Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Considering socio-political framings when analyzing coastal climate change effects can prevent maldevelopment on small islands

## Abstract

Adapting to climate change and sea level rise is challenging on small islands. False adaptation can lead to adverse impacts on natural and societal dynamics. Therefore, an interdisciplinary perspective on the interaction of natural dynamics, societal demands, and political decisions is crucial. In this sense, this study scrutinizes coastal processes and socio-political dimensions of erosion on the reef island Fuvahmulah, the Maldives. The national government and Fuvahmulah’s population have an opposed perception and attribution of the drivers and processes behind Fuvahmulah’s most pressing coastal issue – coastal erosion. To review these perceptions, natural dynamics are recreated with process-based methods and discussed regarding present and projected marine pressures. Population surveys and interviews with actors in coastal development complement the physical insights into erosion on Fuvahmulah and describe the socio-political dimension of climate change adaptation on small islands. This interdisciplinary approach demonstrates how small-islands’ adaptive capacities are typically impaired and disclose the potential of local knowledge to overcome maldevelopment.

## Introduction

There is high confidence and evidence that climate-change induced sea-level rise increases the risk of small islands to suffer from ocean-borne impacts associated with high tides and swells1,2,3. Small low-lying islands and atolls are especially susceptible to the impacts of sea-level rise, as there is often no higher ground available for the local population to retreat to. In the upcoming decades, stronger and more frequent seaborne hazards will follow global heating, making small islands – in extreme situations – potentially uninhabitable2,4. Seaborne hazards on low-lying islands are typically rising sea levels, severe storm surges and swell waves. Likewise, these environmental pressures are constituent parts of the natural dynamics: Coastal environments react to hydrodynamic pressures with a morphodynamic response. Thus, wave-induced currents on the reef platform are an important driver of sediment transport5,6 and the interplay of waves, sea level, reef platform, and sediment production are major elements of reef island genesis and evolution5,7,8,9: For example, when looking at the Maldives, monsoonal wind-wave events have an impact on locally characteristic island morphology, but, in general, distant-source swell waves are dominant in the regional wave climate5,8,10. Swell waves are characterized by a long wave period and play a key role in the island development on reef platforms11,12. Together with constant sediment provision by the coral reef13, the wave-induced sediment transport allows reef islands to mitigate erosion14 and accrete vertically in response to sea level rise11,12. Coral reefs are important coastal protection assets15, because the reef dissipates most of the incoming wave energy16. There is evidence that reefs have grown and can continuously provide protection under rising sea levels11,17 if they remain healthy. As a consequence, healthy coral reefs are a very important – if not the most important – factor of the island’s resilience to withstand sea level rise and marine extreme events. However, global warming and ocean acidification threatens the health of coral reefs18 and with that also the important ecosystem services coral reefs provide.

For small islands, coastal protection is an important ecosystem service. Owing to their geography, small islands’ populations are more exposed to marine hazards and more vulnerable than mainland communities. In addition, marine resources deliver important ecosystem services for island communities, because they provide food and income for the island’s inhabitants19. However, protection from increasing risk of seaborne hazards as well as connection and exploitation of marine resources require concerted development of coastal areas, followed by construction and maintenance of infrastructure. To date, coastal structures and anthropogenic interventions have often resulted in the proliferation of engineering-type hard-coastal protection systems, defending coastal areas up to a certain level of safety from coastal hazards. These interventions are man-made disruptions of the coastal system, often undermining vital ecosystem services and interfering with the morphodynamic interchange on beaches20,21. Nevertheless, such structures are implemented widely22, even though they interrupt the littoral currents or deteriorate coastal environments to large extents. Anthropogenic interventions in natural dynamics put further pressures on the environment and are driven by political decision-making processes and social interests23.

In their first assessment report, the Intergovernmental Panel on Climate Change (IPCC) names three general response strategies to adapt to increased coastal risk and sea level rise24: protection, accommodation and retreat. In the meantime, further alternatives have emerged and in 2019, the Special Report Ocean and Cryosphere in a Changing Climate (SROCC) of the IPCC complements the response strategies by supporting a broader range of options3: protection, accommodation, advance, retreat, ecosystem-based adaptation, or no response. However, especially small island countries are often confronted with a number of issues that further complicate finding sustainable solutions to infrastructure demands25,26: on the one hand, decisions are regularly made from afar by the central government without addressing or integrating local knowledge. On the other hand, island states often lack expertize as well as human and financial resources to allow thorough planning and proper implementation27. Thus, development decisions are embedded in political structures on multiple levels of governance, determining how decisions are made and what actors are involved in the process25,28,29.

## Results

### Study site

Fuvahmulah is an island in the south of the Maldives, located approximately 30 km south of the equator (latitude: − 0.30°, longitude: 73.43°, see Fig. 1). Unlike most other inhabited islands of the Maldives, Fuvahmulah is not part of a ring-shaped atoll and lacks the distinct protection features of oceanward islands8. Fuvahmulah consists of only one main island and its fringing reef. The main island has a size of about 4.4 km by 1.0 km and a coastline of ~11 km length. The island is surrounded by ridges, which are located landward behind the beaches and predominantly vegetated by a fringing palm forest. The island’s ridges reach a height of up to about 4 m above mean sea level. Together with the reef, this natural barrier mostly protects the island from storm surges, swell waves and helps mitigating overwash. Behind these natural coastal protection systems, Fuvahmulah’s inland is at mean sea level and hosts two freshwater lakes.

Among islands of the Maldives, the landmark of Fuvahmulah is its sandy beach Thoondu, a coastal spit situated at the north of the island. As wave direction changes with the seasons, Thoondu responds to these natural dynamics5: it seasonally adjusts its landform shape, alters its sediment volume and regularly meanders from the northern part of the headland in wet season towards the north-east of the island as well as to the northwestern Geiymiskih beach in the dry season. The constant change of Thoondu is well known by locals and contributes to its uniqueness. As consequence, in 2020, the United Nations Educational, Scientific and Cultural Organization (UNESCO) designated Fuvahmulah as a Biosphere reserve.

The closest neighboring island is Hulhumeedhoo (part of Addu Atoll) located about 40 km southwest of Fuvahmulah. The 8510 inhabitants41 of Fuvahmulah are exceptionally isolated – even when considering all 26 atolls in the Maldives, stretching over about 870 km length (north to south) and about 130 km width (west to east). The isolated location and with the reef and vegetated coastal ridges serving as the only natural coastal protection, Fuvahmulah is particularly susceptible to environmental forcing. Fuvahmulah’s location close to the equator makes it less likely to experience cyclones10,42 but the island is considered to be highly exposed to monsoon winds43 and associated wind-waves, as well as to distant-source swells5,8,10,26.

The isolated position makes Fuvahmulah dependent on transportation infrastructure, providing a safe connectivity, an access to the outside world. The transport infrastructure on the island endows the local economy and supply – especially through the shipping of goods. However, until 2002, Fuvahmulah was only accessible by small fishing boats (“dhonis”), which had to navigate through the breaking waves over the reef plateau. The “dhonis” acted as feeders, transporting people, goods, and cargo from Fuvahmulah to larger ships that were waiting offshore. In this procedure, numerous boats capsized and people died. For decades, the people of Fuvahmulah have been requesting the construction of a safe port. The new seaport was constructed on the southeast of the island opening in 2002. The seaport was followed by the construction of an airport for domestic flights in the southwest of the island that opened in 2011.

### Measured coastline erosion on the east side

Ever since the port has been built, people on the island observe and report severe erosion on the east-side of Fuvahmulah44. The Digital Elevation Models (DEMs), derived from aerial imagery, prove this observation and quantify the observed erosion rates and its spatial extent along the east coast. Between 2017 and 2019, the cross-sections along the east side show distinct differences in morphologic changes (Fig. 2): while the northern area close to Thoondu (Fig. 2a) shows a slight increase in beach volume, the other transects depict ongoing erosion (Fig. 2b–d). The increase in sediment volume in the north at Thoondu beach is due to the typical dynamic behavior of the northern beaches in response to varying seasonal wave conditions: during dry season, Thoondu beach moves to the northeastern coast5. The cross-section at the end of Thoondu beach (Fig. 2a) shows the calcareous bulk material forming the beach face. When moving further south, the coast of the central east side has a steeper profile than the northern beaches and consists of gravel to cobblestone material (Fig. 2b) or larger boulders. The two southernmost transects near the seaport have steep edges, where the mainland meets the reef flat (Fig. 2c, d). These transects are located about 700 m (Fig. 2c) and 190 m (Fig. 2d) in the north-west of the seaport’s entrance. At these locations, the subsurface insular bedrock becomes visible. The erosion process is already noticeable from field observations and has also been widely recognized and reported by locals.

Along the southern stretches of the east coast, there are further signs of erosion, such as uprooted trees. For example, adjacent to the most southern cross-section was a palm tree that later fell onto the reef (see Fig. 3). Large parts of Fuvahmulah’s coastal ridge have a coastal forest (“heylhi”), reinforcing the sandy beach profile with its roots. On the east side, waves and currents carve out the sediment under the roots and dislocate broken bedrock material. These observations show that erosion on the east side of Fuvahmulah starts above the reef flat at the bottom of the beach profile. These observations also help to put erosion into context: while the northern transect shows slight sediment accretion, the central part shows shoreline retreat with maximum values of 0.26 m to 0.28 m between 2017 and 2019 (Fig. 2d). However, shoreline retreat was only occasionally measurable and not present over the entire central coastline – in contrast to the southeastern coast, adjacent to the port. This area experiences substantial ongoing erosion along its entire coastline in the order of 0.3 m between 2017 and 2019 with maximum values ~1.33 m (see Fig. 3). The field data quantifies erosion on the southeast of Fuvahmulah in the order of decimeters over two years. Erosion rates on the east coast decline towards the north. In the north-east of Fuvahmulah, the highly dynamic morphology of Thoondu stabilizes the coast.

### Wave climate of the southern Maldives

Both, erosion adjacent to the harbor and sediment dynamics on Thoondu beach, are effects of sediment transport dynamics on the reef. Here, the main driver behind sediment transport are wave-induced currents5,7. Therefore, the local wave climate and possible changes of the wave climate are of particular importance when assessing shoreline changes. As measured wave data for the Maldives is scarce, global wave hindcasts or reanalysis models are the only sources containing long-term wave data of directional sea states for the islands of the Maldives. Information on wave data is provided by several meteorological services, such as the National Centers for Environmental Protection (NCEP) of the American American National Oceanic and Atmospheric Administration (NOAA), the European Centre for Medium-Range Weather Forecasts (ECMWF) or the Collaboration for Australian Weather and Climate Research (CAWCR) (see Fig. 4a). Annual time series from the considered services agree very well among each other, with mean annual correlation coefficients of R ≥ 0.89 (Fig. 4c). To validate the hindcast and reanalysis data, this study uses harmonized and inter-calibrated Satellite Radar Altimetry (SRA) measurements from the Altimeter Data System (ADS) of the Helmholtz Centre Potsdam (GFZ)45 and compares these to the three data sets considered for this study (Fig. 4a). The SRA measurements are available between 1993-04-25 and 2018-06-15. Wave heights from ECMWF’s fifth generation atmospheric reanalysis of the global climate (ERA5) data set are closest to the measured SRA wave height in the region of Fuvahmulah. Even though the locations of the CAWCR output nodes differ from the other data sets, analyzing the spatial variation of the CAWCR time series shows a very high correlation of adjacent data nodes (R ≥ 0.977 ± 0.007, see Fig. 4d). The spatial similarity following from this high correlation of wave parameters allows the data sets to be compared among each other – despite the spatial distance. Likewise, this shows the wave climate on Fuvahmulah is typical for atolls in the south of the Maldives – as well as the associated exposure and impacts. All further analyses of this study make use of the CAWCR data, because the output node is closest to Fuvahmulah and CAWCR provides significant wave heights under Representative Concentration Pathway (RCP) 4.5, and RCP 8 for the twenty-first century46. These future projections allow this study to scrutinize future states of wave climate on the island as well as their impacts and triggered effects.

The wave climate in the area around Fuvahmulah reflects the dry and wet season both by changed wave heights and wave direction. Median wave heights range between 0.98 m in March, as well as 1.71 m in July with a maximum wave height Hs,max of 3.32 m on June 22nd, 1987 (for more statistical data of Fig. 5a, f, for example on boxplot mean and IQR values, see Supplementary Information File). This study defines the period between November to February as dry season and the period between April and September as wet season (Fig. 5a). April and October are considered transition months.

The waves are smaller in the dry season than in the wet season and are dominantly approaching the island from SSW (202.5°, see Fig. 5b–e). There are also other southern and south western portions between November to February, but in terms of occurrence they play a minor role. Significant wave heights Hs are mostly smaller than 1.7 m in the dry season (see Fig. 5c). In the wet season, waves approach the island dominantly from southeast and south-southwest (135° and 202.5°). In the rainy season, significant wave heights increase when compared to the dry season and range between Hs = 1 m − 2 m. Considering the future wave climate in the region, by using the projections of the Coupled Model Intercomparison Project Phase 5 (CMIP5), the data does not significantly differ under the respective RCPs, when compairing wave parameters between the reference time frame 1986–2005 and the last two decades of the twenty-first century (Fig. 5f; reference time frame according to the fifth assessment report of the IPCC47). These results are in line with findings of another study using CAWCR wave projection data on a global scale48.

Considering the results of the statistical wave climate analysis, this study concentrates on scrutinizing waves which propagate from SE and SSW into the computational domain of the numerical models in the following. These wave directions represent the two dominant shares of the wave rose for Fuvahmulah in both seasons. Considering today’s and future significant wave heights Hs, the numerical models use the 99th percentile from the reanalysis data set (Hs,model = 2.3 m ≈ Hs,99% = 2.27 m) as boundary conditions. All conditions are therefore modeled for storm conditions.

### Natural morphodynamics on the reef and anthropogenic interventions

Results from the wave climate analysis serve as boundary conditions for the regional wave models5. The numerical wave models also utilize the DEM data and set the observed and measured morphodynamics into broader context by considering the wave climate’s hydrodynamic forcing. With that, the wave models test the assumption of the harbor infrastructure being the root cause of erosion on the southeast coast, adjacent to seaport. While Delft3D (D3D) calculates the general sediment movement on the Fuvahmulah reef platform for different wave directions, the depth-integrated (2DH) Boussinesq-type wave model gives more detailed information on the role of the port as anthropogenic interference in the natural sediment transport system of the reef (a preceding study gives further insights into the more general dynamic pattern of seasonal and annual sediment transport on the fringing reef of Fuvahmulah5).

D3D confirms that the interplay of sediment supply13 and wave-induced currents provides sediment for the island (Fig. 6). It further indicates where to expect sediment deposits under the given hydrodynamic forcing. As waves approach from θp = 202°, sediment moves towards the island and settles on the west side (Fig. 6a). A smaller sediment depot also forms on the east and southeast of the island under the given premises. Waves from θp = 135° will take sediment from the offshore reef and transport it towards the island (Fig. 6b). The computations for waves from θp = 135° disclose that the sediment stream splits at the southeastern tip of the island, approximately at today’s harbor location. The wave-induced currents distribute the sediment along the east coast of the island as well as on the southwest side. A small portion of sediment also moves around the northern tip of the island onto the lee side.

In general, the D3D computations reveal that wave-induced currents take sediment from the southern reef and transport it towards the south and east coast (without seaport structures being present). D3D also indicates the particular importance of the island’s southeast part for sediment transport along the entire east coast: waves from θp = 202° pick up sediment at the southern reef, transport it towards the east and the southeast coast of Fuvahmulah where it finally accumulates. Then, over seasonal cycles, waves from θp = 135° can take away the accumulated sediment and distribute it along the east side.

The 2DH model highlights the difference in wave driven current patterns leading to sediment transport around the port area. The models also facilitate scrutinizing the changes in sediment transport with and without the harbor infrastructure being present. These computations confirm the processes outlined with the D3D computations and further highlight two factors, contributing to the erosion along the east coast of the island:

The first factor is the available sediment. In both cases – with and without the port structures (breakwater and headland) – waves from θp = 202.5° create a current in front of the port (Fig. 7a, b), transporting sediment over the reef. At the same time, in the area of today’s harbor entrance, the current decelerates and thus allows for the sediment to accumulate in this area. However, with the breakwater present, the structure interrupts the sediment transport and sediment cannot enter this area. In addition, the breakwater reaches up to the reef’s edge. As a consequence, it deflects the current and thus redirects the sediment off the reef into deeper waters.

The second factor is the transport capacity of the east coast’s current: without breakwater, waves from θp = 135° induce a northward current to the area of today’s port entrance (see Fig. 7). This is also the area, where sediment was able to settle from the θp = 202.5° component. Instead, with the harbor present, the breakwater obstructs the emerging current in this area and deflects the velocity momentum off-shore. Yet, a wave-induced longshore current is still present along the east coast independently of the port: when waves from θp = 135° approach the reef and break, they induce radiation stresses and subsequently create this longshore current. With missing sediment from the reef, the currents will likely take sediment from the coast, leading to erosion.

### Governmental and societal framing of coastal development

On Fuvahmulah 93.8% of residents are citizens of the Maldives and as on most other inhabited islands, tourism plays a minor role41. Fuvahmulah’s island dwellers are mainly employed in education and commercial services41. Thus, together with the insights of the field campaigns, Fuvahmulah can be considered a local’s island. In the household survey, the local community perceives erosion as the most pressing issue (closed question, 27% of 345 mentions, see Fig. 8 and Supplementary Information File). At the same time, both the interviews with government officials as well as the recently issued Environmental Impact Assessments (EIAs) on coastal protection on Fuvahmulah reveal that administrations on the national and local level acknowledge erosion as a high priority issue on the island. However, while there is consensus on the need for action among all interviewed actor groups, attribution of the erosion’s root cause varied between the national government representatives and the affected community. On the one hand, the Maldives’ dominant adaptation challenge is sea level rise and the associated impacts. In fact, the Maldives national government actively promotes this narrative of being highly vulnerable to climate change induced sea level rise49 – even in the case of Fuvahmulah. On the other hand, Fuvahmulah’s population is skeptical about the national government’s narrative. The household survey contained an open question (without pre-formulated responses) asking the participants to name visible changes of the natural environment on the island. Here, about 36% of all responses mentioned erosion and of these, 20% also attributed the erosion processes on Fuvahmulah specifically to the harbor construction (see Fig. 8 and Supplementary Information File). These diametrically opposed perceptions on root causes behind coastal adaptation originate from the historical context of national politics and local society.

First experiences with coastal infrastructure in the Maldives dates back to the 1970s, when the first modern coastal protection constructions were built on the islands34. However, coastal protection management was not professionalized before 1987 and 1988 when strong coastal floods caused massive damage to the Maldives’ capital Male’ and on several other islands in the middle of the country (interview with representatives of the Ministry of Environment and Energy (MEE) in 2017). At the same time, establishing coastal protection in the Maldives was accompanied with a rapid increase in population, especially in the Maldives capital Male’. In accordance with the Maldives Decentralization Act in 2010, the national government shifted its focus onto regional development centers, equipped with the necessary infrastructure26 and limiting the migration to the capital50. However, the interviewees still highlight today’s accumulation of resources – economic and professional expertize – in Male’. Decisions are made at the highest level and have an impact down to the local scale. When regarding the implementation of coastal development projects in the Maldives, the planning, implementation and decision-making process are centrally executed and ministerially anchored in the national government without significant involvement of local capacities on the islands. According to actors involved in coastal governance, in general, such top-down processes in decision and implementation are applied for coastal infrastructure projects, for example seaports as well as coastal protection structures. Regional and local government institutions, such as the city council on Fuvahmulah, lack influencing power in the decision-making process. According to the interviewees, the role of local government institutions is limited to informing national-level actors about coastal problems on their island. The lack of power is also expressed by missing financial resources for coastal projects at a council-level (interview with a representative of a state environmental agency in 2017). In addition, infrastructure projects usually require external financial and knowledge resources, provided by international organizations. However, international organizations are legally bound to use the national government as an entry point and cannot initiate projects below the central national-level. Altogether, this stands in contrast to the republic’s decentralization efforts and the corresponding strengthening of local communities.

Moreover, since 2012, the EIA Regulation of the Maldives requires to assess the (adverse) impact of infrastructure projects on the environment. However, quality and compliance to EIA policies – also for example in the tourism sector39 – are traditionally weak40 and have extensively promulgated technical fortification of many inhabited islands in the Maldives20,22,26.

In 2014, government actors of the MEE agreed on taking action against erosion adjacent to Fuvahmulah’s seaport. They signed a grant arrangement for an internationally sponsored infrastructure development fund and initiated the project “Coastal Protection at Gn. Fuvahmulah, Maldives”. This project includes “the development, implementation and maintenance of sustainable coastal protection to prevent erosion and flooding on […] Fuvahmulah. The main objective of this project is to decrease erosion and flooding through a possible […] combination of hard and soft coastal engineering interventions […] to protect the island of Fuvahmulah against flooding due to ongoing coastal erosion and rising sea levels”50. An international consultant supported the MEE together with a local sub-contractor in the site investigation, the feasibility study, the technical design50 and the draft of the legally mandatory Environmental and Social Impact Assessment (ESIA)51.

The technical design report finds human interventions and changes in natural processes disturb the natural balance of the sediment supply along the coastline50. According to the report, these human interventions are more specifically the harbor construction, sand-mining and earlier reef blasting for beach access, while natural processes disturbing the natural balance are “changes in the natural climate”50, but also “may”50 be a possible deterioration of the coral reef. The report also mentions a lacking local sediment source for the coast and predicts the fringing ridge of Fuvahmulah to breach in the next decades, causing “tremendous flood damage to the island”50. Because of this, the report concludes, coastal protection is required and the design should “immediately stop the ongoing erosion”50. A preceding feasibility study was cited in the report, identifying two possible solutions against the ongoing erosion and rendering them reasonable while considering their environmental and social impacts – an offshore breakwater and an onshore revetment. The technical report concludes the revetment to be the most viable option, after presenting preliminary designs and a cost-benefit analysis of both alternatives. The final revetment is about to cover approximately the southern 2.6 km of the 4.0 km long east side and is designed with a lifetime of 50 years, withstanding hazardous seaborne events with a return period of 100 years.

The ESIA51 builds on the technical design report and further underlines the project’s significance in the context of future climate change risks, as well as for the national and local economy. The report also finds the project would enhance the “sense of security and safety” of Fuvahmulah’s island dwellers51. The ESIA initially gives five approaches to coastal adaptation, but finds ecological methods would contain a high uncertainty and that they would be ineffective. Also, the report further dismisses soft measures (also known as sediment-based measures) as unsuitable because a proper (local) sediment source is supposedly missing.  The impact assessment report finally presents the same coastal protection alternatives that have been suggested in the technical design report50. However, in addition to the technical design report, the ESIA also contains stakeholder consultations on the project. Resulting from these consultations, the chapter “project risk mitigation” in the ESIA finds “acceptability of the Project by all stakeholders is an important factor for smooth implementation of any project”51. The first local stakeholders consultation took place in July 2016 on Fuvahmulah and addressed the off-shore and on-shore protection alternatives. The discussions revealed locals prefered the off-shore alternative, but the ESIA objects as “the risk of affecting [Thoondu] beach was not well understood [by the local stakeholders] at that time”. Therefore, the ESIA mentions another consultation with the local government and community in October 2016 “to explain to the stakeholders the preferred alternative that emerged following the feasibility study and proposed in the EIA report” - with the revetment being the “preferred” and to be implemented alternative. In summary, both reports dismiss feasible soft solutions due to false assumptions (missing sediment source, however the reef naturally provides marine aggregates5,9) and undermine the participatory process by promoting the less popular alternative in the end.

In principle, public participation is a key component of EIAs39 and when taken into the decision-making process, public participation also leads to more sustainable implementation of development projects. However, the population survey of this study from March 2017 reveals an unutilized potential to improve the participatory process based on two questions: First, when asking how far politicians work together with the community (open question), 49% of the survey’s respondents found insufficient cooperation of politicians with the community in development projects – while 19% answer to be sporadically involved and 17% report to be regularly involved (see Fig. 8 and Supplementary Information File). Second, a large majority of 83% endorses more involvement in the decision-making process regarding the development of the island, while only 15% of the interviewees do not wish for more participation in this matter. These results disclose an ongoing dissatisfaction of Fuvahmulah’s people and the lacking consideration of their interests, opinions, and knowledge – despite or because of insufficient earlier attempts to include their perspective in the ESIA.

## Discussion

The example of Fuvahmulah epitomizes planning inadequate coastal structures in a reef environment underlies special environmental features and requirements beyond conventional, generic infrastructure design rules and solutions (see the results section “Natural morphodynamics on the reef and anthropogenic interventions” and Fig. 7). Traditionally, human intervention beyond the natural green belt of Fuvamulah used to be highly limited, leaving the protection feature and natural sediment distribution on the reef intact. However, the first hard-engineered protection measures have already led to complications44. Sand-mining is discussed as a factor and cause for coastal erosion on Fuvahmulah. Sand-mining can potentially put anthropogenic pressure on small islands25 and on Fuvahmulah, sand-mining for private purposes was frequently witnessed within the field campaigns around the island. A further reason for increased erosion could be climate variability. However, this study does not include an in-depth assessment on recent climate variability and possible effects on the east-coast of Fuvahmulah, albeit there are tendencies of changing climate impacts (for example sea surface temperature oscillations in the Indian Ocean), which could have affected the formation of Thoondu beach in 20195. However, the most visible and distinct adverse changes on Fuvahmulah’s coasts is local erosion, located adjacent to the seaport in the southeast. Based on the field campaigns and the recorded high-resolution data, this study finds coastal erosion on Fuvahmulah to be a negative consequence from the seaport. The seaport is an anthropogenic intervention into the coral reef system adversely impacting the natural sediment transport and the associated wave-induced redistribution pattern around the island – much more than sand-mining or climate variability.

The perception analysis and population surveys on Fuvahmulah reveal that island dwellers are well aware of the port’s impacts and are doubting the national government’s narrative, framing coastal erosion as climate change and sea level rise impact. Local knowledge helps to identify and overcome local challenges, reflected by both the traditional primacy of island development within the borders of the “heylhi” and veritable attribution of the current coastal erosion. Analyzing the wave-induced sediment processes in numerical wave models corroborates the local people’s assumption that the harbor infrastructure interrupts coastal dynamics and acts as a sediment barrier for the east coast. Local knowledge conveys a distinct local capacity towards economic development in combination with future environmental protection on the island. Implementing and evolving these capacities into a tailored adaptation pathway plan facilitates to consider smaller and less intrusive responses beyond conventional hard-engineering approaches, discloses their potentials and limitations, and leaves further scope of action for the future. Such a more conceptual approach of adaptation pathway planning requires (1) a greater focus on the underlying process, (2) a systematic collaboration between local, national and external experts and stakeholders on the island, (3) ongoing monitoring on the natural and societal impact of the implemented measures, and based on this (4) frequent adjustments on the pathway roadmap by responsible authorities when an adaptation tipping point is reached and the current status (a policy action or actions) will fail. In this regard, dynamic adaptive policy models represent suitable, modern methods, allowing more conceptual approaches to be embedded into continuous and sustainable coastal planning59.

Amalgamating coastal and societal aspects leads to a comprehensive understanding of the ongoing natural and socio-political processes behind coastal issues. These interdisciplinary insights underline that adaptation strategies and measures need to be adjusted to local circumstances. This helps to (1) avoid an interruption of the coastal system, (2) undermine vital ecosystem services and (3) sustain natural coastal protection.

## Methods

### Digital elevation model (DEM)

Based on three field campaigns, Unmanned Aerial Vehicles (UAVs) capture the topography of Fuvahmulah and facilitate analyzing shoreline changes. A DJI Phantom 4 multicopter recorded aerial photos within two field campaigns in March and September 2017. A DJI Phantom 4 Pro was used for the coastal survey in March 2019. The aerial surveys covered the entire coast of Fuvahmulah, except the beaches adjacent to the airport, which are a no-flight-zone. After recording the aerial images of Fuvahmulah’s coast, Agisoft Photoscan Pro (version 1.4.5 build 7354) processes the photos into DEMs by means of Structure from Motion-MultiView Stereo (SfM-MVS)60. DEMs are three dimensional models of the area, containing height and color information, here in about 3.5 cm px−1 resolution. The resulting coastal DEMs were recorded at different points in time, facilitate analyzing temporal changes of sediment volumes on the beach, shoreline location and vegetation. The general workflow in Photoscan is to align the photos, georeference the model with Ground Control Points (GCPs), and create a dense point cloud as well as a mesh of the area61. Photoscan derives the DEM from the dense point cloud, while the mesh serves as base for the orthophoto (for further information on input parameters of each processing step in Photoscan, see Supplementary Materials). To position the models in the World Geodetic System (WGS84), the March 2019 DEM uses a unique set of GCPs around the island. The 2017 model uses the drone’s Global Navigation Satellite System (GNSS) antennas for initial positioning. Afterwards virtual Ground Control Points (vGCPs) of specific landmarks available in both models georeference the 2017 DEMs with the 2019 DEMs. Apart from the erosion, the position of dedicated landmarks between all three field campaigns were reproduced with the expected accuracy61,62 for SfM-MVS procedures in coastal areas.

In general, the inspected erosion starts at the bottom of the beach profile. Waves and currents carve out the sediment under the roots of the coastal forest or dislocate coral rocks. Owing to the specific erosion process on Fuvahmulah, early erosion remains mostly undetected in the DEMs, mostly until the upper part slips or collapses. However, adjacent to the seaport, the erosion is advanced enough to be well visible in the DEMs.

### Hydrodynamic boundary conditions

To evaluate the drivers behind sediment transport, this study uses numerical near-shore wave models. These models require regional wave climate information as hydrodynamic boundary conditions. To derive the wave climate for Fuvahmulah, this study considers data from three global hindcast and reanalysis models from the Collaboration for Australian Weather and Climate Research (CAWCR), National Centers for Environmental Protection (NCEP) of the American National Oceanic and Atmospheric Administration (NOAA), and European Centre for Medium-Range Weather Forecasts (ECMWF) (see Table 1). Data from wave spectra are an appropriate resource to be used as boundary condition in models of finer spatio-temporal resolution63,64. To validate the modeled wave climate data, this study compares each data set with wave heights derived from individual harmonized and inter-calibrated SRA measurements of the GFZ Altimeter Data System45 (see Fig. 4a). The SRA data is taken from latitude 0° and longitude 73° in a radius of 1.5 km between 1993-04-25 and 2018-06-15.

The CAWCR provides hindcast wave data65 generated with the WAVEWATCH III ocean wave model66. The closest output node to Fuvahmulah and the SRA data from GFZ is at 0.0° latitude, 73.2° longitude. The model uses a grid with 0.4° zonal and meridional wide increments. CAWCR also provides estimates for wave heights under different RCP in the twenty-first century. WAVEWATCH III is a numerical model calculating wind wave generation and wave propagation based on the random phase action density balance equation for wave spectra66. The waves are forced by surface wind fields from atmospheric models67, made available by the NCEP as Climate Forecast System Reanalysis (CFSR) and Climate Forecast System Reanalysis v.2 (CFSv2). The Environmental Modeling Center (EMC) of NCEP runs an own hindcast with WAVEWATCH III and their bias corrected wind forcing. EMC maintains WAVEWATCH III and places the code at disposal. WAVEWATCH III is a fork of the ocean wave model (WAM68). The ECMWF uses the WAM model for their ERA5 with wind forcing from their Integrated Forecasting System (IFS)69. The data is available through the Copernicus Climate Change Service (C3S)70. The closest data node to Fuvahmulah for both NCEP and ECMWF’s ERA5 are at 0.0° latitude, 73.5° longitude (see Fig. 1). The SRA data from GFZ were taken from the node at 0.0° latitude, 73.0° longitude (Table 1 juxtaposes the underlying models, key input parameters and output variables of each data set for this study).

To compare the data sets, this study correlates ocean wave time series from all three repositories with a cross-correlation function from the nodes closest to Fuvahmulah. The general solution for a cross-correlation function writes as:

$${C}_{AB}(\tau)=\frac{1}{N-m}\mathop{\sum }\limits_{N=0}^{N-m}(A\left(m{{\Delta }}t\right)-{\mu }_{A})(B(m{{\Delta }}t+\tau)-{\mu }_{B})$$
(1)

with τ = mΔt as the lag time for m sampling time increments and N the number of values in the correlated time series A and B, which are reduced by their respective means μA and μB. The normalized cross-correlation follows from considering the standard deviations σA and σB of each time series as

$${\rho }_{AB}(\tau)=\frac{{C}_{AB}(\tau)}{{\sigma }_{A}(\tau){\sigma }_{B}(\tau)}$$
(2)

The normalized cross-correlation values range from 1 for equal signals over 0 for uncorrelated signals to –1 as a reverse or negatively correlated signal. The corresponding lag time τ of the highest absolute value, here the correlation coefficient R of the cross-correlation, gives information about the time-shift between both signals.

The wave data in each repository contains time series of over 40 years. For the analysis, the time series is split up in annual sections from 1980 until 2019-05 and used separately in the cross-correlation function (Fig. 4b), resulting in 39 annual cross-correlation coefficients. When correlating the hourly data of CAWCR and ERA5 (sample size n = 345,504) with the 3-hourly NCEP time series (n = 115,168), the hourly time series were reduced to the associated 3-h values of the NCEP time series. To estimate the spatial variability of the data, a correlation of different spatial output nodes of the CAWCR data set (Fig. 4c) uses the entire CAWCR time series from 1980 to 2019 (n = 350,640).

This study further evaluates the impact of climate change on the wave climate according to the baseline chosen in the IPCC’s fifth assessment report47 by comparing the mean significant wave heights between the time periods 1986–2005 (historical; n = 204,467 undefined “NaN”, or “Not a Number”, values remain unconsidered) and 2081–2100 (end of the century) under RCP 4.5 and RCP 8.5 projections (n = 230,203 for both; undefined “NaN” values remain unconsidered). These projections are part of the CAWCR repositories71. The data is computed with the same tools as the hindcast data. Yet, the global grid is coarser with spatial increments of 1.0° ($${60}^{\prime}$$) and a temporal resolution of 6 hours. The RCP 4.5 and RCP 8.5 computations for wind waves use surface wind projections of 8 atmospheric models71.

As the CAWCR data set has the output node closest to Fuvahmulah, boundary conditions for the numerical models used in this study were derived from the CAWCR data node north-west of the island at 0.0° latitude, 73.2° longitude (see Fig. 1). The boundary conditions come from the entire, available hindcast data of significant wave heights Hs and associated peak directions θp between 1979 to 2020-02. The analysis of peak directions θp for each season introduces the range of peak directions θp,r. This range comes from a discrete probability distribution of the entire time-series and contains the directions, which combined have the highest 33% occurrence probability (see Fig. 5).

In general, the performance of global ocean wave models is good in the Indian Ocean when compared to buoy data64,72. The comparison to SRA measurements shows that wave data from global models occasionally underestimates wave heights, due to their relatively coarse spatio-temporal resolution in order of kilometers and hours. This improves with higher spatio-temporal resolutions73. Other studies have shown that underpredicted wave heights are especially prominent in extreme events74. In addition, non-linear effects can mitigate increased flood risk associated with sea-level rise75. Since 1992 and in contrast to sea-level rise, wave heights have only increased by about 0 to ~1 cmyr−1 for the Maldives according to a most recent study76 (see also Fig. 5f). In front of this background, wave heights under the respective RCP scenarios are conservative estimates of hazard levels, as the ocean wave models have a resolution of 1.0° and 6 hour time steps versus 0.4° to 0.5° spatial and hourly or 3 hour temporal resolution for the hindcast models. However, even when (1) combining today’s wave heights with a projected sea level rise of about 0.4 m to 0.8 m and (2) comparing it against the topographic height of Fuvahmulah with ~3.5 m above sea level, both factors will lead to more frequent extreme events. In this context, on Fuvahmulah, historically rare extreme events will become common by 2100 under all RCPs2,3.

### Numerical wave models

With the wave climate as boundary conditions, numerical wave models can reconstruct waves and wave-induced currents and thus give insights into sediment transport processes on the reef under different scenarios (for example with or without seaport infrastructure). This study uses two numerical models for different purposes. First, D3D helps to study the general sediment transport processes on the fringing reef platform under the characteristic wave climate of the region. The phase-averaging wave module of D3D uses wave spectra in its governing equations and combines them with sediment transport formulations77. To study the influence of the port infrastructure on the adjacent coastal areas, this study uses the depth-integrated (2DH) Boussinesq-type model BOSZ63. The phase-resolving wave model can directly capture wave transformation and secondary processes such as wave setup and recirculation over the reef. It is therefore suitable to study the current-induced processes in the study area in detail. While the computation of direct sediment transportation is challenging, the general wave-induced current velocities or shear stresses from wave-induced currents serve as the driving forces behind the most prominent bedload and suspended sediment transportation formulations77. Thus, acceleration, deceleration and presence of current fields are direct indicators for sediment accretion and sediment pick-up.

D3D has been used successfully for calculating sediment transport and demonstrated good performance in both experimental and hindcast studies78. The model is in general capable to outline the morphodynamic processes on the Fuvahmulah reef. To remain numerically stable and efficient in terms of computational resources, the settings in this study contain the following simplifications: the model uses a relatively coarse rectilinear grid off-shore, with a refined resolution of Δx = 67 m, Δy = 15 m around Fuvahmulah. To study the morphodynamic processes around the island, D3D uses both the WAVE and the FLOW module. The reef depth and island height are based on topographic and bathymetric surveys within the field campaigns, however the implementation in D3D is idealized: the off-shore section has a constant water depth of h = 100 m. The fringing reef has a constant depth of h = 4 m, which increases on the southern reef platform to h = 17.5 m. In the model, the entire island Fuvahmulah lies at 3 m above mean sea level (MSL). The off-shore reef south of the island has a similar shape in the model compared to the real reef. However, to provide enough cells to transport sediment, the virtual fringing reef in the model is wider than the physical reef around Fuvahmulah. To evaluate sediment transport pathways around the island, the reef was covered with a 1 m layer of erodible sediment at the start of each computation. The underlying assumption is that the reef platform is the main sediment supplier for the island5,13. While the simplifications allow for a larger set of computations, the implementation is still sufficiently detailed to capture the very general interaction between bathymetry, waves and sediment on the reef.

The following simulations in D3D use a PM-spectrum with Hs = 2.3 m, Tp = 17 s and a peak direction of θp = 202.5° and θp = 135° respectively. Despite the complexity of sediment transport on coral reefs79, for a first estimation, D3D runs only with the default (suspended and bed-load) sediment transport formulations77. All simulations use the same sediment properties: the sediment layer thickness is dsed = 1 m, sediment density is ρsed = 2.0 gcm−3 (based on in-situ measurements) and sediment grain size (diameter) is D = 200 μm. The morphological scale factor is fMOR = 2.

The reef bathymetry in the 2DH model comes from records of a single-beam, dual-frequency echo-sounder (Dr. Fahrentholz LituBox 15/200) taken during the first field campaign in 2017. The echo-sounder was able to show water depths up to ~ 80−100 m. However, as water depths increase rapidly beyond the reef plateau, records from the echo-sounder exceeding the maximum depth of the Fuvahmulah reef plateau at the southern tip of ~ 20−25 m were discarded. The minimum depth of the echo-sounder was 4 m due to the draft of the boat. The 2017 Photoscan model provides the nearshore (reef) bathymetry for water depths between mean low tide to mean sea level and the island’s coastal topography (including the coastal ridges). The Generic Mapping Tool80 interpolates the combined elevation and bathymetry data on a 7.5 m by 7.5 m grid, which was also used as a computation grid in the BOSZ model. Offshore waters are truncated to h = 60 m. Similar to the computations in D3D, the hydrodynamic boundary conditions are characterized by a TMA-wave spectrum with a significant wave height of Hs = 2.3 m, a peak period of Tp = 17 s and peak directions of θp = 202.5° and θp = 135°, respectively.

Since this study focuses on the wave driven longshore currents, ocean circulation around Fuvahmulah and the tidal current velocity was disregarded for sediment transport processes (for further information, see Supplementary Materials). This can be considered standard practice when regarding longshore sediment transport and confirms that longshore sediment transport mainly depends on incident waves, while tidal and ocean currents are of minor importance81.

Both numerical approaches reveal the impact of the regional wave climate on sediment transport and thus on erosion. However, some limitations apply to the results of the numerical models: the wave input is based on storm wave heights of two peak directions, while moderate seas also have an influence on sediment transport and beach restoration82. Furthermore, D3D is able to reconstruct experimental and hindcast data of sandy beaches78, but application on reef island morphodynamics is scarce83. For Fuvahmulah, D3D does not implement port structures in its bathymetry, to mimic the natural sediment pathways. The coarse resolution and the idealized bathymetry give a general concept of sediment movement on the reef rather than quantitative volumes of transported sediment. Still, the coarse model is able to reconstruct typical, process-based morphodynamic behavior for an environment like the Fuvahmulah reef5. The 2DH wave model comprises a more detailed study with and without the port structure. However, the results must be interpreted with caution: the 2DH approach calculates current velocities but does not yield transported sediment volumes. Erosion and sedimentation depend on more than only current velocities5. For example, research on organism scale shows that sediment transport on coral reef bottoms is complex79. However, current velocities are the main driver behind sediment transport phenomena and velocity gradients hint at hotspots of erosion and sedimentation77. With this in mind, the computed wave-induced currents are able to illustrate the role of the southeastern reef area under the given wave climate for the sediment transport around Fuvahmulah. They also facilitate a process-based understanding of the seaport’s impact on the adjacent coastline, even without increased sea level or varied wave heights.

### Societal aspects

The analysis of the societal dynamics is based on a mixed-methods approach to convey the social and governance aspects influencing coastal protection and climate change related topics. The approach encompasses (1) a literature review and content analysis of climate change adaptation and coastal protection literature in the Maldives, (2) two surveys with the local population, which are complemented by (3) semi-structured interviews with relevant actors involved in coastal management on the local as well as on the national-level. All participants gave their voluntary consent to participate and have been informed about the objective of the research. The collected data of the second survey and the interviews were analyzed via a structuring content analysis in the software program MAXQDA84. This content analysis categorizes the interview’s responses and those of open questions in the surveys into codes and subcodes, developed in a concept-driven and data-driven approach85. The concept-driven or deductive approach follows a categorization structure established before the field campaign. However, the possibility to respond to open-ended questions also requires post-processing of the responses based on their content – resulting in a data-driven approach.

The first survey and interviews took place from March to April 2017 and were synchronized with the first field campaign, measuring topographic and hydrodynamic data. Intermediate results from both work packages and the associated discussions led to interdisciplinary insights, considered in the second survey (January to February 2019).

### Literature review

The literature review was conducted by content analysis. The analysis focused on publications relevant to climate change adaptation and coastal protection in the Maldives, including peer-reviewed papers, legislation and regulations, as well as gray literature. Legislation and regulations were primarily identified with the support of actors in the Maldives coastal governance system through the semi-structured interviews, ensuring their relevance and completeness. The literature review also encompasses previous EIAs on coastal projects in the Maldives, as these are required to list the legislations and regulations that are applicable to the projects. Most legislation and regulations are available in an English version on the website of the respective ministry, however, documents in the local language Dhivehi have not been considered. The analyzed gray literature covered EIAs, available through the Maldives Environmental Protection Agency (EPA), as well as reports from the Government of the Maldives, Non-governmental Organizations (NGOs) and intergovernmental organizations, such as the United Nations Development Programme (UNDP) Maldives. Special attention was paid to reports, EIAs, and ESIAs concerning Fuvahmulah’s seaport or explicit climate change adaptation measures on the island (see Table 2). The documents were analyzed via a qualitative content analysis85, which focused on the following thematic codes or categories in the context of the coastal governance system in the Maldives: (1) distribution of responsibilities among actors in the context of coastal protection, (2) history of coastal protection measures, (3) framing of coastal protection in the climate change discussion and (4) framing of coastal protection in the climate change discussion framing of coastal protection in the climate change discussion. The analysis disclosed knowledge gaps in the literature and demonstrated the need for an exploratory and qualitative research approach. Therefore, the study also contains population surveys and semi-structured interviews.

### Survey

Two surveys with semi-standardized questionnaires were conducted in Fuvahmulah. The first survey focused on four topics: (1) the local’s understanding of the environment and the coast, (2) their perception of environmental issues with a focus on climate change impacts and comprehension of responsibilities for action, (3) the locals’ perspectives on community life, and (4) their perspectives and attitudes towards coastal protection. The second survey focused on: (1) the locals’ perceptions on community life, (2) their potential for engagement in environmental and coastal protection activities, and (3) how politics shape the way coastal issues are addressed.

The surveys involved going from door-to-door in all eight wards of the island and were supported by local research assistants. The sample size of each survey is equivalent to about every eighth household on the island, while household selection was done by previous randomized sampling, yielding an unbiased representation of local households. Interviewees were not working in local coastal management. The first survey had 116 participants between the age of 14 and 86, while the second survey was answered by 98 inhabitants of Fuvahmulah between 14 and 84 years old. The lower threshold of 14 years results from the fact that climate change is a long-term development and coastal protection measures will have an impact for several generations. As in social science studies on future societal development, we have accordingly extended the age limit to 14 years. Furthermore, this decision is based on the experiences made on the island and made in consultation with the local research assistants – people over 14 years of age start to participate in environmental activities without the guidance of their parents and are therefore considered to be sensitive to the environmental situation of Fuvahmulah. With this in mind, and to avoid bias86, adolescents are considered capable of giving consent to answer the questionnaire themselves. As participants were not divided into experimental groups, the responses of all participants in the survey remain anonymous. However, to further respect their privacy, this minimal-risk study does not include direct verbatim statements from participants aged 14–18.

The questionnaire of each survey contained closed and open questions (see supplementary materials). For closed questions, the questionnaire provides predefined answers, which were analyzed with methods of descriptive analytics. In contrast to closed questions, open questions require the participant to respond freely to the question. The answers were categorized and coded85 after the survey to quantify the results and identify important response patterns. In both surveys, the individual interview time varied strongly, ranging from 20 to 90 min. This can be explained by the qualitative approach of the surveys, utilizing open questions and the varying length and depth of answers which depended, for example, on the interviewee’s interest in the topics.

### Interviews

In each field trip, the interviews of Fuvahmulah’s population were complemented by semi-structured interviews with actors on the national and local level being involved in the management of coastal issues in the Maldives. The selection of interview partners is based on their profession and their expertize on the subject (purposive sampling). Using this non-random technique was necessary due to the limited number of actors in coastal governance, while obtaining the required information to compare the varying opinions and interests among those actors regarding coastal protection.

On the national-level, interview partners were representatives of the Ministry of Environment and Energy (renamed to Ministry of Environment in 2018), members of non-governmental and intergovernmental organizations that are active in the climate change adaptation and coastal management fields, as well as local researchers. Interviews with non-government actors gave in-depth and critical perspectives on the way environmental and coastal issues are addressed by government entities. This information is especially important as it helps to identify informal institutions, such as norms and attitudes towards specific responses to coastal problems that become effective and influence the way coastal problems are addressed.

On the local level, interview partners are current and former members of Fuvahmulah’s council and representatives of non-governmental organizations.

In summary, the results of this study are based on 32 interviews with 18 interviews with actors on the national-level (10 government, 8 non-governmental or intergovernmental organizations and researchers) and 14 on the local level (8 government, 6 non-governmental organizations). The interviews of the first field trip addressed (1) societal dealings with environmental problems, (2) coastal management with focus on decision-making processes, (3) community involvement and public awareness, and (4) politics and climate change. The second interview guide focused on (1) the community in the Maldives, (2) community engagement, and (3) coastal governance. The interviews lasted between 30 and 120 minutes. Owing to the political situation in the Maldives and the small number of experts in this field, not all interviewees agreed to be recorded, but memos and notes capture the interview’s statements. Recorded interviews were transcribed before analysis. In addition, the interviews were fully anonymized – except for the ministry officials – in order to protect the identity of the informants.

### Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

## Data availability

The field data87 from the aerial surveys and bathymetry for numerical modeling is publicly available through https://doi.org/10.5281/zenodo.4304049. Wave climate data is publicly available from the respective services and homepages of CAWCR (CSIRO Data Acess Portal: https://data.csiro.au/collections/collection/CIcsiro:39819), ECMWF (Copernicus Climate Data Store: https://cds.climate.copernicus.eu) and NOAA (WAVEWATCH III®Hindcast and Reanalysis Archives: https://polar.ncep.noaa.gov/waves/hindcasts/). For further information of the climate data, see also Table 1. Source data supporting the social sciences’ finding are provided within the Supplementary Information File of this paper. Other raw data (for example interview transcripts) on the social sciences part that support the findings of this study are protected and not publicly available as they contain personal information that could compromise research participant privacy. Other data from the social science part can be made available from the social science team (B.M.W.R. and A.H.) upon reasonable request.

## Code availability

The aerial images were processed with the Structure-from-Motion MultiView Stereo (SfM-MVS) algorithm of Agisoft Photscan Pro (version 1.4.5., build 7354). Except for the bathymetry of the numerical model, geo-data was processed with Quantum GIS 3.10 A Coruña and 3.16 Hannover on Ubuntu Linux 18.04 and 20.04 and/or Python modules rasterio v.1.1.1. to v.1.1.3. together with the module Fiona v.1.8.13. The Boussinesq Ocean and Surf Zone (BOSZ, version 01-2019) model is available through Volker Roeber upon reasonable request. The Generic Mapping Tool (GMT, version 5.4.5) was used to process the geo-data to be used as bathymetry in BOSZ. Other simulations on the morphodynamics around the reef island was carried out with Delft3D. Delft3D is open-source and available through Deltares (oss.deltares.nl/web/delft3d). In the Delft 3D Suite, we used the modules Deltares FLOWD3D 6.02.13.9162M, SWAN III 40.72ABCDE. Other scripts for data post processing and visualization, are written with the open-source software Python 3.7 in Jupyter Notebooks, using the modules Numpy 1.17.4, Matplotlib (pyplot) 3.1.2, Xarray 0.12.1 with NetCDF 1.5.3, Pandas 0.25.3, cartopy 0.18.0 and cdsapi 0.1.3 and later versions. Cartopy uses OpenStreetMap data (available under the Open Database License, see openstreetmap.org) as well as Stamen open-source maps. Data of the second household survey and interviews were analyzed with the software MAXQDA.

## References

1. Nurse, L. A. et al. in Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel of Climate Change, (eds. Barros, V. R. et al.) 1613–1654 (Cambridge University Press, 2014).

2. Magnan, A. et al. Integrative cross-chapter box on low-lying islands and coasts. (eds. Pörtner, H.-O. et al.) Special Report on Ocean and Cryosphere in a Changing Climate (In press, 2019).

3. Oppenheimer, M. et al. Sea level rise and implications for low lying islands, coasts and communities. (eds. Pörtner, H.-O. et al.) IPCC Special Report on the Ocean and Cryosphere in a Changing Climate (In press, 2019).

4. Storlazzi, C. D., Elias, E. P. & Berkowitz, P. Many atolls may be uninhabitable within decades due to climate change. Sci Report 5, https://doi.org/10.1038/srep14546 (2015).

5. David, C. G. & Schlurmann, T. Hydrodynamic drivers and morphological responses on small coral islands-the Thoondu spit on Fuvahmulah, the Maldives. Front. Marine Sci. 7, 885 (2020).

6. East, H. K., Perry, C. T., Beetham, E. P., Kench, P. S. & Liang, Y. Modelling reef hydrodynamics and sediment mobility under sea level rise in atoll reef island systems. Glob. Planet. Change 192, 103196 (2020).

7. Kench, P. S. & Brander, R. W. Response of reef island shorelines to seasonal climate oscillations: South Maalhosmadulu atoll, Maldives. J. Geophys. Res.: Earth Surface 111, (2006). agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2005JF000323https://doi.org/10.1029/2005JF000323.

8. Kench, P. S., Brander, R. W., Parnell, K. E. & McLean, R. F. Wave energy gradients across a Maldivian atoll: Implications for island geomorphology. Geomorphology 81, 1–17 (2006).

9. Kench, P. S. & Mann, T. Reef island evolution and dynamics: insights from the indian and pacific oceans and perspectives for the spermonde Archipelago. Front. Marine Sci. 4, 145 (2017).

10. Wadey, M., Brown, S., Nicholls, R. J. & Haigh, I. Coastal flooding in the Maldives: an assessment of historic events and their implications. Nat. Hazard. 89, 131–159 (2017).

11. East, H. K., Perry, C. T., Kench, P. S., Liang, Y. & Gulliver, P. Coral reef island initiation and development under higher than present sea levels. Geophys. Res. Lett. 45, 11265–11274 (2018).

12. Masselink, G., Beetham, E. & Kench, P. Coral reef islands can accrete vertically in response to sea level rise. Sci. Adv. 6, eaay3656 (2020).

13. Ryan, E. J., Hanmer, K. & Kench, P. S. Massive corals maintain a positive carbonate budget of a Maldivian upper reef platform despite major bleaching event. Sci. Rep. 9, (2019).

14. Tuck, M. E., Ford, M. R., Kench, P. S. & Masselink, G. Sediment supply dampens the erosive effects of sea-level rise on reef islands. Sci. Rep. 11, 6515 (2021).

15. Beck, M. W. et al. The global flood protection savings provided by coral reefs. Nat. Commun. 9, https://doi.org/10.1038/s41467-018-04568-z (2018).

16. Ferrario, F. et al. The effectiveness of coral reefs for coastal hazard risk reduction and adaptation. Nat. Commun. 5, https://doi.org/10.1038/ncomms4794 (2014).

17. Beetham, E., Kench, P. S. & Popinet, S. Future reef growth can mitigate physical impacts of sea-level rise on atoll islands. Earth’s Future 5, 1002–1014 (2017).

18. Hughes, T. P. et al. Coral reefs in the Anthropocene. Nature 546, 82–90 (2017).

19. Moustahfid, H., Marsac, F. & Gangopadhyay, A. Impacts of Climate Change on Fisheries and Aquaculture–synthesis of Current Knowledge, Adaptation and Mitigation Options, Vol. 627, Technical Paperclimate Change Impacts, Vulnerabilities and Adaptations: Western Indian Ocean Marine Fisheries (Fisheries and Aquaculture Department of the Food and Agriculture Organization of the United Nations, Rome, Italy, 2018). ISSN 2070-7010.

20. Kench, P. S. Compromising Reef Island Shoreline Dynamics: Legacies of the Engineering Paradigm in the Maldives. (eds. Cooper, J. A. G. & Pilkey, O. H.) Pitfalls of Shoreline Stabilization: Selected Case Studies 165–186, (Springer Netherlands, Dordrecht, 2012) https://doi.org/10.1007/978-94-007-4123-2_11.

21. Luijendijk, A. et al. The state of the world’s beaches. Sci. Rep. 8, https://doi.org/10.1038/s41598-018-24630-6 (2018).

22. Duvat, V. K. E. & Magnan, A. K. Rapid human-driven undermining of atoll island capacity to adjust to ocean climate-related pressures. Sci. Rep. 9, https://doi.org/10.1038/s41598-019-51468-3 (2019).

23. Patrício, J., Elliott, M., Mazik, K., Papadopoulou, K.-N. & Smith, C. J. DPSIR–two decades of trying to develop a unifying framework for marine environmental management? Front. Marine Sci. 3, 177 (2016).

24. IPCC. The IPCC Response Strategies. (eds. Bernthal, F. et al.) First Assessment Report (FAR), 269 (Island Press, Cambridge, United Kingdom and New York, NY, USA, 1990).

25. Ratter, B. M., Petzold, J. & Sinane, K. Considering the locals: coastal construction and destruction in times of climate change on Anjouan, Comoros. Nat. Resource. Forum 40, 112–126 (2016).

26. Magnan, A. K. & Duvat, V. K. E. Towards adaptation pathways for atoll islands. Insights from the Maldives. Reg. Environ. Change 20, https://doi.org/10.1007/s10113-020-01691-w (2020).

27. Petzold, J. & Magnan, A. K. Climate change: thinking small islands beyond Small Island Developing States (SIDS). Climatic Change 152, 145–165 (2019).

28. Petzold, J., Ratter, B. M. W. & Holdschlag, A. Competing knowledge systems and adaptability to sea-level rise in The Bahamas. Area 50, 91–100 (2018).

29. Hamin, E. et al. Pathways to coastal resiliency: The adaptive gradients framework. Sustainability 10, 2629 (2018).

30. Cai, W. et al. Increased frequency of extreme indian ocean dipole events due to greenhouse warming. Nature 510, 254–258 (2014).

31. Dangendorf, S. et al. Persistent acceleration in global sea-level rise since the 1960s. Nat. Climate Change 9, 705–710 (2019).

32. Nunn, P. D., Klöck, C. & Duvat, V. Seawalls as maladaptations along island coasts. Ocean Coastal Management 205, 105554 (2021).

33. Ratter, B., Hennig, A. & Zahid. Challenges for shared responsibility–Political and social framing of coastal protection transformation in the Maldives. DIE ERDE-J. Geogr. Soc. Berlin 150, 169–183 (2019).

34. MEE & UNDP. Survey of Climate Change Adaptation Measures in Maldives-Integration of Climate Change Risks into Resilient Island Planning in the Maldives Project. Final Report, Ministry of Housing and Environment, the Maldives (MEE), United Nations Development Programme (UNDP). Author: Ahmed Shaig (2011).

35. IPCC. Climate Change 2014: Impacts, Adaptation, and Vulnerability (Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2015).

36. Noble, I. et al. Adaptation needs and options. (eds. Field, C. et al.) Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, vol. AR 5, 833–868. The Intergovernmental Panel on Climate Change (Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2015).

37. Duvat, V. K. Human-driven atoll island expansion in the Maldives. Anthropocene 32, 100265 (2020).

38. Gussmann, G. & Hinkel, J. What drives relocation policies in the Maldives? Clim. Change 163, 931–951 (2020).

39. Zubair, S., Bowen, D. & Elwin, J. Not quite paradise: Inadequacies of environmental impact assessment in the Maldives. Tour. Manage. 32, 225–234 (2011).

40. Gussmann, G. & Hinkel, J. A framework for assessing the potential effectiveness of adaptation policies: Coastal risks and sea-level rise in the Maldives. Environ. Sci. Policy 115, 35–42 (2021).

41. NBS. Maldives Population and Housing Census. Summary, National Bureau of Statistics (NBS), Ministry of Finance and Treasury, Male’ 20379, Republic of Maldives (2014).

42. Trigo, R. M. & Gimeno, L. Weather Pattern Changes in the Tropics and Mid-Latitudes as an Indicator of Global Changes. In Letcher, T. M. (ed.) Climate Change, chap. 8, 165–180, (Elsevier, Amsterdam, 2009) https://doi.org/10.1016/B978-0-444-53301-2.00008-7.

43. MEE. Environment & Social Assessment & Management Framework Fuvahmulah. Climate Resilience and Environment Sustainability Project. Final Report, (2014). Project carried out under the Climate Change Trust Fond, the World Bank; last accessed Dec. 27 through documents.worldbank.org.

44. Ministry of Environment, E. & Water. Foahmulaku Beach Erosion Survey & Coastal Protection Report. Field Survey Report, (2006).

45. Schöne, T., Esselborn, S., Rudenko, S. & Raimondo, J.-C. Radar altimetry derived sea level anomalies–the benefit of new orbits and harmonization. (eds. Flechtner, F. M. et al.) System Earth via Geodetic-Geophysical Space Techniques, 317–324, (Springer Berlin Heidelberg, Berlin, Heidelberg, 2010) https://doi.org/10.1007/978-3-642-10228-8_25.

46. Hemer, M. A., Katzfey, J. & Trenham, C. E. Global dynamical projections of surface ocean wave climate for a future high greenhouse gas emission scenario. Ocean Modell. 70, 221 – 245 (2013).

47. IPCC. Summary for policymakers. (eds. Stocker, T. et al.) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 1–30, (Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2013) https://doi.org/10.1017/CBO9781107415324.004.

48. Hemer, M. A., Fan, Y., Mori, N., Semedo, A. & Wang, X. L. Projected changes in wave climate from a multi-model ensemble. Nat. Clim. Change 3, 471–476 (2013).

49. Malatesta, S. & Schmidt, M. Environmental policy and climate change vulnerability in the Maldives: from the ‘lexicon of risk’ to social response to change. Island Stud. J. 12, 53–70 (2017).

50. DHV, R. H. Coastal protection at Gn. Fuvahmulah. Deliverable–2e. Technical Design, Amersfoort, the Netherlands (2016).

51. Energy, M. & (MEECO), E. C. Environmental Impact Assessment Report For Coastal Protection at Gn. Fuvahmulah, Maldives. Technical Report, Male’, Republic of Maldives (2016). In Association with Royal HaskoningDHV.

52. CERC. Shore Protection Manual, vol. 2 (Coastal Engineering Research Center (CERC) Dept. of the U.S. Army, Waterways Experiment Station, Corps of Engineers, Coastal Engineering Research Center, 1984).

53. USACE. Coastal Engineering Manual (EM 1110-2-1100), vol. 1–6 (U.S. Army Corps of Engineers (USACE), Washington D.C., United States of America, 2002).

54. Perry, C. T. et al. Caribbean-wide decline in carbonate production threatens coral reef growth. Nat. Commun. 4, https://doi.org/10.1038/ncomms2409 (2013).

55. Duvat, V. K. et al. Trajectories of exposure and vulnerability of small islands to climate change. WIREs Climate Change 8, e478 (2017).

56. Bisaro, A. & Hinkel, J. Governance of social dilemmas in climate change adaptation. Nat. Clim. Change 6, 354–359 (2016).

57. David, C. G., Schlurmann, T. & Roeber, V. Coastal Infrastructure on Reef Islands–the Port of Fuvahmulah, the Maldives as Example of Maladaptation to Sea-Level Rise? (eds. Goseberg, N. & Schlurmann, T.) Coastal Structures 2019, 874–885, (Bundesanstalt für Wasserbau, Hannover, Germany, 2019) https://doi.org/10.18451/978-3-939230-64-9_087.

58. van Slobbe, E. et al. Building with Nature: in search of resilient storm surge protection strategies. Nat. Hazard. 66, 1461–1480 (2013).

59. Haasnoot, M., Kwakkel, J. H., Walker, W. E. & ter Maat, J. Dynamic adaptive policy pathways: a method for crafting robust decisions for a deeply uncertain world. Glob. Environ. Change 23, 485–498 (2013).

60. Westoby, M., Brasington, J., Glasser, N., Hambrey, M. & Reynolds, J. ‘Structure-from-Motion’ photogrammetry: a low-cost, effective tool for geoscience applications. Geomorphology 179, 300–314 (2012).

61. David, C. G. et al. Structure-from-motion on shallow reefs and beaches: potential and limitations of consumer-grade drones to reconstruct topography and bathymetry. Coral Reefs 40, 835–851 (2021).

62. Casella, E., Drechsel, J., Winter, C., Benninghoff, M. & Rovere, A. Accuracy of sand beach topography surveying by drones and photogrammetry. Geo-Marine Lett. 40, 255–268 (2020).

63. Roeber, V. & Bricker, J. D. Destructive tsunami-like wave generated by surf beat over a coral reef during typhoon haiyan. Nat. Commun. 6, https://doi.org/10.1038/ncomms8854 (2015).

64. Bruno, M. F., Molfetta, M. G., Totaro, V. & Mossa, M. Performance assessment of ERA5 wave data in a swell dominated region. J. Marine Sci. Eng. 8, 214 (2020).

65. Durrant, T., Hemer, M., Smith, G., Trenham, C. & Greenslade, D. CAWCR Wave Hindcast - Aggregated Collection. v1. Last accessed: 2020-03-29, permalink: hdl.handle.net/102.100.100/137152?index=1 (2019).

66. Tolman, H. L. User Manual and System Documentation of WAVEWATCH III version 3.14. Technical Note, (NOAA-NCEP, 1999).

67. Saha, S. et al. The NCEP Climate Forecast System Reanalysis. Bulletin of the American Meteorological Society 91, 1015–1058 (2010).

68. WAMDIG. The WAM Model–A Third Generation Ocean Wave Prediction Model. J. Phys. Oceanogr. 18, 1775–1810, (1988) https://doi.org/10.1175/1520-0485 (1988)018 < 1775:TWMTGO > 2.0.CO;2.

69. ECMWF. ECMWF Wave Model. In IFS Documentation - CY41R2, chap. VII, 1–83 (European Centre for Medium-Range Weather Forecasts (ECMWF), Shinfield Park, Reading, RG2 9AX, England, 2016).

70. Copernicus Climate Change Service. ERA5: Fifth Generation of ECMWF Atmospheric Reanalyses of the Global Climate. Copernicus Climate Change Service Climate Data Store (CDS). Last date of access: 2019-12-06 https://doi.org/10.24381/cds.adbb2d47 (2017).

71. Hemer, M., Trenham, C., Durrant, T. & Greenslade, D. CAWCR Global wind-wave 21st century climate projections. v1. Last accessed: 2020-03-29, https://doi.org/10.4225/08/55C991CC3F0E8 (2015).

72. Muhammed Naseef, T. & Sanil Kumar, V. Climatology and trends of the Indian Ocean surface waves based on 39-year long ERA5 reanalysis data. Int. J. Climatol. 40, 979–1006 (2020).

73. Wiese, A. et al. Synergy of wind wave model simulations and satellite observations during extreme events. Ocean Sci. 14, 1503–1521 (2018).

74. Hildebrandt, A., Schmidt, B. & Marx, S. Wind-wave misalignment and a combination method for direction-dependent extreme incidents. Ocean Eng. 180, 10–22 (2019).

75. Arns, A. et al. Non-linear interaction modulates global extreme sea levels, coastal flood exposure, and impacts. Nat. Commun. 11, https://doi.org/10.1038/s41467-020-15752-5 (2020).

76. Timmermans, B. W., Gommenginger, C. P., Dodet, G. & Bidlot, J.-R. Global wave height trends and variability from new multimission satellite altimeter products, reanalyses, and wave buoys. Geophys. Res. Lett. 47, e2019GL086880 (2020).

77. van Rijn, L. C. Principles of Sedimentation and Erosion Engineering in Rivers, Estuaries and Coastal Seas. (Aqua Publications, Amsterdam, The Netherlands, 2005).

78. Lowe, R. J., Falter, J. L., Monismith, S. G. & Atkinson, M. J. A numerical study of circulation in a coastal reef-lagoon system. J Geophys. Res.: Oceans 114. https://doi.org/10.1029/2008JC005081agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2008JC005081 (2009).

79. Pomeroy, A. W. M. et al. Sediment transport in the presence of large reef bottom roughness. J. Geophys. Res.: Oceans 122, 1347–1368 (2017).

80. Wessel, P. & Luis, J. F. The GMT/MATLAB Toolbox. Geochem. Geophys. Geosyst. 18, 811–823 (2017).

81. Burcharth, H. F., Hawkins, S. J., Zanuttigh, B. & Lamberti, A. in Environmental Design Guidelines for Low Crested Coastal Structures (eds. Burcharth, H. F., Hawkins, S. J., Zanuttigh, B. & Lamberti, A.) 203–333, (Elsevier Science Ltd, 2007).

82. Barnard, P. L. et al. Extreme oceanographic forcing and coastal response due to the 2015–2016 El Niño. Nat. Commun. 8, https://doi.org/10.1038/ncomms14365 (2017).

83. Shope, J. B. & Storlazzi, C. D. Assessing morphologic controls on atoll island alongshore sediment transport gradients due to future sea-level rise. Front. Marine Sci. 6, 245 (2019).

84. Kuckartz, U & Rädiker, S. Analyzing Qualitative Data with MAXQDA (Springer, 2019).

85. Kuckartz, U. Qualitative Text Analysis: A Systematic Approach. In Kaiser, G. & Presmeg, N. (eds.) Compendium for Early Career Researchers in Mathematics Education, 181-197, (Springer International Publishing, Cham, 2019) https://doi.org/10.1007/978-3-030-15636-7_8.

86. Liu, C., Cox, R. B., Washburn, I. J., Croff, J. M. & Crethar, H. C. The effects of requiring parental consent for research on adolescents’ risk behaviors: A meta-analysis. Journal of Adolescent Health 61, 45–52 (2017).

87. David, C. G., Ballesteros, P. & Schlurmann, T. Coastal Digital Elevation Models and Transects of the Reef Island Fuvahmulah, the Maldives [Dataset]. Zenodo (2020) https://doi.org/10.5281/zenodo.4304049.

## Acknowledgements

This study took place in the project “Dealing with change in SIDS: societal action and political reaction in sea level change adaptation in Small Island Developing States (DICES)”, grant no. SCHL 503/17-1 (C. Gabriel David, Torsten Schlurmann) and RA 585/19-1 (Arne Hennig, Beate M. W. Ratter). The project is framed within the priority programm (SPP-1889) – regional sea level change and society of the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG). Volker Roeber acknowledges financial support from the Isite program Energy Environment Solutions (E2S), the Communauté d’Agglomération Pays Basque (CAPB) and the Communauté Région Nouvelle Aquitaine (CRNA) for the chair position HPC-Waves; as well as the support from the University of Hawai’i at Mānoa for the Affiliate Graduate Faculty position. Zahid was employed as Deputy Director General Climatology at the Maldives Meteorological Service (MMS) at the time of the project, but retired from his position. The authors would like to thank Ali Ahmed, Pablo Ballesteros, Tatiana Ivanova, René Klein, Nina Kohl, Manò Schütt, Ibrahim Shiyan (Panda), Jailam Zahir as well as Marion and Uwe Zander for their help in the field campaigns. In the Maldives, the authors were supported by the Maldives Meteorological Service (MMS), Fuvahmulah Island Council and Fuvahmulah DIVE School. Pablo Ballesteros and René Klein are staff of Ludwig-Franzius-Institute, Leibniz University of Hanover (LuFI-LUH) and supported the field campaign with their technical support. Nina Kohl and Manò Schütt were student assistants at LuFI, helping to record data on Fuvahmulah. Ali Ahmed and Jailam Zahir are locals from Fuvahmulah and assisted the in the coastal surveys (Ali) and in the interviews and household survey (Jailam). The authors also thank Tilo Schöne of GFZ Potsdam, providing the SRA data as well as Jean Bidlot (ECMWF), Mark Hemer (Australian Commonwealth Scientific and Industrial Research Organisation, CSIRO) and Todd Spindler (NCEP-EMC) for their help in accessing climate reanalysis data. Also, Elisa Casella and Alessio Rovere gave valuable insights and feedback to UAV-based photogrammetry, while Tobias Kersten helped with geodetic questions before the third field campaign and with post-processing the GNSS data. Furthermore, the authors acknowledge Jannek Gundlach for his feedback on and Astrid Kartes and Jonas Briese for their assistance with Delft3D. Both, Astrid and Jonas, were students at LuFI during the project.

## Funding

Open Access funding enabled and organized by Projekt DEAL. This study took place within the German Research Foundation’s (DFG) priority programm “regional sea level” (SPP-1889).

## Author information

Authors

### Contributions

C.G.D. conducted the natural science/engineering part of field campaigns, research and analysis, while A.H. carried out the corresponding social science part. V.R. developed the BOSZ model used in this study and supervised the numerical modeling. T.S. and B.R. (co-)designed the research project, were responsible for funding resources and reporting and provided guidance throughout the entire research. C.G.D. wrote the manuscript with input from A.H., while B.R., V.R., T.S. and Zahid edited and contributed to the final manuscript. Zahid helped in designing the research project, coordinating the field surveys and provided a local perspective to the evaluation and analysis.

### Corresponding author

Correspondence to C. Gabriel David.

## Ethics declarations

### Competing interests

The authors declare no competing interests.

### Ethics statement

All participants gave their informed consent and we complied with all relevant ethical regulations. Ethical affairs have been handled and declared during the funding phase of the project. Ethics committee and institutional review board approval was determined not to be required for this project.

Peer review informationNature Communications thanks Paula Camus and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Rights and permissions

Reprints and Permissions

David, C.G., Hennig, A., Ratter, B.M.W. et al. Considering socio-political framings when analyzing coastal climate change effects can prevent maldevelopment on small islands. Nat Commun 12, 5882 (2021). https://doi.org/10.1038/s41467-021-26082-5

• Accepted:

• Published:

• DOI: https://doi.org/10.1038/s41467-021-26082-5