnerability across different countries. However, in flood damage modeling, vulnerability remains the most uncertain element in the risk chain (de Moel & Aerts, 2011). Whereas damage modeling has improved over the years, the uncertainty about depth-damage functions and their effects on final risk estimates remains high. More research is needed to develop innovative approaches for damage modeling (e.g., Merz, Kreibich, & Lall, 2013). Especially in developing countries, the validation and calibration of vulnerability are major challenges that have barely been addressed.
A range of options are available to policymakers to address flood risk and build a safer future (Fraser et al., 2016). As populations grow, cities expand, and climate changes, the challenge is to reduce current flood risk while also ensuring that new development happens in a resilient way. It is crucial to consider a diversified risk management approach that considers engineering measures (e.g., levees), green infrastructure solutions (e.g., wetlands and mangroves), and risk management approaches such as early warning systems and evacuation procedures. With climate change leading to potentially more severe and more frequent floods in many parts of the world, comprehensive understanding of and effective actions against flood risk will be increasingly critical.
Adger, W. N., Dessai, S., Goulden, M., . . ., Wreford, A. (2009). Are there social limits to adaptation to climate change Climate Change, 93, 335–354.Find this resource:
Aerts, J. C. J. H., Botzen, W. J. W., Emanuel, K., Lin, N., & de Moel, H., & Michel-Kerjan, E. O. (2014). Evaluating flood resilience strategies for coastal megacities. Science, 344(6183), 473–475.Find this resource:
Angel, S., Parent, J., Civco, A., & Blei, D. (2013). Atlas of urban expansion. Available www.lincolninst.edu/subcenters/atlas-urban-expansion.
Bates, P. D., Dawson, R. J., Hall, J. W., Horritt, M. S., Nicholls, R. J., & Wicks, J., & Hassan, M. A. A. M. H. (2005). Simplified two-dimensional numerical modelling of coastal flooding and example applications. Coastal Engineering, 52, 793–810.Find this resource:
Bouwer, L. M., Crompton, R. P., Faust, E., Hoppe, P., & Pielke, R. A. Jr. (2007). Confronting disaster losses. Science, 318(5851), 753.Find this resource:
Burby, R. J., Deyle, R. E., Godschalk, D. R., Olshansky, R. B. (2000). Creating hazard-resilient communities through land-use planning. Natural Hazards Review, 1, 99–106.Find this resource:
Cid, A., Castanedo, S., Abascal, A.J., Menéndez, M., & Medina, R. (2014). A high-resolution hindcast of the meteorological sea level component for southern Europe: The GOS dataset. Climate Dynamics, 43, 2167–2184.Find this resource:
Center for International Earth Science Information Network (CIESIN). (2016). Gridded population of the world, version 4 (GPWv4): Population density. Palisades, NY: CIESIN.Find this resource:
Ciullo, A., Viglione, A., Castellarin, A., Crisci, M., & Di Baldassarre, G. (2017). Socio-hydrological modelling of flood-risk dynamics: comparing the resilience of green and technological systems. Hydrological Sciences Journal, 62, 880–891.Find this resource:
Coles, S. (2001). An introduction to statistical modeling of extreme values. Springer Series in Statistics. London: Springer London.Find this resource:
Cutter, S. L., Boruff, B. J., & Shirley, W. L. (2003). Social vulnerability to environmental hazards. Social Science Quarterly, 84, 242–261.Find this resource:
de Moel, H., & Aerts, J. C. J. H. (2011). Effect of uncertainty in land use, damage models and inundation depth on flood damage estimates. Natural Hazards, 58, 407–425.Find this resource:
de Moel, H., Jongman, B., Kreibich, H., Merz, B., Penning-Rowsell, E., & Ward, P. J. (2015). Flood risk assessments at different spatial scales. Mitigation and Adaptation Strategies for Global Change, 20(6), 865–890.Find this resource:
Di Baldassarre, G., Kooy, M., Kemerink, J. S., & Brandimarte, L. (2013). Towards understanding the dynamic behaviour of floodplains as human-water systems. Hydrology and Earth Systems Sciences Discussions, 10, 3869–3895.Find this resource:
Di Baldassarre, G., Viglione, A., Carr, G., Kuil, L., Yan, K., Brandimarte, L., & Blöschl, G. (2015). Debates—perspectives on socio-hydrology: Capturing feedbacks between physical and social processes. Water Resources Research, 51, 4770–4781.Find this resource:
Dietrich, J. C., Bunya, S., Westerink, J. J., . . ., Roberts, H. J. (2010). A high-resolution coupled riverine flow, tide, wind, wind wave, and storm surge model for Southern Louisiana and Mississippi. Part II: Synoptic description and analysis of Hurricanes Katrina and Rita. Monthly Weather Review, 138, 378–404.Find this resource:
Dow, K., Berkhout, F., Preston, B. L., Klein, R. J. T., Midgley, G., & Shaw, M. R. (2013). Limits to adaptation. Nature Climate Change, 3, 305–307.Find this resource:
Fraser, S., Jongman, B., Balog, S., Simpson, A., Saito, K., & Himmelfarb, A. (2016). Making a riskier future: How our decisions are shaping future disaster risk. Washington, DC: GFDRR.Find this resource:
Gerland, P., Raftery, A. E., Sevčíková, H., . . ., Wilmoth, J. (2014). World population stabilization unlikely this century. Science, 346, 234–237.Find this resource:
Haigh, I. D., Wijeratne, E. M. S., MacPherson, L. R., Pattiaratchi, C. B., Mason, M. S., Crompton, R. P., & George, S. (2014). Estimating present-day extreme water level exceedance probabilities around the coastline of Australia: tides, extra-tropical storm surges and mean sea level. Climate Dynamics, 42, 121–138.Find this resource:
Hallegatte, S., Bangalore, M., Bonzanigo, L., . . ., Vogt-Schilb, A. (2015). Shock waves: Managing the impacts of climate change on poverty. Washington, DC: World Bank.Find this resource:
Hallegatte, S., Green, C. H., Nicholls, R. J., & Corfee-Morlot, J. (2013). Future flood losses in major coastal cities. Nature Climate Change, 3, 802–806.Find this resource:
Hallegatte, S., Vogt-Schilb, A., Bangalore, M., & Rozenberg, J. (2016). Unbreakable: building the resilience of the poor in the face of natural disasters. Washington, DC: The World Bank.Find this resource:
Hinkel, J., Lincke, D., Vafeidis, A. T., . . ., Levermann, A. (2014). Coastal flood damage and adaptation costs under 21st-century sea-level rise. Proceedings of the National Academy of Sciences of the USA, 111, 3292–3297.Find this resource:
Hirabayashi, Y., Mahendran, R., Koirala, S., . . ., Kanae, S. (2013). Global flood risk under climate change. Nature Climate Change, 3, 1–6.Find this resource:
Intergovernmental Panel on Climate Change (IPCC). (2012). Managing the risks of extreme events and disasters to advance climate change adaptation. A special report of Working Groups I and II of the Intergovernmental Panel on Climate Change, New York: Cambridge University Press.Find this resource:
Jongman, B. (2014). Unravelling the drivers of flood risk across spatial scales (PhD Thesis). Amsterdam: Vrije University.Find this resource:
Jongman, B., Hochrainer-Stigler, S., Feyen, L., . . ., Ward, P. J. (2014). Increasing stress on disaster-risk finance due to large floods. Nature Climate Change, 4, 264–268.Find this resource:
Jongman, B., Koks, E. E., Husby, T. G., & Ward, P. J. (2014). Increasing flood exposure in the Netherlands: implications for risk financing. Natural Hazards Earth System Sciences, 14, 1245–1255.Find this resource:
Jongman, B., Wagemaker, J., Romero, B., & de Perez, E. (2015). Early flood detection for rapid humanitarian response: Harnessing near real-time satellite and Twitter signals. ISPRS International Journal of Geo-Information, 4, 2246–2266.Find this resource:
Jongman, B., Ward, P. J., & Aerts, J. C. J. H. (2012). Global exposure to river and coastal flooding: Long-term trends and changes. Global Environmental Change, 22, 823–835.Find this resource:
Jongman, B., Winsemius, H. C., Aerts, J. C. J. H., . . ., Ward, P. J. (2015). Declining vulnerability to river floods and the global benefits of adaptation. Proceedings of the National Academy of Sciences of the USA, 112(8), E2271–E2280.Find this resource:
Kew, S. F., Selten, F. M., Lenderink, G., & Hazeleger, W. (2013). The simultaneous occurrence of surge and discharge extremes for the Rhine delta. Natural Hazards Earth System Sciences, 13, 2017–2029.Find this resource:
Kind, J. M. (2014). Economically efficient flood protection standards for the Netherlands. Journal of Flood Risk Management, 7, 103–117.Find this resource:
Koks, E. E., Bočkarjova, M., de Moel, H., & Aerts, J. C. J. H. (2015). Integrated direct and indirect flood risk modeling: Development and sensitivity analysis. Risk Analysis, 35, 882–900.Find this resource:
Koks, E. E., Jongman, B., Husby, T. G., & Botzen, W. J. W. (2014). Combining hazard, exposure, and social vulnerability to provide lessons for flood risk management. Environmental Science & Policy, 47, 42–52.Find this resource:
Kuik, O., Buchner, B., Catenacci, M., Goria, A., Karakaya, E., & Tol, R. (2008). Methodological aspects of recent climate change damage cost studies. Integrated Assessment, 8(1).Find this resource:
Lenderink, G., Mok, H. Y., Lee, T. C., & van Oldenborgh, G. J. (2011). Scaling and trends of hourly precipitation extremes in two different climate zones—Hong Kong and the Netherlands. Hydrology and Earth System Sciences, 15, 3033–3041.Find this resource:
Lenderink, G., & van Meijgaard, E. (2008). Increase in hourly precipitation extremes beyond expectations from temperature changes. Nature Geoscience, 1, 511–514.Find this resource:
Linard, C., Kabaria, C. W., Gilbert, M., . . ., Snow, R. W. (2017). Modelling changing population distributions: An example of the Kenyan Coast, 1979–2009. International Journal of Digital Earth, 10(1), 1–13.Find this resource:
Mechler, R., & Bouwer, L. M. (2014). Understanding trends and projections of disaster losses and climate change: is vulnerability the missing link? Climatic Change, 133(1), 23–35.Find this resource:
Mechler, R., Hochrainer-Stigler, S., Aaheim, A., Salen, H., & Wreford, A. (2010). Modelling economic impacts and adaptation to extreme events: Insights from European case studies. Mitigation and Adaptation Strategies for Global Change, 15, 737–762.Find this resource:
Merz, B., Kreibich, H., & Lall, U. (2013). Multi-variate flood damage assessment: a tree-based data-mining approach. Natural Hazards Earth System Sciences, 13, 53–64.Find this resource:
Mills, E. (2005). Insurance in a climate of change. Science, 309, 1040–1044.Find this resource:
Muis, S., Güneralp, B., Jongman, B., Aerts, J. C. J. H., & Ward, P. J. (2015). Flood risk and adaptation strategies under climate change and urban expansion: A probabilistic analysis using global data. Science of the Total Environment, 538, 445–457.Find this resource:
Muis, S., Verlaan, M., Winsemius, H. C., Aerts, J. C. J. H., & Ward, P. J. (2016). A global reanalysis of storm surges and extreme sea levels. Nature Communications, 7.Find this resource:
Munich Re. (2016). NatCatSERVICE Database. Munich Reinsurance Company Geo Risks Research, Munich. Retrieved from https://www.munichre.com/touch/naturalhazards/en/about/index.html.Find this resource:
O’Neill, B. C., Kriegler, E., Riahi, K., . . ., van Vuuren, D. P. (2014). A new scenario framework for climate change research: The concept of shared socioeconomic pathways. Climatic Change, 122, 387–400.Find this resource:
Petherick, A. (2011). The second tsunami. Nature Climate Change, 1, 90–91.Find this resource:
Poussin, J. K., Bubeck, P., Aerts, J. C. J. H., & Ward, P. J. (2012). Potential of semi-structural and non-structural adaptation strategies to reduce future flood risk: Case study for the Meuse. Natural Hazards Earth System Sciences, 12, 3455–3471.Find this resource:
Reice, S. R. (2003). The silver lining: The benefits of natural disasters. Princeton, NJ: Princeton University Press.Find this resource:
Resio, D. T., & Westerink, J. J. (2008). Modeling the physics of storm surges. Physics Today, 61(9).Find this resource:
Rijkswaterstaat. (2011). Syntheserapport Gebiedspilots Meerlaagseveiligheid. Retrieved from https://www.rijksoverheid.nl/documenten/rapporten/2011/11/29/syntheserapport-gebiedspilots-meerlaagsveiligheid.
Sampson, C. C., Smith, A. M., Bates, P. D., Neal, J. C., Alfieri, L., & Freer, J. E. (2015). A high-resolution global flood hazard model. Water Resources Research, 51, 7358–7381.Find this resource:
Schumann, G. J.-P., Bates, P. D., Neal, J. C., & Andreadis, K. M. (2014). Technology: Fight floods on a global scale. Nature, 507, 169.Find this resource:
Scussolini, P., Aerts, J. C. J. H., Jongman, B., Bouwer, L. M., Winsemius, H. C., de Moel, H., & Ward, P. J. (2016). FLOPROS: An evolving global database of flood protection standards. Natural Hazards Earth System Sciences, 16, 1049–1061.Find this resource:
Smith, D. I. (1994). Flood damage estimation—a review of urban stage-damage curves and loss functions. Water SA, 20, 231–238.Find this resource:
Soz, S. A., Kryspin-Watson, J., & Stanton-Geddes, Z. (2016). The role of green infrastructure solutions in urban flood risk management. Washington, DC: The World Bank. Retrieved from https://openknowledge.worldbank.org/handle/10986/25112 License: CC BY 3.0 IGO.Find this resource:
Trapp, R. J., Diffenbaugh, N. S., Brooks, H. E., Baldwin, M. E., Robinson, E. D., & Pal, J. S. (2007). Changes in severe thunderstorm environment frequency during the 21st century caused by anthropogenically enhanced global radiative forcing. Proceedings of the National Academy of Sciences of the USA, 104, 19719–19723.Find this resource:
United Nations International Strategy for Disaster Reduction (UNISDR). (2011). Global assessment report on disaster risk reduction 2011. Revealing risk, redefining development. Geneva, Switzerland: UNISDR.Find this resource:
United Nations International Strategy for Disaster Reduction (UNISDR). (2015). Global assessment report 2015. Making development sustainable: the future of disaster risk management. Geneva, Switzerland: UNISDR.Find this resource:
van den Honert, R. C., & McAneney, J. (2011). The 2011 Brisbane floods: Causes, impacts, and implications. Water, 3, 1149–1173.Find this resource:
van Vuuren, D. P., Kriegler, E., O’Neill, B. C., . . ., Winkler, H. (2014). A new scenario framework for climate change research: Scenario matrix architecture. Climatic Change, 122, 373–386.Find this resource:
Vousdoukas, M. I., Voukouvalas, E., Annunziato, A., Giardino, A., & Feyen, L. (2016). Projections of extreme storm surge levels along Europe. Climate Dynamics, 47, 3171–3190.Find this resource:
Wahl, T., Jain, S., Bender, J., Meyers, S. D., & Luther, M. E. (2015). Increasing risk of compound flooding from storm surge and rainfall for major US cities. Nature Climate Change, 5, 1093–1097.Find this resource:
WAVES Partnership. (2016). Managing coasts with natural solutions : Guidelines for measuring and valuing the coastal protection services of mangroves and coral reefs. Washington, DC: The World Bank.Find this resource:
Winsemius, H. C., Aerts, J. C. J. H., van Beek, L. P. H., . . ., Ward, P. J. (2015). Global drivers of future river flood risk. Nature Climate Change, 6, 381–385.Find this resource:
Zhang, Y., Hong, Y., Wang, X., Gourley, J. J., Gao, J., Vergara, H. J., & Yong, B. (2013). Assimilation of passive microwave streamflow signals for improving flood forecasting: A first study in Cubango River Basin, Africa. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 6, 2375–2390.Find this resource: