Generic adaptation pathways for coastal archetypes under uncertain sea-level rise

Adaptation to coastal flood risk is hampered by high uncertainty in the rate and magnitude of sea-level rise. Subsequently, adaptation decisions carry strong risks of under- or over-investment, and could lead to costly retrofitting or unnecessary high margins. To better allocate resources timely and effectively, and achieve long-term sustainability, planners could utilise adaptation pathways, revealing the path-dependencies of adaptation options. This helps to identify low-regret short-term decisions that preserve options in an uncertain future, while monitoring to detect signals to adapt. A major barrier to the application of adaptation pathways is limited experience. To facilitate this, here we generalize this pathways approach for six common coastal archetypes, resulting in generic pathways suitable to be adjusted to local conditions. This provides a much richer analysis of coastal adaptation than provided by any previous analysis, by assessing the solution space and options over time for a variety of coastal regions. Based on this analysis, we find that the number of adaptation options declines while sea-level rises. For some archetypes, it becomes clear that long-term thinking is needed now, about if, how and when to move to transformative options, such as planned retreat, which may presently not be considered or acceptable. Our analysis further shows that coastal adaptation needs to start earlier than anticipated, especially given time required for local debate and choice and to implement measures.


Introduction
Uncertainty about the future complicates and can even paralyze decision making on adaptation. One such large uncertainty is quantifying the rate and magnitude of sea-level rise [1][2][3]. Along with uncertain changes in future population, economic developments and societal values, this results in deep uncertainties. Depending on climate change mitigation, by 2100, mean sea-level may further rise by 0.26 to 0.98 m 4 , with a low probability, higher tail of possible rise due to accelerated ice sheet melting [3,4]. Even in case emissions are reduced as defined in the Paris Agreement, sea-levels will continue to rise, although more slowly [5,6].
Adapting to sea-level rise typically entails large-scale investments with long planning and implementation time, and potentially large societal impacts for current and future generations. In the face of deep uncertainty, a 'wait and see' approach to adaptation is often taken, until uncertainty is reduced [7]. However, this could result in untimely adaptation, which may be less effective, and could limit future adaptation options [8].
To support decision making under uncertainty, an adaptation pathways approach was devised [8,9]. Adaptation pathways are sequences of linked (portfolios of) actions that can be implemented as conditions change. Typically, they start with low-regret actions that maintain future options [10] when uncertainty is high. As time progresses and ambient conditions change, this initial low-regret adaptation action may reach a threshold when it no longer performs acceptably, i.e., an adaptation tipping point occurs [11]. Subsequently a switch to another action is needed to continue to achieve objectives; a pathway of adaptation decision emerges. Identifying thresholds is important for optimal adaptation, therefore monitoring to detect early signs of change is required [12].
Adaptation pathways support decision making under uncertainty in three main ways. First, they can help overcome the policy paralysis due to uncertainty, by putting adaptation decisions into manageable steps over time, starting with low-regret actions. Second, the visualization of alternative pathways and their costs and benefits makes the path-dependency of options explicit [8], showing that past decisions open some options and could foreclose others [13]. This helps to recognize the risk of lock-in situations, minimize costly retrofitting and achieve long-term sustainability. Third, adaptation pathways deal explicitly with timing and thereby help to define not only what decisions but also when decisions are needed for adaptation.
So far, adoption of adaptation pathways to sea-level rise includes the UK Thames Estuary 2100 plan [9], the Dutch Delta Program in the Rhine-Meuse delta [14], the Bangladesh Delta Plan, the township of Lakes Entrance in Australia [10] and the Hutt river in New Zealand [15]. In spite of their proven potential to support decision making under uncertainty, application of adaptation pathways remains uncommon [15,16]. One reason for this may be the challenge of the complexity of exploring and evaluating the wide range alternative pathways into the medium and long-term future, rather than the short-term where coastal management decisions are often focused.
Hence, our goal is to create generalized adaptation pathways applicable to a wide range of environments (referred to as archetypes) and common adaptation methods. We do not consider governance or socioeconomic conditions as these can be very local in nature. Thus, our motivation is to provide a broad framework and method to construct pathways, thus enabling coastal managers to develop their pathways specific to their coastline and management goals. This advances the science by moving forward from the generic traditional 'protect-accommodate-retreat' options and considering how in reality these options can be achieved over time, while extending planning timescales, and considering path-dependency and uncertainty.
We create this pathways framework by (i) defining common archetypes through geomorphic setting and land use (section 2.1); (ii) describing the four-step method of creating pathways through identifying the hazards, management goals, adaptation types and their thresholds (section 2.2); instigating steps 1-3 of the methodology (in section 3); deriving and analyzing pathways for step 4 of the methodology (in section 4); and (iv) exploring how these generic pathways may be made site specific and relevant to coastal managers (in section 5).

Method
To derive a typology of generic adaptation pathways for coastal adaptation to sea-level rise, we created a set of common coastal archetypes for which generic adaptation pathways can be developed, and then designed and illustrated the potential adaptation pathways.

Derivation of coastal archetypes
Physically, adaptation options principally depend on geomorphology and land use. Using existing classifications for geomorphology [17][18][19][20][21] and land use [22], we divide these into three sub-categorizations for geomorphology and two for land use. Our three low-lying coastal geomorphic settings are: • Open: a coast with sediment, without river mouths; • Delta: a deltaic coast with wetlands; • Estuary: an estuarine coast with wetlands.
Cliffed environments are not considered as they are not low-lying or significantly threatened by sea-level rise. Small island settings are also excluded as these may contain the geomorphic features above or, depending on size, require a different approach to adaptation at island level.
The two land use types considered are: • Urban: A densely populated coast, with substantial and/or costly building stock, and/or tourist attractions, where sea-level rise would result in significant damage and disruption. Adaptation would typically have a high benefit-to-costs ratio.
• Rural: A predominantly agricultural coast, typically of lower value than urban areas, with sparser dwellings, low population density and limited tourism. Sea-level rise could result in disruption of local livelihoods (but without regional or national implications), but not in significant infrastructure damage. Adaptation would typically have lower benefit-to-cost ratio than in urban areas.
These geomorphology and land use types were combined to form six coastal archetypes (figure 1). Archetypes describing purely natural coastlines were not considered for the pathways analysis, as adaptation pathways are much less likely to be necessary.

Derivation of adaptation pathways to sea-level rise
To design adaptation pathways for the coastal archetypes, we follow the steps described in the Dynamic Adaptive Policy Pathways approach [8]. First, we specify the management aims and analyse the impacts of sea-level rise for the different archetypes. Second, adaptation options are identified to address the aims and impacts. Each adaptation option is assessed against its effectiveness to reduce the following impacts that are most relevant for coastal systems (e.g., [23][24][25]; see also supplementary material is available online at stacks.iop.org/ERC/1/ 071006/mmedia): • submergence (the permanent covering of water over the land), • temporary flooding from extreme events (the temporary covering of the land or a wetland), • erosion (the permanent destruction of land due to attack from sea water), • rising groundwater levels (the raising of the water table and impeded drainage)and • salinization (an increase in the salt content of the soil, ground water or inland water bodies).
Thirdly, we define threshold conditions that make a specific measure viable (an opportunity tipping point), and threshold conditions for which the measure fails, making additional or other actions necessary (an adaptation tipping point). We thus considered reasons to adapt, rather than limits (e.g. [26]) or barriers (e.g. [27]) for adaptation. These reasons to adapt are defined as: • Engineering design conditions: when design conditions are exceeded and measure effectiveness decline; • Space and material availability: where there is insufficient space to build a defence or to allow for retreat, or where there are insufficient raw materials available; • Cost-benefit conditions: when costs exceed benefits; • Social (un)acceptability: when a lack of government or stakeholder support for adaptation inhibits action or generate strong opposition or social conflict with (part of the) population or stakeholders, or when support generates opportunities to implement a measure; • Economic productivity: where the economic production or service level has insufficient yield or quality to be viable (e.g. food production).
Fourthly, pathways are designed by structurally sequencing adaptation options while considering (a) the relative amount of sea-level rise they are able to address as indicated through the tipping point conditions and (b) the path-dependency of options. In addition, narratives were written describing sequences of adaptation options as sea-levels rise. The pathways are then visualized in a pathways map for each archetype and illustrated with pathways found in literature.

Coastal archetypes and adaptation options
The suitability of adaptation options and pathways depends on the six broad archetypes representing the combinations of dominant geomorphology (open coast, delta, and estuary) and land use (urban and rural). Figure 1 illustrates the archetypes, the direct and indirect impacts of sea-level rise they already experience or could experience in the future (see also Supplementary Material), and examples of real-world occurrence.
For each archetype, sea-level rise has typical physical and socio-economic impacts, depending on geomorphology and land-use respectively. For example, in terms of our archetypes, in urban areas, sea-level rise may result in erosion of open coasts with beaches (archetype 1a) and thereby a decrease in the beach recreational carrying capacity which may have economic (coastal tourism) and/or social (leisure) consequences. Conversely, along rural open coasts (archetype 1b), the loss of natural values supported by the beach may be more prominent and can be quantified in terms of affected ecosystem services. Consequently, these archetypes require separately analysis, reflecting different management aims and thus adaptation goals, measures and pathways.
Following the methodology described in section 2.2, step 1 aims to describe the management aim. This typically depends on land use: • In urban areas, the management aim is to reduce coastal flood, erosion and local water levels, i.e., to protect livelihoods and promote industry and tourism and reduce expected damages in coastal infrastructures.
• In rural areas, the management aim is to safeguard food production from temporary flooding, erosion, salinization and rising ground waters, and to defend smaller, local communities and industries from temporary flooding and erosion. It does not necessarily aim to address permanent inundation. In areas of high natural values, adaptation aims at ensuring accommodation space for ecosystem facing accelerated erosion (e.g., wetland migration).
Step 2 of the method (section 2.2) aims to identify adaptation options to address impacts of sea-level rise. A list of thirteen common and proven adaptation options was compiled (table 1; Supplementary Material) and divided into three categories following the IPCC [28]: protect, accommodate and retreat. We have deliberately not included 'attack' or 'advance' as a fourth option which is at times used as a method of defence or due to land claim, as it is often site specific and a special case of protect. Attack may have similar tipping points but at different relative timings to protect.
To be consistent with the archetypes, the adaptation is considered in more generic functional terms. For instance, breakwaters and wetlands are wave dissipation structures. Also, early warning systems are appropriate across all coastal archetypes, so they are omitted.
The third step of the method (section 2.2) is to define adaptation thresholds and tipping points. Most adaptation options address several impacts of sea-level rise, and have several reasons for opportunity and adaptation tipping points (table 1). The thresholds of these tipping points were assessed in terms of a relative to sea-level rise: low (e.g. less than 0.3 m), medium (e.g. 0.3-0.8 m) and high (e.g. more than 0.8 m). The boxes of figure 2 thereby present the relative amount of sea-level rise the adaptation options can address before management aims may start to fail. The exact values for sea-level rise are location specific. Note that this is deliberately independent of the time dimension, so as to allow analysis without assuming specific climate scenarios (or associated socio-economic conditions which typically consider up to 2100), for which rates of sealevel rise vary [29,30].
Reasons for opportunity tipping points vary. For example, constructing a storm surge barrier normally takes decades for planning and implementation (e.g. [9]). The use of nature-based options such as planting mangroves or wetlands requires not only time to grow and stabilize to become effective, but also space and sufficient sediment supply [31]. Additionally, adaptation tipping points may be determined when, for example, sea-levels become too high for an engineered defence (such as a storm-surge barrier), prompting the need for new adaptation, such as a fix barrage and pumps. Table 1. Possible adaptation options, the impacts they address and their opportunity and adaptation tipping points across the six coastal archetypes studied. Impacts: P=permanent flooding (submergence); T=temporal flooding due to extreme event; E=erosion, G=rising groundwater levels; S=salt water intrusion. Reasons for opportunity and adaptation tipping points: D=engineering design; $=cost-benefit considerations; M=space and material availability; A=social acceptability; Y=economic productivity. X indicate that the adaptation option is more or less common for a given archetype, respectively. o indicates a less common adaptation for an archetype. Uncertainty in the sea-level rise conditions or timing of a tipping point is indicated with a dotted line.

Generic pathways for coastal archetypes
Step 4 of the methodology described in section 2.2 involves pathway design. These pathways are described in sections 4.1-4.4 for each archetype. The adaptation pathways for each archetype are shown in figure 2. In many cases, we could not find an adaptation pathway for a coastal locality as few exist. Hence, we also illustrate possibilities from the past or other (non-pathways) plans.

Urban open coast
For an 'urban open coast' archetype potential impacts of sea-level rise include erosion, temporal flooding from extreme events, submergence and rising groundwater. Adaptation options thus aim to protect from flooding and erosion and to maintain the coast for recreation and tourism. Today, the most common adaptation falls under the 'protect' category (see table 1 and table SM1 in Supplementary Material), but accommodation through flood proofing and, planned retreat by enforcing no-build zones are becoming more widely considered.
A common pathway for this archetype, when erosion is the main impact, starts with beach nourishment to maintain the coastline and protect the area from flooding. Nourishment volumes increase or become more frequent as sea-level rise accelerates, as expected on the Dutch coast [14]. For high-end sea-level rise, beaches may need to be almost continuously nourished, which may be unacceptable for inhabitants, tourists and nature, and thus reach an adaptation tipping point for social reasons. This could be avoided by adopting a meganourishment based-strategy as in the Dutch 'sand engine' approach [32]. However, there may be a threshold as a wide beach in front of an urban coast may not be accepted. Ultimately, a solution here must recognize the tradeoff between the higher costs associated with continuous nourishments, the stronger modification of the shoreline caused by mega-nourishment [33], and the social acceptability of an option. Other reasons for adaptation tipping points for nourishments are lack of cost-effective resources (i.e. sand [34]) and high energy costs [35]. These tipping points may lead to combining nourishment with controlled retreat measures such as planned no-build zones or managed realignment in selected locations. Such a pathway was devised northern Portugal (Aveiro), where costs, effects on the ecosystem and the availability of sand determine adaptation tipping points and the switch from nourishment to planned realignment in combination with flood proofing of infrastructure [34].
A pathway addressing flooding as the main impact will consist of first using protection measures, such as wave dissipation structures or flood gates in high-risk areas to mitigate storm-induced floods under low sea-level rise, and then moving to dikes or seawalls as flood frequency becomes unacceptable. Simultaneously, adaptation could also start with planned no-build zones/set-back line (e.g. as was proposed in Cape Town [36]) and flood proofing new infrastructure and buildings (e.g., elevating houses on piles, as common in the US and Asia). This could be combined with protection for existing buildings (e.g., south east Queensland) [37], as elevation of existing parts of the city could be more expensive and socially unacceptable or not technically possible. With higher sea-levels, planned realignment and relocation are possible, although the lack of space for realigning may present a tipping point. Such pathways that start with accommodate through changes in land use and building regulations, and later switch to either protection with barriers, or planned retreat have been mapped for Lake Entrance in Australia [10,38].

Urban deltas
Historically, many 'urban deltas' were drained and pumped to remove excess water and lower groundwater levels. Subsequently dikes were built to protect against flooding. Human interventions extend beyond the deltaic coastal zone, such as upstream damming (Mississippi delta, US), drainage (Rhine-Meuse delta, Netherlands), groundwater abstraction (Mekong delta, Vietnam), which may cause subsidence [39] and thus a larger relative rise of sea-level. Consequently, many deltas are already following a specific pathway, and are locked into limited future options.
Continuing on the pathway of protection through dikes in combination with drainage and pumping is a common pathway in urbanized deltas (e.g., deltaic part of the Netherlands [14]; Jakarta, Indonesia [40]). Nevertheless, nature-based defences to reduce waves are increasingly considered [41] to reduce flood risk, and could thus shift the pathway.
A simultaneous or complementary pathway for no to low levels of sea-level rise could start with accommodation, including flood proofing or elevating infrastructure for low levels of sea-level rise, allowing for occasional flooding. For example, in the Mekong Delta, 'accommodate' options, such as floodproofing and raising property, could postpone dike construction [42]. Additionally, accommodate measures could be combined with breakwaters to ensure reduced flood risk and/or to extend the threshold of adaptation so that an adaptation point occurs later in time.
Hard defences such as dikes could occur with any level of sea-level rise, but would be increasingly necessary with low to medium levels of sea-level rise, as accommodation options reach tipping points which limit their efficiency. As flood barriers long enough to protect deltas are expensive [42], they are not considered an option for this archetype. In practice, they are limited to parts of the delta that resemble the estuary archetype, where they aim to protect areas of particularly high exposure (e.g. Ho Chi Min City [43]) to be cost effective. Closed barriers or storm surge barriers that frequently need to close can have adverse impact on port functioning, which is a future concern for the port of Rotterdam in the Rhine-Meuse delta [11].
As floodgates, floodproofing and wave dissipation structures reach their tipping point, local land raising becomes an increasing possibility, and could be undertaken as urban areas are renewed. This renewal acts as a threshold for an opportunity tipping point. Conversely, adaptation tipping points will mainly be determined by cost-benefit conditions, space and material availability (e.g. sand) and social unacceptability of dislocation and loss of cultural value in the relinquished districts [11]. Planned retreat would be either a last resort (and could be used simultaneously with land raising), used in risk sharing across a wider area or through set-back lines to gradually relocate infrastructure to higher ground.

Urban estuaries
In 'urban estuaries', such as Elbe/Hamburg, Thames/London and Hudson/New York, fluvial and coastal flooding may coincide. The management aim is to protect the city, industry and port from inundation or temporary flooding, and to a lesser extent from extreme events and rising ground waters. Thus protection and accommodate are more common adaptation types over retreat.
One pathway may involve flood retention areas for low levels of sea-level rise, thus reducing river discharge (e.g. Netherlands). With increasing sea-levels, quay walls will have to be raised (e.g., Tai O, Hong Kong [44]), and large protection infrastructure may be required, such as a storm surge barrier.
A storm surge barrier already exists in the Thames Estuary. To continue to protect London, the low-regret option identified was to raise existing defences, enabling the possibility of raising them further in the future, in addition to incorporating structural flexibility and reconsidering safety margins. Only with much higher sealevel will a new downstream barrier be built [9]. For rural areas of the estuary, planned retreat is considered, but this is limited due to lack of space [45].
Alternatively, a pathway set on the 'protect' trajectory, could start with no-build zones, floodproofing of infrastructure, or floodgates. With higher sea-water levels dikes and storm surge barriers are needed if retreat is not preferred. This can be illustrated with the plans for some localities around New York City. Post Hurricane Sandy in 2012 an overall policy of 'no-retreat' was defined [46]. Alternative pathways include protection through floodwalls and reclaimed natural barriers (dunes and wetlands), and accommodation through flood proofing and elevation of infrastructure [47]. Storm surge barriers are considered an option at a later stage [48]. Thus, a multi-pronged approach opens possibilities.

Rural open coast/delta/estuary
Archetypes 'rural open coast/delta/estuary' have similar and fewer adaptation options and pathways and are therefore discussed together (figure 2). Impacts are similar to those in their urban counterparts, but preferred adaptation options are reduced and/or tipping points are different due to lesser socio-economic consequences because of lower population and infrastructure density. Adaptation typically focuses on maintaining food productivity and the natural environment benefits. Low cost-benefit ratios may limit adaptation pathways.
Pathways for rural areas emphasise accommodate and retreat options before protect. For example, to maintain food production as sea-level rises and salinity and groundwater levels increase, a typical short-term measure is to improve or continue to maintain field drainage, possibly complemented with pumps (in figure 2 this is considered as part of the current situation). Productivity may be further enhanced by switching to flood tolerant or salt tolerant crops, or to aquaculture (e.g., southwest Bangladesh [49], Mekong delta). On the longterm, if sea-level continues to rise and flooding becomes permanent, managers are left with options to relocate or raise the land. Raising land may be undertaken through river diversion, such as being done or planned for in rural parts of the Mississippi delta [34], the southwest of Bangladesh [49] and the Ebro delta [50].
Another pathway could start with low-cost green protection measures with for example reed beds or mangroves, to dissipate waves and reduce erosion and flooding. For example, pathways for the Danube and Ebro deltas first consider green protection with reeds combined with raising the land via strategic sediment measures, with a later option of set-back lines within a planned realignment of the coastline [50]. Along parts of the coast in the UK (e.g., The Wash), Germany (e.g., Langeoog Island), and the Netherlands (e.g., Westerschelde), managed realignment is implemented to restore saltMarshes and to aid coastal defence [51][52][53].
Selecting preferred pathways is based on trade-offs between different criteria reflecting management aims such as food production or mitigation of potential infrastructural damage. At the same time decisions on adaptation are also driven by other incentives, such as economic development. In rural south west of Bangladesh, this triggers the implementation and development of pathways with dykes, drainage and pumps [54]. For the Mekong Delta, accommodate/retreat pathways have been explored, consisting of adapting agriculture to enhance yield, diversifying livelihoods to ensure other sources of income, and migrating to less hazardous areas [55]. However, current governance focuses on protection options, like raising dikes, to enable socio-economic development, which benefits triple-cropping agriculture on the short-term, but may lead to reduced productivity in the long-term without costly fertilizer, thus penalising poor farmers [56]. In the end, choosing for flood protection through dikes may lead to path-dependencies that could result in non-inclusive outcomes [56] and ultimately reduce the possibility to pursue accommodate and retreat [55].

Meta-pathways for archetypes
Based on our analysis (figure 2), common traits of coastal adaptation pathways emerge (figure 3). In urban environments, the immediate priority is to protect, by either soft or hard measures. The path of protection tends to be self-reinforcing, because by virtue of the 'levee effect'; people and assets tend to accumulate in protected areas, in turn requiring higher protection, in a feedback loop [57]. Accommodation could extend the effectiveness of protective measures, but in the end stronger protection may be needed, and retreat remains the last option if protection is not possible or preferred anymore. In rural land hard protection is difficult to motivate: present interventions are minimal and mostly in the direction of accommodate, with a possibility to delay the tipping point through the combination with protection measures (of relatively small investment). However, with medium to high sea-levels retreat remains the last option, unless new technologies delay the tipping point and extend the lifetime of accommodate measures. Figure 3 indicates that adaptation tipping points will occur sooner in rural than in urban areas, as different resources are available. Social acceptability is a major barrier to switching adaptation types [26], and economic analysis often suggest that money may be spent more effectively elsewhere [27]. This indicates that, with time, retreat is a realistic outcome for both urban and rural areas, but may come earlier for rural and for different reasons.

Towards local pathways
Our generic pathways (figure 2) provide a framework to develop site-specific adaptation plans to sea-level rise. First a coastal manager needs to identify their coastal archetype from the six options. In practice, hybrid and nested archetypes exist besides our six archetypes. Many morphological classifications have a hierarchical structure, where one morphological type or land use may be embedded in another [21]. For example, a delta system could comprise a sandy beach at the delta front (e.g., Ebro delta). This is representative of 'open coast' geomorphology, nested within a 'delta' geomorphology. While these complexities cannot be considered in the scope of our archetypal analysis, we recommend that, upon applying our archetypes for the design of localised adaptation pathways, any subareas within a larger coastal archetype should be considered as a coastal archetype in their own, depending on size and relevance and on the management scopes. Thus, options from multiple archetypes may need to be considered in real world cases.
Similar land use types nest within another. For example, if a nuclear power station was situated on a rural coast (e.g., Sizewell in Suffolk, UK), coastal adaptation there might follow the path of urban coast, as high protection standards are required. Another example is low-lying farmland which maybe a valuable asset and therefore protected. Hence each feature of morphology and land use must be considered in a wider context. Then following the methodology described in section 2.2, local managers must clearly define their management goal (step 1). Next, the full range of adaptation options need to be explored (step 2), taking account of local perspectives. Local adaptation pathways require specific information to select and complement the adaptation measures that most align with the case context, and to define their threshold sea-level rise conditions (step 3) at which adaptation and opportunity tipping points occur (similar to table 1), as seen for the local pathway of Lakes Entrance [10] (their table 1). These tipping points should take account of the possible rate of regional sea-level rise and its effects (e.g. number of days inundated), as well as other processes and criteria which influence decision making. Next, the pathways from figure 2 can be adjusted to local conditions (step 4), first at a generic level by selecting the relevant pathways and adjusting the tipping point conditions; and then towards more detailed levels, possibly with site specific adaptation measures (e.g. split the adaptation step 'protection through dikes' into dikes up to 0.5 m of sea-level rise, followed by dikes up to 1 m of sea-level rise). The result is a set of nested pathways that describe different levels of detail.
At local level, system-specific information and stakeholder participation are vital in debating and selecting adaptation measures (e.g. in Lakes Entrance this was achieved through a telephone survey to identify the important features on the local environment [10]-see their figure 2), and to define their tipping points with respect to future sea-levels, other drivers of change and other criteria which influence decision making. With this local information in combination with the typology of pathways, local adaptation pathways can be designed. For successful implementation, pathways need to be complemented by good, continuous governance [16,58,59], where all stakeholders work towards the overall management goal, rather than their own narrow objectives.

Conclusions
Adaptation pathways boost flexibility and sustainability in decision making for coastal adaptation, yet they are limited in application due in part to lack of experience and the complexity involved in their generation. To aid take up at local level, pathways have been generated generically for six coastal archetypes, and illustrated with examples at local settings.
We illustrate that presently adaptation options decrease with rising sea-levels, unless we radically change our approach to coastal adaptation by exploring considering adaptation pathways and making the necessary preparations to timely adapt. The pathways analysis also shows that, for high sea-levels, options will need to be considered that are not presently acceptable, but may be needed in the end. This helps to avoid unsustainable investments with potential for lock-in. In urban areas, there is a greater motivation to protect and accommodate rather than retreat. However, accommodation cannot continue forever, and in the long-term, protect, or planned retreat are options that could become more common [60]. Inaction could lead to unplanned retreat [61] or lack of adaptation options in the end.
Exploring adaptation pathways to sea-level rise can help coastal planners to evaluate the sustainability of their investments for coastal adaptation under uncertainty. We show that this approach allows for a richer analysis of the operation space for coastal adaptation than has be done with static assessments, and takes into account the uncertainty and timing of adaptation needs.
Rarely do people adapt to sea-level rise alone, with many factors influencing the need to change. Additional criteria, such as higher economic development or the effects on the natural environment are also considered, and will influence how the pathways result in practice. Our generic pathways serve as inspiration as to what is physically possible, but local decision making and stakeholder engagement is key to determine what is acceptable.
By just taking account of physical constraints, the lead time of measures and adaptation planning frequently needs to start earlier than anticipated, especially as rapid sea-level rise is a risk and may require larger time consuming adaptation efforts. Local stakeholder engagement to enable effective decisions making would further extend this time. Therefore, with potential accelerated sea-level rise [3], exploring pathways and monitoring to detect signals for adaptation becomes more urgent as then time available for planning and implementation will be less.