A coastal exposure index for Ireland: relative hazard exposure and the protective role of coastal habitats Authors: Kevin Walsh, Larissa Macedo Cruz de Oliveira, Jennifer Keenahan, Aaron Lim Published: 2026 ABSTRACT The concurrent rise of global sea levels and coastal population growth is intensifying exposure to climate-driven hazards. In Ireland, where most major urban centres are on, or connected to the coast, the nature and severity of hazard impacts vary across diverse coastal landscapes. A spatially explicit assessment of coastal exposure is therefore critical for effective risk management and climate adaptation planning. This study presents a national-scale Coastal Exposure Index for Ireland, developed using the Coastal Vulnerability Model within the InVEST® software suite. The index integrates physical and climatic variables, including 30 years of wind and wave records, sea-level rise rates, and storm surge levels. Coastal habitats, including dunes, machair, saltmarsh and seagrasses, were directly represented to assess their potentially protective role. Exposure levels of coastal populations and cultural heritage sites were also evaluated. A critical appraisal of the InVEST® approach was undertaken, including a sensitivity analysis to assess how model modifications affect outcomes. Approximately 8575 km of coastline was assessed at an alongshore spacing of 250 m, with exposure reported on a relative scale. Results reveal that 38% of the coast benefits from potential exposure reduction provided by coastal habitats. The west coast exhibits the highest relative exposure values and contains the largest extent of “Very High Exposure” coastline. The model equally identifies exposure distribution along the east coast, with several eastern local authorities having some of the highest proportions of “Very High Exposure” coastline relative to their overall length. Coastal populations within each exposure category were estimated at multiple distances from the coast, revealing almost 10,000 residents in both Dublin City and Galway City live within 500 m of a “Very High Exposure” coastline. Cultural heritage sites were similarly mapped across categories and distances, with several exposure hotspots identified along the west and northwest coasts. By assessing the potentially protective role of coastal habitats, this study supports the integration of nature-based solutions in into coastal zone management and provides a national-scale screening assessment of relative exposure to guide evidence-based adaptation strategies.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s11069-026-08361-w. Introduction Coastal zones are among the most densely populated and rapidly urbanising environments on Earth, a trajectory that has left only 15.5% of the world’s coastal regions in a natural intact state (Cosby et al. 2024; Neumann et al. 2015; Williams et al. 2022). Coastal populations face escalating risks from sea level rise (SLR), which is accelerating due to human-induced warming, driving glacial melting and thermal expansion of ocean waters (IPCC 2023). This risk is compounded in certain urban coastal areas which are experiencing higher levels of relative SLR due to subsidence, often caused by groundwater or hydrocarbon extraction (Nicholls et al. 2021). Under high-emission scenarios, global mean sea levels are projected to rise 1.1 m above 1900 levels by 2100 (Oppenheimer et al. 2019), and without substantive adaptation and mitigation strategies, coastal hazards will become more frequent and severe. The fifth assessment report by the Intergovernmental Panel on Climate Change (IPCC 2014) outlines a conceptual framework for understanding how to assess and manage climate risks. Risk is defined as the likelihood of an occurrence of a hazard, and is a function of the hazard itself, the exposure, and the vulnerability (Field et al. 2014). Exposure refers to the places and things which may be adversely affected by climate hazards, while vulnerability refers to the propensity to be adversely affected, encompassing the sensitivity to the hazard and the capacity to adapt (Field et al. 2014; Rocha et al. 2023). Numerous strategies to reduce coastal risk exist, which have traditionally relied on “hard”, engineered solutions such as dikes or seawalls for necessary coastal protection. Although vital in certain areas, these structures are often short-term and expensive interventions. They can also lead to unintended environmental consequences, such as beach narrowing and habitat degradation (Dugan et al. 2008; Silver et al. 2019). Alternative approaches to coastal protection involve employing nature-based solutions (NbS) defined as “actions to protect, sustainably manage and restore natural and modified ecosystems that address societal challenges effectively and adaptively, simultaneously providing human well-being and biodiversity benefits” (Cohen-Shacham et al. 2016, p. 2). Coastal ecosystems, for example, sand dunes, kelp forests and saltmarshes, deliver a wide range of services to humanity, including food provision, water purification, and opportunities for recreation and emotional well-being (Millennium Ecosystem Assessment 2005). They also act as natural buffers, often serving as the “first line of defence” against powerful waves, winds and storm surges (Hanley et al. 2014; Spalding et al. 2014). As such, they represent a clear example of NbS for coastal protection. Various methodologies for assessing vulnerability to climate risks exist, with index-based approaches commonly employed for national scale assessments. The Coastal Vulnerability Index (CVI) was first conceptualised by Gornitz (1991) to assess vulnerability to SLR along coastline of the United States of America, and has since been adapted to numerous coastlines globally at varying scales (e.g. Bagdanavičiūtė et al. 2015; Hammar-Klose and Thieler 2001; López Royo et al. 2016; Mclaughlin and Cooper 2010; Pendleton et al. 2010; Rocha et al. 2020). The earliest studies focused on the physical aspects of the coast, such as the geomorphology, relief, rock type, and the forcing mechanisms driving change, such as waves and SLR (Rocha et al. 2023). The concept of coastal vulnerability has since evolved, with contemporary approaches recognising the importance of incorporating not only the physical characteristics of the coast but also the socioeconomic factors that influence the sensitivity, and adaptive capacity of coastal populations and infrastructure (Bevacqua et al. 2018; Kantamaneni et al. 2018; Rocha et al. 2023; VanZomeren and Acevedo-Mackey 2019). This study aims to produce a detailed, national map of the present day, relative exposure to climate hazards around the coastline of the Republic of Ireland (ROI). The potential for coastal habitats to reduce exposure is emphasised through a “no-habitat” scenario, highlighting where NbS for coastal protection may be most beneficial. The results show the distribution of relative exposure levels, which when overlain with population and national monument maps, indicate the total number of residents and sites of cultural heritage in each exposure category. Results are presented for each Local Authority (LA) along the ROI coastline, as these are the statutory bodies responsible for coastal management implementation in Ireland. To the best of the authors’ knowledge, this study presents the first unified, nationwide assessment of the relative exposure of Ireland’s coast to climate hazards, providing essential insights to inform policy, management, adaptation, and resource allocation. Study area Ireland is an island nation located in northwestern Europe (Fig. 1), where approximately 40% of the population live within 5 km of the coast and over 130,000 people reside in areas less than 5 m above mean sea level (Center For International Earth Science Information Network-CIESIN-Columbia University 2021; CSO 2016). Facing the Northeastern Atlantic Ocean, Ireland’s western, wave-dominated coastline is predominantly rocky and indented, with bay-headland configurations common (Cooper et al. 2004; Devoy 2008; Haslett 2016). In contrast, the eastern coast on the Irish Sea and the southeastern coast on the Celtic Sea, are generally lower lying and comprised of “soft”, often unconsolidated sediments, deposited during the Quaternary (Devoy 2008; Masselink et al. 2014).Fig. 1Republic of Ireland Coastline with coastal Local Authorities, broad shoreline type, and sea boundaries around the Irish coast. Local Authority outlines source: Tailte Éireann (2024a), Country boundaries source: European Commission, Eurostat (2024). Shoreline type source: Geological Survey Ireland (2024), World Sea Boundaries Source: Marine Institute, (2024) Republic of Ireland Coastline with coastal Local Authorities, broad shoreline type, and sea boundaries around the Irish coast. Local Authority outlines source: Tailte Éireann (2024a), Country boundaries source: European Commission, Eurostat (2024). Shoreline type source: Geological Survey Ireland (2024), World Sea Boundaries Source: Marine Institute, (2024) Extreme winds, and coastal flooding and erosion have been identified as climate hazards which pose the greatest risk to Ireland (Environmental Protection Agency 2025). Erosion, leading to loss or damage to property, infrastructure, or habitats, has been widely reported along parts of the east and southeastern coastlines (Caloca-Casado 2018; Monteys et al. 2024; Senatilleke et al. 2026). The west is also impacted by coastal erosion, although given it is mostly a rocky coast, high erosion rates tend to be focused within sedimentary bays or beaches (Cooper et al. 2004; Gault et al. 2011). Historically, solutions to coastal erosion in Ireland have involved hard, engineered structures such as rock revetments, sea walls and groynes (O’Hagan and Cooper 2002), although recent strategy proposals highlight the need for more sustainable solutions (OPW 2023). Coastal flooding also poses a significant hazard in Ireland, which when coupled with fluvial flooding can lead to widespread damage in urban areas such as Cork City, Galway City and Dublin City (Devoy 2008; Flood and Sweeney 2012; Paranunzio et al. 2022; Tyrrell and Hickey 1991). SLR will intensify the extent and impact of these events, with some estimates indicating that a 1 m rise in sea level would affect up to 350 km2 of Ireland’s coastal zone, with potential economic costs of €1.1 billion (Flood and Sweeney 2012). One metre of SLR is also predicted to lead to the loss of up to 30% of the nation’s coastal wetlands, further exacerbating the impacts from climate hazards (Devoy 2008). Ireland has a rich archaeological history, with over 150,000 records in the Register of National Monuments (data.gov.ie/dataset/national-monuments-service-archaeological-survey-of-ireland). These archaeological sites are protected under the Historic and Archaeological Heritage and Miscellaneous Provisions Act (2023), with site management and conservation conducted by the Office of Public Works (OPW) (O’Sullivan and Downey 2023). Many of these sites are threatened by climate hazards, with coastal sites identified as being particularly at risk to structural damage and erosion (Daly 2019; Jones et al. 2023). Whilst numerous coastal heritage sites may inevitably be lost, the Irish Government has emphasised the need to develop a “risk register”, to identify exposed sites and direct resources to areas where preventative actions may be viable (Daly 2019). The OPW is also responsible for monitoring, modelling and assessing coastal hazards and has completed several national coastal hazard assessments. These include shoreline erosion projections, coastal flood hazard mapping under present and future sea-level rise scenarios, extreme water level modelling, and identified sites potentially exposed to wave-overtopping (OPW 2021, 2018, 2013, 2010). These datasets provide national information on coastal erosion and flooding hazards but do not explicitly assess the vulnerability or exposure of coastal populations, ecosystems or assets (Flood and Schechtman 2014). Although regional-scale studies have evaluated coastal vulnerability and risk in selected areas of the Irish coastline (Caloca-Casado 2018; Chalençon et al. 2026; Paranunzio et al. 2022), no nationwide assessment of coastal vulnerability has yet been published. Numerous studies have examined the planning, adaptation, and management strategies Ireland could adopt to address the impacts of coastal climate hazards (Devoy 2008; Falaleeva et al. 2011; Flood and Schechtman 2014; Kopke and O’Mahony 2011; Lawlor and Cooper 2024). These, along with The Irish Coastal Change Management Strategy (OPW 2023), underscore the need to identify where risk from coastal change is greatest, requiring an integrated approach that combines hazard characterisation with broader considerations central to effective coastal zone management. In response, the present study aims to address this need by examining the following research questions (RQs): RQ1: What is the spatial distribution of relative exposure levels to coastal hazards along the Irish coast? RQ2: What are the controls on coastal hazard distribution across Ireland? RQ3: Where can coastal habitats reduce relative exposure and mitigate against the potential impacts of these hazards? RQ4: What populations and cultural heritage sites are located in highly exposed coastal regions? Materials and methods The open-source InVEST® (Integrated Evaluation of Ecosystem Services and Tradeoffs) software suite is developed by the Natural Capital Project (https://naturalcapitalproject.stanford.edu/) and contains several, spatially explicit models designed to map, assess and valuate the services natural ecosystems provide to humanity. One such model, the Coastal Vulnerability Model (CVM) is based on the traditional CVI approach. It requires input of physical variables such as shoreline type, natural habitats, elevation and bathymetry as well as climatological data such as wind speed, wave power, storm surges, and rates of sea level change. Details on the specific variables employed in this assessment can be found in Tables SI-1 and SI-2 of the Supplementary Information. The ultimate product of the CVM is a series of points at a user-defined interval along a coastline. In line with previous studies (Al Ruheili and Boluwade 2023; Silver et al. 2019; Thomas and Lekshmy 2025) a 250 m alongshore line spacing was chosen for this assessment of the ROI coastline. The coastline was derived from the seaward limit of the 2024 National Statutory Boundaries dataset (Tailte Éireann 2024b), and largely represents the High Water Mark. The coastal points created along this line contain a semi-qualitative ranking (on a 1–5 scale) for each variable included in the model (Table 1).Table 1Coastal exposure indexCoastal Exposure IndexVery LowLowModerateHighVery High12345Shoreline typeBedrock outcropBuilt Environment: Causeway, Dam, Embankment, PierGlacial TillGravels, Estuarine & Marsh silts and claysAlluvium, Peat, Lacustrine and Glaciomarine sediments, Marine sands and gravels, Windblown sands. Built Environment: Airfield/Airport, Industrial, Landfill, Made ground, UrbanRelief81 to 100 Percentile61 to 80 Percentile41 to 60 Percentile21 to 40 Percentile0 to 20 PercentileCoastal HabitatsHigh, Wide DunesHigh, Narrow Dunes,Low, Wide Dunes, Machair. Wide SaltmarshNarrow, Low Dunes. Narrow Saltmarsh. SeagrassNo habitatWind, Wave, Storm Surge and Sea Level Rise (SLR) Exposure0 to 20 Percentile21 to 40 Percentile41 to 60 Percentile61 to 80 Percentile81 to 100 Percentile Table 1Coastal exposure indexCoastal Exposure IndexVery LowLowModerateHighVery High12345Shoreline typeBedrock outcropBuilt Environment: Causeway, Dam, Embankment, PierGlacial TillGravels, Estuarine & Marsh silts and claysAlluvium, Peat, Lacustrine and Glaciomarine sediments, Marine sands and gravels, Windblown sands. Built Environment: Airfield/Airport, Industrial, Landfill, Made ground, UrbanRelief81 to 100 Percentile61 to 80 Percentile41 to 60 Percentile21 to 40 Percentile0 to 20 PercentileCoastal HabitatsHigh, Wide DunesHigh, Narrow Dunes,Low, Wide Dunes, Machair. Wide SaltmarshNarrow, Low Dunes. Narrow Saltmarsh. SeagrassNo habitatWind, Wave, Storm Surge and Sea Level Rise (SLR) Exposure0 to 20 Percentile21 to 40 Percentile41 to 60 Percentile61 to 80 Percentile81 to 100 Percentile Coastal exposure index The model then calculates a CEI value for each coastal point based on the geometric mean of the ranked variables (Eq. 1):1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}CEI= {\left({R}_{Shoreline type}\cdot {R}_{Relief}\cdot {R}_{Habitats}{\cdot R}_{Wind}{\cdot R}_{Wave}{\cdot R}_{Surge}\cdot {R}_{SLR}\right)}^{1/7}\end{array}$$\end{document} This multiplicative approach yields a range of CEI values on the same 1–5 scale as the individual variables, facilitating straightforward interpretation. It also accounts for the non-linear interactions among these variables, as opposed to an additive approach which could produce a misleading representation of exposure, whereby a high index rank in one variable offsets low ranks in others, potentially masking genuine exposure (Al Ruheili and Boluwade 2023; Silver et al. 2019). Data mining and processing Shoreline type and relief Shoreline Type was classified from the “1:50,000 Quaternary Sediments” layer from the Geological Survey of Ireland (GSI, 2021). Each of the 52 coastal units were grouped into 10 geological classes (Table 1) based on sediment characteristics and on the potential erodibility of the substrate. One class, termed “Built Environment”, is applied where the coast has been heavily modified through built structures, such as piers or embankments. Exposure ranks for each class were determined based on similar studies (Arkema et al. 2013; Hammar-Klose and Thieler 2001; López Royo et al. 2016; Mclaughlin and Cooper 2010; Pendleton et al. 2010; Rocha et al. 2020), with sub-classes of the “Built Environment” classified individually in line with Chalençon et al. (2026). Where coastal segments encompass multiple units, an intermediate value is calculated within the CVM from the average of ranked units within half the shoreline spacing (125 m). Given the increasing threats from storm surges and SLR, the topographic relief of a coastline serves as a key indicator of its potential exposure to hazards such as flooding or permanent inundation (Gornitz 1991; Mclaughlin and Cooper 2010). To account for the diverse topography of a coastline, the CVM calculates the average elevation value within a user-defined radius around each shoreline point. A 10 m resolution national Digital Terrain Model (DTM) was utilised for this assessment, which was generated from a point cloud, derived from both camera and LiDAR sensors, with an accuracy of ± 0.25 m and ± 1.5 m respectively (Tailte Éireann 2021). A 125 m elevation averaging radius was used to calculate the mean elevation of DTM cells surrounding each shoreline point, relative to Malin Head Ordnance Datum (OD). This radius was chosen to account for minor discrepancies between the statutory coastline used in this assessment and the coastline captured by the DTM, for example where erosion or anthropogenic modification has altered the coast. By default, the CVM clips the DTM to the statutory coastline and sets negative elevations (below mean sea level) and no-data cells to zero. This not only excludes land within the DTM seaward of the statutory boundary but can also mask genuinely low-lying land below mean sea level, potentially underestimating exposure at affected points. To avoid this, the baseline results presented here instead use relief values calculated independently via zonal statistics in QGIS, applying the same 125 m radius but excluding no-data cells from the average rather than treating them as zero. The default CVM relief calculation was retained as a variable in the sensitivity analysis, to assess how changing this parameter affects overall exposure results. Wind and waves Formatted to align with the CVM methodology, a dataset of 30 years (1995–2024) of wind and wave records surrounding the coast of Ireland was developed, using 3-hourly statistics on wind speed and direction from the Copernicus European Regional ReAnalysis (CERRA) dataset (Copernicus Climate Change Service 2021), and wave data from the Atlantic- European North West Shelf- Wave Physics Reanalysis (NWSHELF) dataset (European Union-Copernicus Marine Service 2020). These were chosen due to their superior spatial resolution, (5.5 km for CERRA and 1.9 km for NWSHELF), allowing for better performance in coastal regions when compared with coarser datasets, such as WaveWatch III or ERA5 (European Union-Copernicus Marine Service 2020; Verrelle et al. 2022). The temporal range of 1995–2024 was selected to provide a contemporary view of present wind and wave conditions, with 30-years chosen in accordance with the recommended averaging period for climatic analysis (Re et al. 2023; WMO 2015). Numerous, extreme weather events occurred during this time frame, see Bricheno et al. (2025); Camus et al. (2024); O’Brien et al. (2018); and Olbert and Hartnett (2010) for further information on such events. To account for the complex geometry of some coastlines, the CVM computes the relative exposure of the coastline to wind and waves based on the directionality of the highest wind speeds and strongest waves. The model casts ‘Fetch Rays’, which subdivides the compass rose into 16, equiangular sectors to group data based on the direction from which the wind or waves are coming from. Average wind speeds in the 90th percentile of the dataset are used to convey storm conditions. A Relative Exposure Index (REI) is calculated for the wind speeds using:2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}REI= {\sum}_{n=1}^{16}{U}_{n}{P}_{n}{F}_{n}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${U}_{n}$$\end{document} is average wind speed (ms−1) of highest 10% wind speeds in the nth sector; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${P}_{n}$$\end{document} is the percentage of the highest 10% of wind speeds that blow in direction of the nth sector and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${F}_{n}$$\end{document} is the fetch distance in meters in the nth sector. The relative exposure to waves \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${(E}_{w})$$\end{document} for each coastal point is determined by the maximum of the weighted average of wave energy values across each 16 sectors for both oceanic (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{w}^{o})$$\end{document} and local waves \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${(E}_{w}^{l})$$\end{document}.3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{E}_{w}=max({E}_{w}^{o}, {E}_{w}^{l})\end{array}$$\end{document} Each directional sector is classified as either exposed to ocean or local waves based on whether the fetch ray from that directional sector extends to the ‘Maximum Fetch Distance’ without reaching a landmass. The ‘Maximum Fetch Distance’ in this study was chosen at 10 km, with fetch rays reaching beyond this distance classified as exposed to the open ocean. This distance was chosen to prioritise calculating wave exposure from the NWSHELF dataset, computing local wave conditions only in enclosed bays or inlets along the coast. Relative exposure to oceanic waves in calculated via:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{E}_{w}^{o}= {\sum}_{k=1}^{16}{H[F}_{k}]{P}_{k}^{o}{O}_{k}^{o}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${H[F}_{k}]$$\end{document} = 1 if the fetch ray (F) in the angular sector k reaches 10 km, and 0 for those sectors where F is less than 10 km. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${P}_{k}^{o}$$\end{document} = the average of the highest 10% of relative wave exposure values in the angular sector k, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${O}_{k}^{o}$$\end{document} = the percentage of time those waves occurred within angular sector k. For all waves, the relative wave energy index (P) was calculated by:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}P= \frac{1}{2} {H}^{2} T\end{array}$$\end{document}where H is the wave height [m] and T is the wave period [s]. For the wave data, statistics for the highest 10% of relative wave energy values (calculated using Eq. 5), their associated directions, and the proportion of these values within each sector, were extracted to a CSV from the NWSHELF NetCDF data. These data were then combined with the wind records, joined to the nearest pixel-centroid location from the wind grid. The processing of these data was completed in Python, following the workflow detailed in Figure SI-1 of the Supplementary Information. Further details on the formatting of the wind and wave dataset, as well as the calculation of local wave exposure, can be found in the InVEST® CVM user documentation (Natural Capital Project 2025). Once the relative exposure values for wind and waves are computed for each shoreline point, the index ranks are determined from the quintile ranges of all values for each element (Table SI-2). Coastal habitats The coastal habitats included within this assessment are Saltmarshes; Seagrasses; Dunes; and Machair. Machair is a dune-type habitat unique to parts of Scotland and western Ireland, defined as “areas of level, stable coastal dune grassland over calcareous soils” (Bassett and Curtis 1985, p.1). These habitats vary considerably in terms of their spatial footprint, morphology, and connectivity with the coast, all of which influence the potential protective capability of each habitat. To account for some of this variability, dune and habitats saltmarsh (the most prevalent habitats in this study) were assessed in terms of their mean width. This was calculated in GIS by creating a series of approximately shore-normal transects spaced at 50 m, and determining the mean width across all transects within each habitat polygon. Dune habitats disconnected from the coast by more than 100 m, either due to being bisected by a road or through inaccurate mapping, were removed to avoid potentially misrepresenting habitat protection in those areas. The four classes of coastal habitats included in this assessment were ranked based on their protective potential (Table SI-3), which was informed by those suggested in the InVEST® user documentation (Natural Capital Project 2025) and through a review of the relevant literature (Chen et al. 2007; Debaine and Robin 2012; Hanley et al. 2014; Itzkin et al. 2021; Jordan and Fröhle 2022; Leonardi et al. 2018; Shepard et al. 2011). Those habitats with a higher protective potential are assigned lower exposure index ranks. Such habitats include High, Wide Dunes, defined here as dunes with a mean width of 100 m or higher, and a mean elevation of 4 m or higher (Table 2). The width threshold was selected based on previous work by Pantusa et al. (2018). The height threshold was derived from recent work by Cotrim et al. (2026), which revealed total water levels (accounting for surge, astronomical tide and wave set-up) around Ireland during a 100-year return period (RP) storm event do not exceed 4 m (Cotrim et al. 2026). Habitats such as Narrow, Low Dunes; Narrow Saltmarsh (< 100 m mean width); and Seagrasses, were assigned higher exposure rank values, as these will may have limited wave and surge attenuation effects during storms, but may still contribute to shoreline stabilisation (Chen et al. 2007; Itzkin et al. 2021; Leonardi et al. 2018; Shepard et al. 2011).Table 2Habitat size classificationMean transect width per polygon< 100 m>= 100 mMean elevation (metres above Malin Head OD)< 4 mLow, Narrow DunesLow, Wide Dunes>= 4 mHigh, Narrow DunesHigh, Wide Dunesn/aNarrow SaltmarshWide Saltmarsh Table 2Habitat size classificationMean transect width per polygon< 100 m>= 100 mMean elevation (metres above Malin Head OD)< 4 mLow, Narrow DunesLow, Wide Dunes>= 4 mHigh, Narrow DunesHigh, Wide Dunesn/aNarrow SaltmarshWide Saltmarsh Habitat size classification For shoreline points where two or more adjacent habitats occur, the InVEST® model creates an array R containing all ranks Rk, 1 ≤ k ≤ N, where N is the number of habitats. The habitat exposure rank is determined using:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{R}_{hab}=4.8-0.5 \sqrt{{(1.5\underset{k=1}{\mathit{max}}(5-{R}_{k}))}^{2}+ \sum_{k=1}^{N}{(5- {R}_{k})}^{2}-( \underset{k=1}{max}(5- {R}_{k}))^{2}}\end{array}$$\end{document}where habitats with the most protective effect are weighted 1.5 times higher than other habitats near that shoreline segment. The resulting value, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${R}_{hab}$$\end{document}, is then used within the geometric mean calculation to determine the CEI. To determine the role of habitats in reducing exposure, a “no-habitat” scenario is calculated within the CVM by applying the maximum value of 5 for \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${R}_{hab}$$\end{document} to all coastal points. Storm surge As part of the CVM, a storm surge potential variable is calculated based on the assumption that storm surge elevation is a function of the amount of time wind blows over relatively shallow water (Natural Capital Project 2025). The CVM therefore computes the distance between the coastline to the edge of the continental shelf, with larger distances deemed more exposed to higher storm surges. To test the reliability of the results of the CVM storm surge variable, the ranks were compared against hydrodynamically modelled surge data from the “Global sea level change indicators from 1950 to 2050 derived from reanalysis and high resolution CMIP6 (Coupled Model Intercomparison Project Phase 6) climate projections” dataset (Copernicus Climate Change Service 2022a). Absolute surge levels from a 100-year RP storm event were extracted from the 1985–2014 reanalysis data, modelled at a 0.1° (~ 12 km) horizontal resolution along the coast. Surge values from the CMIP6 grid were joined to the coastal points, with each point inheriting the surge value form the nearest grid cell centroid. The storm surge exposure levels correspond to the quintile ranges of these surge levels around the ROI coastline, with the resulting ranks from the CVM storm surge and CMIP6 storm surge assessed through a Spearman’s rank correlation. These showed a strong negative correlation, indicating the unsuitability of the CVM surge proxy for Ireland. As such, the surge ranks from the CMIP6 dataset were applied in the CEI calculation, which is taken as the baseline model. The results of relative exposure values derived from the default CVM surge values tested in the sensitivity analysis. Sea level rise Modelled rates of SLR were extracted from the “Global sea level change time series from 1950 to 2050 derived from reanalysis and high resolution CMIP6 climate projection” dataset (Copernicus Climate Change Service 2022b). This dataset contains annual rates of mean sea level change from historic reanalysis data and accounts for glacial isostatic adjustment based on work by Peltier et al. (2015). Mean sea level change rates (mm yr−1) were extracted for the Irish coast from the same 30-year time range as covered by the storm surge data (1985–2014) with exposure levels determined from the quintile ranges of the change rates within these data points. Population and cultural heritage exposure To ensure the most representative portrayal of population distribution around the coast, a coastal population raster for the ROI was created using a dasymetric mapping approach (Mennis 2003). This method seeks to account for the limitation of the aggregation of population statistics across census designated “blocks”, which may misrepresent the true population distribution at the coast (Arkema et al. 2013; Mennis 2003). Small Area Population Statistics (SAPS) for 2022, provided by the Central Statistics Office (CSO 2023a), were used to calculate the total population for each area located within 1 km of the coastline. The total number of residential buildings within each coastal area was determined from the GeoDirectory database (GeoDirectory 2023) and an average population per building was assigned. The results were rasterised at a 100 m resolution, with each pixel representing the sum of the average population in each building within that cell. Coastal populations for each exposure category were determined by calculating the total population within 125 m, 250 m and 500 m of each coastal point. Voronoi polygons were used to construct these inland buffers, ensuring a more accurate attribution to the nearest coastal exposure category than would be possible with a conventional, uniform buffer. The three buffer distances correspond to 50%, 100%, and 200% of the fixed shore point spacing (250 m) were evaluated to account for potential positional uncertainty in the statutory coastline. Although many of Ireland’s population centres are located at or near the coast, much of the country’s shoreline remains sparsely populated. In many of these more remote areas, particularly along the west coast, numerous historical and archaeological sites are found which represent significant social and cultural value for the nation (Jones et al. 2023). The National Monuments Service provides a continually updated record of archaeological sites for Ireland. This dataset was downloaded with records classified as “redundant” removed. As with population, cultural heritage sites within 125 m, 250 m, and 500 m of each exposure category were identified for Ireland’s 17 coastal LAs. Sensitivity analysis Determining coastal exposure through an index-based approach requires numerous choices when deciding what variables to include, how to rank those variables, and how to display the results in an interpretable manner. No universal method exists, nor would one necessarily be desirable, given that coastal regions vary in physical, environmental and climatic characteristics, meaning hazard-driving conditions in one region may not apply equally elsewhere. Given the many possible approaches which could be taken in computing a CEI or CVI, testing an index's sensitivity to variations in how input variables are ranked, or how the resulting score is classified, is important for informed interpretation of the results (Chalençon et al. 2026; Pantusa et al. 2022; Re et al. 2023; Thomas and Lekshmy 2025; VanZomeren and Acevedo-Mackey 2019). Leave-one-out analysis To test the sensitivity of the CEI results to changes in the input variables, a Leave-one-out (LOO) analysis was conducted. For each run, one variable was omitted from the CEI calculation, with the new relative exposure score taken from the geometric mean of the 6 remaining variables. Pearson’s and Spearman’s rank correlations were conducted, comparing each new LOO exposure score with the baseline model. Changes in coastal points designated “Very High Exposure” were calculated per Local Authority, to aid in determining which variables have the most influence in each area. Alternative variables Following the approaches described above, the default Relief and Surge exposure variables produced by the CVM were replaced for the baseline model (CEI_baseline). To test the sensitivity to model to these changes, the results of the alternative variable ranks, and CEI ranks derived from these variables, were compared against the baseline model. An additional variable was also introduced, the total water level (TWL). TWL encompasses the astronomical tide, the wave setup, and the storm surge residual (Cotrim et al. 2026; Vousdoukas et al. 2016). TWL therefore presents a more accurate predictor of coastal flooding potential than surge residual alone, particularly for meso- or macro-tidal environments such as Ireland. To account for this, a second index (CEI_TWL) was created wherein the storm surge variable was replaced by a TWL variable, derived from Cotrim et al. (2026). This dataset includes extreme water level values from a 100-year RP storm event for the period of 1985–2021, modelled at approximately 1 km along the European coastline. These were joined to the nearest coastal point, with the resulting TWL ranks derived from the quintile ranges of these values. The relative exposure results for this third index (CEI_CMIP6, comprised of the CMIP6 surge values and the default relief ranks), and the default InVEST output (CEI_CVM), were compared against the baseline. Alternative ranking methods As with all index-based approaches, the designation of the variable ranks is a crucial component of the final exposure or vulnerability score. Ranks may be defined using user-specified intervals (e.g. Gornitz 1991; Hammar-Klose and Thieler 2001; Koroglu et al. 2019; Mclaughlin and Cooper 2010; Pantusa et al. 2022), weights derived from multi-criteria decision-making methods such as the Analytical Hierarchical Process (AHP) (Chalençon et al. 2026; Mani Murali et al. 2013; Rocha et al. 2020), or through statistical classification of the data. The latter is the default method for the CVM, with quintile ranges applied to the variables consisting of continuous values. Similarly, the final exposure ranks must be expressed in interpretable classes (e.g. from Very Low—1 to Very High—5), which again relies on statistical methods of sorting the data, frequently through percentile ranges (e.g. Arkema et al. 2013; Hammar-Klose and Thieler 2001; Koroglu et al. 2019; Pantusa et al. 2022). To test the sensitivity of the CEI to changes in the ranking method of such variables, four classification methods were compared: quintile ranges (employed in baseline model); equal intervals; Jenks natural breaks; and geometric intervals. The resulting CEI was calculated from the geometric mean of the newly classified variables, with final relative exposure ranks derived from the same method used to classify the variables. Results Coastal exposure to climate hazards A 250 m alongshore line spacing produces 34,303 coastal points, corresponding to approximately 8575 km of coastline. Traditionally, ROI is divided into 26 counties, 14 of which are coastal. Dublin, the most populous, is subdivided into four administrative areas (three coastal), and Galway City has its own LA. The results of the CVM are presented for these 17 coastal LAs (Fig. 2A). Each coastal point contains a relative CEI value, determined by the geometric mean of the seven variables. These were subsequently classified into 5 ranks based on the quintile ranges across all coastal points: “Very Low” (1.31–2.32), “Low” (2.32–2.56), “Moderate” (2.56–2.83), “High” (2.83–3.18), and “Very High” (3.18–4.65).Fig. 2Coastal Exposure Index for the Republic of Ireland showing levels of relative exposure to coastal hazards (Map A); and relative exposure levels for the “no-habitat” scenario (Map B) for Ireland’s 17 coastal Local Authorities Coastal Exposure Index for the Republic of Ireland showing levels of relative exposure to coastal hazards (Map A); and relative exposure levels for the “no-habitat” scenario (Map B) for Ireland’s 17 coastal Local Authorities Due to its indented, crenulate nature, LAs on the western coast have the longest coastal stretches, with Galway, Cork, Donegal, Mayo and Kerry County Councils accounting for 82% of the total assessed coastline. Mayo records the greatest extent of “Very High Exposure” (414.5 km), while Galway County has the highest length of both “High Exposure” (330 km) and “Moderate Exposure” (297 km) coastline. Cork County contains the largest lengths of coastline classified as “Low Exposure” and “Very Low Exposure” measuring 435 km and 485 km, respectively (Fig. 3).Fig. 3Bar chart depicting coastal length (km) by relative exposure category for each Local Authority Bar chart depicting coastal length (km) by relative exposure category for each Local Authority The proportion of coastline within each exposure level (Fig. 4) provides a different perspective, particularly for LAs with relatively short coastlines. Approximately 68% of the 5.5 km coastline of Leitrim is deemed “Very High Exposure”. In Meath, which has 11 km of coastline, 57% is categorised as “High Exposure”. Dublin City has approximately 52 km of coastline, 26% of which is “Moderate Exposure”. Cork County has the largest overall proportion of “Low Exposure” and “Very Low Exposure” coastal points, at 29% and 32% of the coastline, respectively.Fig. 4Bar chart depicting the proportion of coastline by relative exposure category for each Local Authority Bar chart depicting the proportion of coastline by relative exposure category for each Local Authority The mean CEI value across all LAs is 2.74, with a standard deviation of 0.52 (see Table SI-4 in the Supplementary Information). Mean CEI values within each LA are lowest in Cork (2.49) and highest in Leitrim (3.40). The seven variables which determine the CEI (Fig. 5) reveal trends in the physical nature of the coast and the distribution of metocean and climatic forcing elements around Ireland. Exposure levels based on Shoreline Type (Fig. 5A) tends to be higher in the east and lower in the west of the country. The mean value of exposure based on Shoreline Type for all LAs is 2.3 (Table SI-4), with mean values highest in Meath (4.9) and lowest in Cork (1.6). Relief (Fig. 5D), averaged over a 125 m radius around each coastal point, has a mean value of 9.9 m (relative to Malin Head OD), ranging from − 7.1 to 485.2 m across the coast, peaking at the Croaghaun Cliffs on Achill Island, County Mayo. Mean values for Relief are lowest in Dublin city (2.5 m) and highest in Clare (14.0 m).Fig. 5Ranked relative exposure levels for each of the seven variables within the Coastal Exposure Index for the Republic of Ireland Ranked relative exposure levels for each of the seven variables within the Coastal Exposure Index for the Republic of Ireland Exposure levels for Wind (Fig. 5B) and Waves (Fig. 5C) show more spatial variability. Wind REI levels, calculated using Eq. 2, average at 22,093 and range from 0 to 143,737, with highest wind REI levels seen in Donegal. Meath has the highest mean wind exposure (41,483) and Dublin City has the lowest mean values (10,493). Relative wave energy values range from 0 to 305.2 nationally, highest in Kerry. Mean values for relative wave energy are 17.9, with Clare having the highest mean wave energy values (51.3) and Dublin City having the lowest (0.5). The exposure levels for Storm Surge (Fig. 5E) and SLR (Fig. 5F) depict broad, regional trends across the country. Storm surge levels for a 100-year RP event range from 1.0 to 1.7 m, peaking around Malin Head, County Donegal (Table SI-4). Nationally, mean storm surge heights are 1.3 m, with mean heights lowest in Cork (1.1 m) and highest in Galway City (1.6 m). The average SLR rates for Ireland from 1985 to 2014 range from 0.3 to 2.2 mm yr−1. Mean values across the nation are 1.8 mm yr−1, with the southern counties of Cork and Kerry having the highest mean rates of SLR, at 2.1 mm yr−1 and Donegal having the lowest (1.0 mm yr−1). Sensitivity analysis Leave-one-out analysis The resulting ranks from each iteration of the LOO Analysis reveal the relative influence of each variable in the CEI, by performing a Pearson’s correlation (r) and Spearman’s rank correlation (ρ) against the baseline model (Table 3). Removal of the Shoreline Type variable produces the largest change in ranks across all coastal points, with 57.1% shifted one rank and a mean rank shift of 0.61. Leaving out Coastal Habitats provides the least overall change, with a ρ of 0.95 and a mean rank shift of 0.18 across all points.Table 3Results of leave-one-out sensitivity analysis—relative change to final exposure value when reducing each variable by one rankVariablePearson correlation (r) w/ baseline modelSpearman correlation (ρ) w/baseline model% of coastal points with rank shifted ± 1% of coastal points with rank shifted ± 2Mean rank shiftShoreline Type0.870.8357.13.80.61Relief0.910.8548.04.90.53Wind0.920.8545.84.60.50Wave0.920.8546.05.20.51Coastal Habitats0.980.9517.30.70.18Storm Surge0.910.8644.64.30.49Sea Level Rise0.920.8539.93.00.43 Table 3Results of leave-one-out sensitivity analysis—relative change to final exposure value when reducing each variable by one rankVariablePearson correlation (r) w/ baseline modelSpearman correlation (ρ) w/baseline model% of coastal points with rank shifted ± 1% of coastal points with rank shifted ± 2Mean rank shiftShoreline Type0.870.8357.13.80.61Relief0.910.8548.04.90.53Wind0.920.8545.84.60.50Wave0.920.8546.05.20.51Coastal Habitats0.980.9517.30.70.18Storm Surge0.910.8644.64.30.49Sea Level Rise0.920.8539.93.00.43 Results of leave-one-out sensitivity analysis—relative change to final exposure value when reducing each variable by one rank The percentage of coastal points retaining the “Very High Exposure” rank for each LA (Figure SI-2) indicates which variables most strongly influence this upper 80th percentile exposure category. Dun Laoghaire-Rathdown is most sensitive to Shoreline Type, Relief, and Storm Surge, with no points retaining the rank when each of these variables are removed. In Wexford, Wind and Waves have the greatest influence, with 42.6% and 48.0% of points, respectively, retaining the rank following their removal. Fingal is most sensitive to the Coastal Habitats variable, with 56.8% of points retaining the rank when this parameter is omitted. In Cork County, SLR has the largest effect, with only 41.2% of points remaining in the “Very High Exposure” category after its removal. Alternative variables Pearson’s correlation values from the continuous data, and Spearman’s rank correlation values for the ranked data, for the alternative variables are shown in Table 4. A notable outcome is the very strong correlation (r = 0.96) between the CEI_baseline relief values (calculated via zonal statistics in QGIS) and those calculated via the CV. However, the Spearman’s rank correlation (ρ = 0.54), reveals the statistical ranking produces considerable shifts in the Relief ranks across all coastal points, highlighting the sensitivity of the model based on these quintile breaks in the data. The results also reveal the strong negative correlation between the default storm surge proxy calculated by the CVM, and the modelled surge residual values (r = − 0.72, ρ= − 0.68). The CMIP6 surge residual and the total water levels, modelled by Cotrim et al. (2026), show moderate to strong Pearson’s and Spearman’s correlations (r = 0.66, ρ= 0.69).Table 4Pearson’s correlation and Spearman’s rank correlation of alternative variables tested within the Coastal Exposure IndexComparisonPearson’s correlation (r)Spearman’s rank correlation (ρ)Alternative VariablesRelief: Baseline model relief vs CVM default relief0.960.54Surge: Baseline model CMIP6 storm surge residual vs CVM default− 0.72− 0.68Water level: CMIP6 storm surge residual vs total extreme water level0.660.69Alternative CEI outcomesCEI_baseline vs CEI_CVM0.650.60CEI_baseline vs CEI_CMIP60.780.73CEI_baseline vs CEI_TWL0.960.93 Table 4Pearson’s correlation and Spearman’s rank correlation of alternative variables tested within the Coastal Exposure IndexComparisonPearson’s correlation (r)Spearman’s rank correlation (ρ)Alternative VariablesRelief: Baseline model relief vs CVM default relief0.960.54Surge: Baseline model CMIP6 storm surge residual vs CVM default− 0.72− 0.68Water level: CMIP6 storm surge residual vs total extreme water level0.660.69Alternative CEI outcomesCEI_baseline vs CEI_CVM0.650.60CEI_baseline vs CEI_CMIP60.780.73CEI_baseline vs CEI_TWL0.960.93 Pearson’s correlation and Spearman’s rank correlation of alternative variables tested within the Coastal Exposure Index When accounting for all coastal points, comparison of the overall CEI values reveals a moderate to high correlation between the baseline model (CEI_baseline) and the default InVEST® output (CEI_CVM) (r = 0.65, ρ= 0.60). A high correlation exists between the baseline model and the model employing the default relief value and the CMIP6 surge value (CEI_CMIP6) (r = 0.78, ρ = 0.73). The highest correlation occurs between the CEI_baseline and the CEI_TWL, with strong Pearson’s and Spearman’s correlations (r = 0.96, ρ= 0.93). Examining how each scenario affects the designation of “Very High Exposure” coastal points at the LA level reveals more spatially specific differences (Figure SI-3). Dun Laoghaire–Rathdown retains no “Very High Exposure” points under either the CEI_CVM or CEI_TWL scenarios, suggesting this area is particularly sensitive to both the water level variable specification and the combined effect of default relief and surge parameters. In contrast, most LAs show broadly consistent “Very High Exposure” designations under the CEI_CMIP6 scenario, with the notable exception of Mayo, which retains only 40% of its baseline “Very High Exposure” points. Alternative ranking methods Histogram graphs of each of the five continuous variables, i.e. Wind; Waves; Relief; Storm Surge; and SLR (Figure SI-5) display key distributional trends. Wind, Waves and Relief display a strong positive skewness, with most coastal points containing lower values, and few areas encountering extreme high values. Storm Surge displays a trimodal distribution, with a slight skew to the right. SLR is skewed to the left, with most points inheriting rates in the higher end of the distribution (~ 2 mm yr−1). Comparison of different methods for ranking these variables, with their resulting CEI values (calculated via Eq. 1), are shown in Figs. 6 and 7.Fig. 6Pearson and Spearman correlation matrices for Coastal Exposure Index values and ranks under alternative classification and ranking methodsFig. 7Boxplots depicting the distribution of Coastal Exposure Index values when different statistical methods for ranking the continuous variables are employed Pearson and Spearman correlation matrices for Coastal Exposure Index values and ranks under alternative classification and ranking methods Boxplots depicting the distribution of Coastal Exposure Index values when different statistical methods for ranking the continuous variables are employed Pearson’s and Spearman’s values (Fig. 6) indicates the strongest correlation between the quintile ranking methods occurs with the geometric interval ranking method (r = 0.68, ρ = 0.60). The other methods correlate much higher with each other than with the quintile, with geometric interval and Jenks natural breaks having the strongest correlation (r = 0.88, ρ = 0.78). The distribution of relative exposure shifts substantially depending on the ranking method used (Figure SI-6), with the “High Exposure” and “Very High Exposure” categories showing the most pronounced differences. Under the equal interval method, just 97 points nationally fall into the “Very High Exposure” category, with 482 points under the geometric interval method. In both cases, these points are concentrated in the west and northwest, with none appearing along the east coast. The natural breaks method produces a much larger national total of 2023 points in this category, of which only 20 are located along the east coast. Protective role of coastal habitats The results of the CVM reveal 38% of coastline of Ireland (3299 km) is within the potential protective distance of the coastal habitats included in this assessment. Saltmarshes have the most influence, with 8268 coastal points (4134 km) and 2026 points (1013 km of coastline) within the 500 m protective distance of a Narrow Saltmarsh and a Wide Saltmarsh, respectively. Approximately 477 km of coastline is within 250 m of High, Wide Dunes; 245 km for High, Narrow Dunes; 34 km for Low, Wide Dunes; and 218 km for Low, Narrow Dunes. Machair influences 203.75 km of coastline; and Seagrasses 887 km. The lowest Habitat exposure rank (1.1) occurs in Donegal, in an area of High, Wide Dune backed by Wide Saltmarsh. Mean values of Habitat exposure for Ireland are 4.4, with lowest mean values seen in Meath (3.5) and highest mean values in Leitrim (5.0) (Table SI-5). To assess the role in habitats in reducing relative exposure, the same quintile ranges from the baseline CEI model were applied to derive the no-habitat scenario rankings. The exposure levels in the no-habitat scenario are presented in Fig. 8, with Mayo again having the highest total amount of “Very High Exposure” coastline, increased to 472.8 km.Fig. 8Bar chart of coastal length (km) by exposure category for each Local Authority for the no-habitat scenario Bar chart of coastal length (km) by exposure category for each Local Authority for the no-habitat scenario When considering the proportion of coastline within each exposure category for each LA, the no-habitat scenario results in notable increases in “Very High Exposure” levels (Fig. 9). In Meath, this category raises to 89%, while Galway City increases to 64%. Dublin City experiences a 15%, increase, reaching 29.5% “Very High Exposure”, and Fingal has a 13% increase. Without habitats, more than half of the coastline in Sligo, Leitrim and Louth is classified as “Very High Exposure” (Table SI-6).Fig. 9Bar chart of the proportion of coastline by exposure category for each Local Authority in the no-habitat scenario Bar chart of the proportion of coastline by exposure category for each Local Authority in the no-habitat scenario Exposed coastal populations and cultural heritage sites As of 2022, 64% of the population of the ROI resided in the 17 coastal LAs assessed in this study (CSO 2023b). The dasymetric mapping approach and Voronoi polygon buffers reveal an estimated 57,367; 137,114; and 308,313 people reside within 125 m; 250 m; and 500 m of the coastline, respectively. Cork County contains the highest numbers of residents within 125 m (n = 11,309) and 250 m (n = 22,800) of the coast. The largest populations within 500 m of the coastline are in Fingal (n = 48,482). The total number of residents in each relative exposure category, for each LA, is presented in Table SI-7. In the no-habitat scenario, coastal populations across all LAs residing within a “Very High Exposure” area increase by 2830 for the 125 m buffer, 7861 for the 250 m buffer, and 19,908 for 500 m buffer. The total sum of registered national monuments located in each buffer distance and exposure category are presented in Table SI-9. Within “Very High Exposure” zones, 780; 1192; and 1797 sites occur at the 125 m; 250 m; and 500 m buffers, respectively. The greatest numbers of sites in these zones are located in Sligo, with 208; 291; and 383 registered monuments in the same three buffers. A total of 191 individual classifications of national monument exist within these zones, with “Midden” occurring most frequently for each buffer distance (n = 114 at 125 m; n = 121 at 250 m; n = 140 at 500 m). In the no-habitat scenario, sites within “Very High Exposure” areas rises to 908; 1393; and 2118 across the three buffer distances, with Sligo again having the highest number of sites (Fig. 11). A total of 201 individual monument classifications exist in this scenario, with middens remaining the most common (n = 114 at 125 m; n = 151 at 250 m; n = 174 at 500 m).Fig. 10Total Coastal Populations within 500 m of a "Very High Exposure" coastline, with habitats (Map A) and in the no-habitat scenario (Map B). Dasymetric mapping results with Coastal Exposure Index results for Galway City (Map C) and Dublin City (MapD)Fig. 11National Monuments within 125 m of a “Very High Exposure” Coastline, sum for each coastal Local Authority with examples of such sites around the coast of Ireland Total Coastal Populations within 500 m of a "Very High Exposure" coastline, with habitats (Map A) and in the no-habitat scenario (Map B). Dasymetric mapping results with Coastal Exposure Index results for Galway City (Map C) and Dublin City (MapD) National Monuments within 125 m of a “Very High Exposure” Coastline, sum for each coastal Local Authority with examples of such sites around the coast of Ireland Discussion RQ1: What is the spatial distribution of relative exposure levels to coastal hazards along the Irish coast? The results of the CVM (Fig. 2) reveal the relative exposure to climate hazards along the ROI coastline is highly spatially variable. However, some broad trends are visible from the results. In general, exposure tends to be higher in the west and northwest, both in terms the CEI value and the overall coastal length. Of the 6695 coastal points deemed “Very High Exposure” (CEI > 3.18), 25% are located in Mayo, 24% in Galway County and 12% in Donegal. The distribution of exposure ranking tends to be negatively skewed for Mayo and Galway County, with coastal points tending towards higher exposure (Fig. 3). The southern county of Cork, and Donegal in the north, have the longest stretches of coastline deemed “Very Low Exposure” (CEI ≤ 2.32), with levels negatively skewed. The CEI results align closely with existing literature on coastal hazard distribution around Ireland. Devoy (1990) for example, highlighted broad regions along the coast which are vulnerable to SLR and erosion, most of which correspond with zones of “Very High” or “High Exposure” in the CEI e.g. Clew Bay in County Mayo (Fig. 12A). Consistency is also apparent at the national scale, as seen in OPW assessments identifying erosion-prone coastlines and wave overtopping hotspots (OPW 2018, 2010). When these zones are segmented and joined to their nearest coastal point, 81% of CEI points within them are classified as “High Exposure” or “Very High Exposure” under the baseline model (Table SI-11). While the OPW assessments are not fully comprehensive, they provide reasonable confidence that the CEI is capturing genuine patterns of hazard exposure. The results were also compared against the GSI Shoreline Change study (Caloca et al. 2025; Monteys et al. 2024), which analysed portions of the coasts of Wicklow and Dublin over the period of 2000–2021. This revealed coastal segments experiencing “Moderate” (0.25–1 m yr−1) and “High” erosion (≤ 1 m yr−1). When compared against the results of the baseline model, 83% of the points deemed “High Erosion” were ranked as “High Exposure” or “Very High Exposure” in the CEI (Table SI-11). This comparison, however, also revealed that 81% of points in regions experiencing net accretion (≤ 9.56 m yr−1) were likewise classified as “High Exposure” or “Very High Exposure” in the CEI. This underscores that although the CEI is useful as a screening metric for identifying areas potentially prone to erosion, it cannot capture the complex dynamics of sediment transport along the coast. The CEI maps “High Exposure” or “Very High Exposure” exposure levels in other areas known to be under threat from coastal hazards. The Maharees Peninsula (Fig. 12B) is a tombolo landform on the west coast, in County Kerry and contains several sites where storm-driven erosion has been documented, from both cliffs and sand dunes, with one section of beach having an average retreat rate of 0.63 m yr−1 between 1900 and 2015 (Farrell et al. 2018, 2016). The exposed nature of the area is captured by the CEI, with CEI values ranging from 2.42 to 4.12, and 92% of the peninsula deemed “High Exposure” or “Very High Exposure”. The CEI identifies hotspot locations around the Irish coast where relative exposure to climate hazards is greatest. As shown, many of these hotspots are areas with a documented history of coastal erosion or recurring flooding, reinforcing the validity of the index. The analysis also highlights several potentially lesser-known sites that may not have previously been recognised as highly exposed. By assessing the entire coastline of the ROI in relative terms, this study provides a strategic framework through which national bodies such as the OPW, alongside LAs, can visually identify and prioritise areas for adaptation or protection, as well as locations where intervention may be unnecessary or inappropriate. RQ2: What are the controls on coastal hazard distribution across Ireland? The seven variables comprising the CEI (Fig. 5) reveal distinct spatial trends in the relative distribution of coastal hazards around Ireland. This spatial behaviour reflects the underlying processes generating each variable, the resolution of the input data, the nature of the coastline, and how the model ranks each variable. See Figure SI-4 for Pearson’s and Spearman’s rank correlation matrices of the 7 variables utilised in the baseline CEI. Potential shoreline erodibility, represented by the Shoreline Type variable, exhibits broad regional trends. Hard, rocky coastlines of low erodibility dominate the west and southwest, while the east coast is characterised by beach sands and marine gravels of high, or very high, erodibility. This is consistent with documented shoreline retreat rates and longstanding erosion issues in counties such as Wexford, Wicklow and Dublin (Caloca et al. 2025; Caloca-Casado 2018; Carter et al. 1987; Monteys et al. 2024). Relief exposure exhibits greater spatial variability, although it has a moderate correlation with Shoreline Type (ρ = 0.49), since lower-lying regions are generally comprised of softer sediment. Ireland’s position in the northeastern Atlantic means the majority of storm systems approach from the west and southwest (Bricheno et al. 2025; Camus et al. 2024). Consistent with this setting, the highest wind speeds occur along the west and northwest coasts. Analysis of the CERRA dataset (Copernicus Climate Change Service 2021) shows that 90th-percentile wind speeds predominantly originate from the southwest–northwest sectors, reaching 22.16 m s⁻1 off the southwest coast. High winds also affect the east coast, although generally with lower magnitudes than those observed in the west, and peak winds on the east coast are more commonly associated with southeasterly directions. Relative wave energy exhibits a moderate correlation with wind exposure (r = 0.59), with high-energy swell waves impacting the west coast and lower-energy, locally generated wind waves dominating along the east coast (Gallagher et al. 2014; O’Connell et al. 2024). Despite this, the highly indented western coastline provides considerable local shelter, captured through the fetch ray function in the CVM. This results in many western coastal points exhibiting low potential wave energy, such that the upper quintiles of wave exposure encompass areas influenced by both swell-dominated and wind-wave-dominated conditions. This is fitting, as both wind waves and swell waves during storm events can exhibit comparable erosive or flooding potential, depending on the dominant wave climate affecting a given stretch of coastline (Lobeto et al. 2024). The national trends of Storm Surge and SLR are largely explained by the spatial resolution of the CMIP6 model data, approximately six kilometres at the coast (Copernicus Climate Change Service 2022a). These are large-scale oceanographic and climatic processes that produce similar index ranks across adjacent points over hundreds of kilometres. Although the precise rates of sea-level change for the entire coastline of Ireland are currently unavailable, the modelled rates from the CMIP6 dataset are broadly consistent with known rates of sea level change around the country. For example, a recent analysis of a century of tidal gauge data (1925–2024) by McLoughlin et al. (2026) revealed highest rates of SLR in Cork, in the south of the country, and lowest rates in the north which experiences relatively higher rates of glacial isostatic adjustment (Devoy 2008; McLoughlin et al. 2026; Peltier et al. 2002). A strong negative correlation (r = − 0.73) between the SLR and Surge variables exists (Figure SI-4), with records from the CMIP6 dataset revealing relative exposure to a 100-year RP storm surge event is highest in the north (1.74 m) and lowest in the south (1.00 m). This distribution is in line with previous studies of storm surge residual heights around ROI (Olbert and Hartnett 2010; Wang et al. 2008). From analysis of 48 surge events from 1959 to 2005, Olbert and Hartnett (2010) found that surge heights were lower in the south of the country, along the Celtic sea, due to a negative correlation between wind-speed and pressure driven surges (Olbert and Hartnett 2010). As with SLR, binning the surge values into 5 classes based on quintile ranges results in many LAs being assigned just one rank. More detailed modelling of these variables, particularly for SLR which could include local rates of subsidence, may reveal more heterogenous distributions (Nicholls et al. 2021). The quintile binning of continuous data can also produce strong positive rank correlations, as is the case for the Wind and Waves variables which exhibit a Spearman’s correlation of ρ = 0.9. This collinearity effectively applies greater weighting to coastal points exposed to the prevailing wind and wave direction. This was deemed appropriate for this multi-hazard index, as strong winds and powerful waves represent distinct but equally hazardous coastal processes that may independently drive exposure. However, this assumption may not apply universally, and a cautious interpretation of the CVM output is recommended in this regard. In those locations where extreme winds may not be deemed hazardous, the default CVM should be interpreted with this limitation in mind. The LOO analysis conducted in this study highlights how each variable contributes to exposure, with removal of the Shoreline Type variable producing the largest change across all coastal points (Table 3). This is particularly evident when viewing the distribution of coastal points deemed “Very High Exposure” for each LOO scenario (Figure SI-2), which reveals significant changes to this category for eastern LAs. This suggests the potential erodibility of the coastal substrate is a key component of hazard exposure on the east coast of Ireland. Conversely, removal of the SLR variable often produces no change in the percentage of points classified as “Very High Exposure” for many LAs. The highest change for this variable is in Cork, which is as expected given this is where rates are highest. Removal of Wind or Waves produces similar results, which is largely attributed to their shared spatial distribution. Notably, few LAs exhibit significant changes in “Very High Exposure” coastal points when Coastal Habitats are removed, underscoring the highly modified nature of much of the Irish coastline. The alternative variables employed within this assessment were selected based on a critical appraisal of the CVM methodology and results. Alternating just one or two variables can lead to significant changes, which was particularly apparent when comparing the baseline model (CEI_baseline) against the default InVEST® output (CEI_CVM) This is largely attributable to the negative correlation between the surge proxy variable and the modelled CMIP6 surge residual levels. Changes to the methodology employed when deriving the Relief variable produced some significant changes in certain LA’s, notably in Mayo and Donegal which retained 40% and 45%, respectively, of points deemed “Very High Exposure” when the CVM Relief was used. This localised reduction suggests the default CVM Relief variable may underrepresent very low-elevation regions, and we recommend that future versions of the CVM adopt the same methodology used in the baseline model when calculating Relief exposure. Assessment of the alternative ranking methods reveals the sensitivity of index-based approaches to changes in these classification styles. The results appear to support the quintile ranking approach, since many of the continuous variables are positively skewed (Figure SI-5). As a result, ranking these via the equal interval and natural breaks methods produce disproportionately large numbers of coastal points assigned “Very Low Exposure” or “Low Exposure” ranks. These methods under represent potential exposure along the east coast, despite this being a region that experiences significant coastal hazards in places, including some of the highest reported shoreline retreat rates nationally (Caloca et al. 2025; Monteys et al. 2024). These findings reinforce the validity of the CVM's default quintile ranking method, particularly for Wind, Waves, and Relief. However, this approach may be less suitable for variables with more normal or multimodal distributions, such as storm surge. For SLR it may introduce errors, since SLR rates (particularly when derived from global models) can be relatively uniform over large distances. Under these conditions, small differences in SLR rates can result in substantial shifts between quintile classes despite having little practical significance. Where SLR rates are known to be spatially uniform, excluding this variable from the index may be more appropriate. Alternatively, applying user-defined SLR thresholds based on meaningful physical or management criteria, rather than statistical class breaks, may provide a more robust ranking. While the sensitivity analysis (and comparison with regional studies) strengthens confidence in the reported findings, effective interpretation of the CEI nonetheless requires local knowledge, as geomorphological context and dominant coastal processes must be understood before determining appropriate management responses. By identifying the key controls on hazard exposure along specific coastline sections, the CEI provides a foundation for developing targeted and evidence-based adaptation strategies. RQ3: Where can coastal habitats reduce relative exposure and mitigate against the potential impacts of these hazards? The protective function of coastal habitats are most commonly discussed in terms of wave attenuation, shoreline stabilisation, and flood or inundation mitigation (Arkema et al. 2013; Barbier 2016; Narayan et al. 2017; Shepard et al. 2011; Spalding et al. 2014). Dunes have long been associated with coastal defence (Hanley et al. 2014). Larger dune systems, such as “White Dunes” which may reach elevations of 10–30 m have significant wave attenuation effects (Jordan and Fröhle 2022). Seagrass meadows also contribute to wave and current attenuation, especially in shallower waters where the plants occupy a large proportion of the water column (Jordan and Fröhle 2022). In some instances, combined seagrass and kelp forest habitats have been shown to reduce wave heights by 25–45% (Narayan et al. 2017). Saltmarsh, seagrass, and dune vegetation also enhance shoreline stability. Plant roots improve soil cohesion, while in the case of saltmarshes, the vegetation canopy slows water velocity, thereby reducing shear stress near the seabed (Hanley et al. 2014; Jordan and Fröhle 2022). Saltmarshes and dune habitats can also autonomously adapt to SLR, provided sediment supply is adequate and there is sufficient landward space for the habitat to migrate (Cott et al. 2012; Hanley et al. 2014; Nicholls and Klein 2005; Pethick 1993). The CEI reveals 3299 km of the coastline of the ROI potentially benefits from habitat protection, with areas containing two or more classes of habitat experiencing the greatest reduction in exposure values in the CEI. An example of such an area is surrounding the Belmullet Peninsula in County Mayo (Fig. 13). This region, which contains some of the highest CEI values nationally, experiences reduction in relative exposure levels of up to 0.83. CEI reduction is most pronounced surrounding a large dune system, which is backed by saltmarsh and seagrass habitats which provide lower levels of CEI reduction. The no-habitat scenario highlights that in relative terms, coastal habitats have a significant impact in reducing CEI levels in the eastern LAs of Dublin City and Fingal. In Dublin City, the largest effect is in the area surrounding Bull Island, a 5 km long sand spit which formed in the last two centuries as a result of the accumulation of sediment following the construction of two large walls surrounding the entrance to Dublin Port (Mathew et al. 2019). Bull Island contains both saltmarsh and dune habitats, and is designated a Special Area of Conservation under the E.U. Birds Directive (NPWS 2014). The area is backed by a relatively densely populated urban area, and with SLR rates (from 1997 to 2016) in the region estimated at 6.48 mm yr−1 (Shoari Nejad et al. 2022), the presence of a healthy, accreting saltmarsh could have a substantial effect in reducing exposure to flooding and inundation (Möller and O’Leary 2025). The no-habitat scenario also highlights the risk posed by habitat degradation. Many of Ireland's coastal habitats are in unfavourable condition (OPW 2022), and coastal habitats in particular are subject to numerous anthropogenic and climate pressures (Casal and McCarthy 2023). The examples previously stated in the Maharees Peninsula, County Kerry (Fig. 12B); and Burrow Beach, County Dublin are dune systems which have been degraded in recent years, largely driven by human activities that have led to vegetation loss (Farrell et al. 2016; Stori and O’Mahony 2021). The consequences of such vegetation loss can be considerable; for instance, Rossbeigh beach Co. Kerry, experienced a 43.2% reduction in vegetated dune cover between 2001 and 2019 (Nuyts et al. 2020). Employing coastal habitats as NbS for coastal defence offers advantages over hard engineered structures alone. Unlike fixed infrastructure, habitats can adapt to changing conditions; saltmarshes can adjust to rising sea levels where sediment supply is adequate, and dune systems can recover from storm erosion within relatively short timescales (Kraus 1996; Spalding et al. 2014). Hybrid or tandem approaches to coastal protection may also be effective, evidence suggests that engineered structures fronted by saltmarsh experience reduce maintenance costs, with estimated savings of €11.5 million per year (Norton et al. 2018). With approximately 38% of the ROI’s coastline experiencing some degree of exposure reduction due to coastal habitats, and with this reduction often quite pronounced in urbanised areas, the preservation and enhancement of these habitats as NbS for coastal protection should be a key management priority. RQ4: What populations and cultural heritage sites are located in highly exposed coastal regions? The CEI (Fig. 2) reveals hotspots within each LA, revealing where coastal climate hazards are most likely to have adverse impacts. To assess who and what is exposed, total population and cultural heritage sites within 125 m, 250 m and 500 m of “Very High Exposure” coastal points were analysed (Figs. 10, 11). Collectively, 22.5% of Ireland's population resides in Dublin's three coastal LAs, with an estimated 20,062 people in these LAs residing within the 500 m “Very High Exposure” buffer zone (Fig. 10D). Key hotspots include Sandymount (n = 5209) and Clontarf (n = 4329) in Dublin City; Sutton and Baldoyle in Fingal (n = 4683); and Blackrock in Dun Laoghaire-Rathdown (n = 1483). Galway City on the west coast (Fig. 10C) contains 9971 people—~ 11.4% of the city’s population—residing in the 500 m “Very High Exposure” buffer zone. Population numbers in this zone increase across all LAs (except Leitrim) for the no-habitat scenario (Table SI-8), with a 81% increase in Fingal, where habitats—such as the dune systems at Burrow Beach and Portmarnock—potentially reduce exposure for ~ 7297 people. While urban areas dominate in terms of exposed populations, the cultural heritage analysis reveals that remote and sparsely populated coastlines contain significant numbers of exposed assets. Of the 8871 registered cultural heritage sites within 500 m of the coast, 20% (1797) are within “Very High Exposure” zones, with the majority concentrated along the west coast (Table SI-9). A notable hotspot is the uninhabited island of Inishmurray, County Sligo (Fig. 11A), which contains a medieval monastic settlement and 132 registered monuments, including inscribed cross stones (n = 57) and leachts (n = 20), defined as “grave, cairn or sepulchral monument” (National Monuments Service 2023). The no-habitat scenario produces an 18% increase in heritage sites within the 500 m “Very High Exposure” buffer zone, totalling 2118 (Table SI-10), with the highest proportional increases in Wexford (114%) and Fingal (95%). Previous studies, such as the CHERISH project, have highlighted the need for a registry of potentially at risk coastal heritage sites around Ireland (CHERISH 2023; Jones et al. 2023). By extending this assessment to a national scale, the CEI provides a consistent framework for prioritising exposed heritage assets and identifying sites warranting further investigation (Fig.11). Fig. 12Close-up examples of exposure levels at four different sites around Ireland. Area surrounding Clew Bay and Achill Island, County Mayo (Map A). Dingle Peninsula, Co. Kerry (Map B).Wicklow Head to Arklow, County Wicklow (Map C). County Cork coastline (Map D) Fig. 13Potential reduction in exposure to climate hazards by coastal habitats—example from Belmullet Peninsula, County Mayo Close-up examples of exposure levels at four different sites around Ireland. Area surrounding Clew Bay and Achill Island, County Mayo (Map A). Dingle Peninsula, Co. Kerry (Map B).Wicklow Head to Arklow, County Wicklow (Map C). County Cork coastline (Map D) Potential reduction in exposure to climate hazards by coastal habitats—example from Belmullet Peninsula, County Mayo Reducing the impacts of climate hazards on coastal populations and culturally significant sites along the ROI coastline requires careful planning and the targeted allocation of limited resources. As outlined in the Strategic Pillar 3 of National Coastal Change Management Strategy Report (DHLGH and OPW, 2023) five categories of response have been identified: Do Nothing; Hold the Line; Managed Realignment; Managed Retreat; and Advance the Line. For many heritage sites, where preventing long-term loss may be costly and ultimately unavoidable, non-intervention may prevail. For coastal populations, the need for intervention is more pressing, as inaction risks lives and livelihoods. As a screening tool, the results of the CEI can aid coastal managers in making these decisions. It should not however, be used in isolation, as the population numbers and cultural heritage sites reported here are based on primarily on proximity to “Very High Exposure” coastlines. Site specific information on factors such as hazard probability, flood pathways, existing coastal defence standards, land ownership, social vulnerability, projected costs, and governance feasibility, should also be included when selecting a management response. It should be noted that exposure alone does not equate to vulnerability. Some areas may benefit from functional coastal defences, while others may possess greater socioeconomic capacity to anticipate and recover from hazard events (Cutter et al. 2012; Fitton et al. 2022). While the CEI provides a robust national-scale assessment of relative exposure, integrating demographic and socioeconomic variables in future assessments would move beyond exposure towards a more complete evaluation of coastal vulnerability, strengthening its value as a decision-support tool for coastal management. Limitations Whilst every attempt was made to ensure the most accurate and up-to-date data was used when completing the CEI, some uncertainties remain due to limitations in the underlying input data, and the methodologies employed. The quaternary sediments layer (GSI 2021), for example, provides the most comprehensive national-scale representation of shoreline type available. The exposure ranks assigned to this reflect the potential erodibility of the coastal substrate rather than observed erosion. No national database of erosion rates currently exists for Ireland, meaning comparison of the rankings is limited to regional or localised studies of documented erosion. Such studies rely on remote sensing techniques such as aerial photography (Chalençon et al. 2025) or satellite imagery (Caloca et al. 2025; Monteys et al. 2024; Riaz et al. 2026; Senatilleke et al. 2026) and utilise the Digital Shoreline Analysis System (DSAS) developed by United States Geological Survey (USGS) (Himmelstoss et al. 2021). Comparing the results of these studies with the shoreline type layer provides some degree of confidence in the exposure rankings. For example, comparison of the Shoreline Type layer against results of the GSI Shoreline Change study (Caloca et al. 2025; Monteys et al. 2024) reveals that for High Erosion (≤ 1 m yr−1) sites, 72% and 10% are classified as “Marine beach sands” and “Windblown sands” respectively, with erosion rates up to − 4.97 m yr−1 reported (Caloca et al. 2025; Monteys et al. 2024). The same comparison, however, highlights local inaccuracies often arising from minor misalignments between the coastline used in the Quaternary Sediments dataset, and the coastline applied in this assessment, which may misrepresent the true nature of the coast in some areas. Due to the limited temporal range and pixel resolution of satellite or aerial imagery, shoreline change studies are typically limited to soft coasts, leading to erosion rates on rocky coasts around Ireland remaining poorly constrained. However site specific rocky-cliff studies reporting net rates of 0.0037 m yr−1 (Cullen 2019), and previous national estimates as low as 0.0001 m yr−1 (Devoy 2008), indicate reduced exposure to erosion on rocky coasts. This, however, will vary considerably with lithology, structural weaknesses, and orientation relative to dominant wave directions. Differentiating rock types at a national scale is therefore limited by this uncertainty, though such distinctions, alongside localised sediment accretion, may be more appropriately captured through regional or local assessments. In addition, there is currently no comprehensive, up-to-date national inventory of coastal protection infrastructure in Ireland. The absence of this information may lead to an overestimation of exposure in areas where effective engineered defences are already in place. In line with previous studies (e.g. Gornitz 1991; Mclaughlin and Cooper 2010) the underlying substrate was utilised, rather than the descriptive geomorphological category, since many of these categories correlate with separate variables already included within the InVEST® CVM. In previous CVI studies, geomorphology is typically included as an individual variable, including elements such as “Indented coasts”, “Low cliffs”, and “salt marsh” (Hammar-Klose and Thieler 2001; Koroglu et al. 2019; López Royo et al. 2016; Pantusa et al. 2022; Parthasarathy and Natesan 2015; Pendleton et al. 2010). Such categories may be appropriate in individual CVI studies, however applying such categories in the CVM risks an over-representation of certain variables, such as elevation. Together with the habitat type, the geology of the coast and it’s relief essentially combine to define its geomorphology. The indented nature of the coast, and it’s aspect in relation to incoming storm waves and winds, is also accounted for through the use of fetch rays. Therefore, we would suggest that future studies employing the CVM use shoreline type, or coastal substrate, rather than geomorphology, and the variable nomenclature be revised in the CVM documentation. The CVM requires the application of generic protective distances and qualitative rankings for each coastal habitat class. These values are necessarily simplified and may be somewhat arbitrary, as protective capacity can vary substantially between coastal habitats (Koch et al. 2009). For example, the protection provided by dunes is a function not only of their height and width, but also other physical parameters such as their aspect ratio, the foredune slope, the vegetation density and maturity, and the width of the beach below the dune, along with other factors such as storm duration and intensity (Itzkin et al. 2021; Sigren et al. 2014; Toledo et al. 2024). Although saltmarshes are known to have wave and surge attenuation and shoreline stabilisation properties, this may be limited to large saltmarshes (approximately 10 km or greater in width), particularly during storms (Leonardi et al. 2018). The density and health of the saltmarsh vegetation, channels within the saltmarsh, and surge magnitude are also highly influential factors (Leonardi et al. 2018; Shepard et al. 2011). Seasonality can also impact protective effects, with seagrasses shown to have higher wave attenuation properties when the plants are flowering (Chen et al. 2007). The application of protective distances uniformly in all directions may also overestimate habitat influence in areas that are adjacent to, but not directly protected by, coastal habitats. Furthermore, this approach assumes that all mapped habitats are in good ecological condition and capable of providing protective functions. Unfortunately, this is not the case in Ireland with recent national habitat assessments by the National Parks and Wildlife Service (NPWS), completed under Article 17 of the EU Habitats Directive (European Council 1992), indicating the overall conservation status of coastal habitats such as Marram Dunes and Mediterranean salt meadows as “Unfavourable-Inadequate”, and “Unfavourable-Bad” for habitats such as Atlantic salt meadows (NPWS 2025). The status of individual sites is not reported however and as such was not included in this assessment, however in locations where habitats are degraded, the no-habitat scenario may provide a more realistic representation of current exposure. In either scenario, caution is recommended when interpreting the results, and detailed, localised studies are required to provide a full accounting of coastal protection for each habitat. The majority of spatial data used to represent coastal habitats were sourced from the NPWS. However, this dataset is not exhaustive. For example, there is currently no national spatial dataset describing the extent and distribution of kelp forests around the Irish coastline. Research by Schoenrock et al. (2020) on kelp forests suggests they may be widespread. As with seagrass habitats, kelp forests are known to attenuate wave energy (Arkema et al. 2013; Jordan and Fröhle 2022). Should suitable spatial data become available, the inclusion of kelp, or other habitats within the CEI would further improve the representation of natural coastal protection. The InVEST® CVM offers a readily accessible and widely applicable framework for screening coastal exposure at a national scale. However, certain model outputs may be overly simplified. For example, calculation of the storm surge potential relies solely on the distance from the coast to the edge of the continental shelf. This limitation has been noted by previous studies (Al Ruheili and Boluwade 2023; Ballesteros and Esteves 2021). The storm surge output from the CVM was not utilised in this study and was instead replaced with exposure rankings derived from quintile ranges of the CMIP6 reanalysis data (Copernicus Climate Change Service 2022b). However, in terms of hazard exposure, the most extreme water levels arise from the co-occurrence of a storm surge and a high spring tide (Cotrim et al. 2026; Olbert et al. 2013; Vousdoukas et al. 2016). Omitting tidal influence therefore risks misrepresenting flooding exposure during storm events, particularly in macro-tidal environments characteristic of Ireland's west and northeastern coasts (Devoy 2008). Sensitivity testing of the model when including the TWL variable (Table 4) reveals only a marginal change in the overall CEI values when correlated with the baseline model (r = 0.96), reflecting the moderate to strong correlation between the ranks of the CMIP6 storm surge levels and the TWL variable (ρ = 0.69). Nevertheless, we recommend that future applications of the CVM critically evaluate the storm surge component by comparing results with modelled surge levels, or total water levels, where available. It should also be noted that this study does not account for river discharge or floodplain connectivity, as would be required for a comprehensive coastal-flood risk assessment (Ganguli and Merz 2019). The use of quintile-based classification to derive exposure rankings represents a further limitation. This approach was deemed to be appropriate for Ireland, as it facilitates relative comparison across the entire country. However, it does not account for the varied contributions of individual variables to overall exposure, as recognised by authors who have adopted customised class boundaries (Ballesteros and Esteves 2021). In addition, quintile classification constrains comparisons of CEI results across different temporal scales. For example Re et al. (2023) compared the CEI results from two outputs, using wind and wave records from two different time periods. Although these records were different for all coastal points in the two scenarios, the quintile ranges of the outputs were broadly similar, resulting in little change to overall exposure levels (Re et al. 2023). Comparable testing conducted in this study using projected rates of sea-level rise and storm-surge levels produced similar outcomes, indicating that this index-based approach is not well suited to direct comparison across multiple time horizons. Conclusion This study provides the first national-scale index of relative exposure to climate hazards along the coastline of the ROI. The results demonstrate that exposure is highly site-specific and controlled by a combination of shoreline type, metocean forcing and climatic variables. Exposure patterns around Ireland therefore cannot be simplified to a binary distinction between a sheltered east coast and an exposed west coast. While it is generally true that Atlantic storm systems exert a greater influence on the west, and that many highly exposed cultural heritage sites are concentrated in western counties such as Mayo, Sligo, Galway, and Kerry, the analysis reveals that eastern LAs also contain disproportionately high levels of exposure relative to their coastline length. These are also the most densely populated coastal regions with eastern LAs of Dublin City, Dún Laoghaire–Rathdown, Fingal, Louth, Meath, Wexford, and Wicklow, despite comprising just 8% of the ROI coastline, account for 44% of coastal populations residing within 500 m of a “Very High Exposure” coastline. The results highlight that it is the nature of what and who is exposed that ultimately defines coastal risk. In terms of population exposure, portions of Galway City and Fingal emerge as particularly significant hotspots. Approximately 61% of Galway City’s 33.5 km coastline is classified as having “Very High Exposure”, with almost 10,000 residents within 500 m of these coastal areas. The potential susceptibility of this area to coastal hazards was underscored during Storm Éowyn in January 2025, when the city narrowly avoided widespread flooding during a 2.6 m storm surge, which happened to coincide with a mid-tide (University of Galway 2025). Fingal, in County Dublin, contains over 16,000 residents in the 500 m buffer zone for the no-habitat scenario, with over 7000 people potentially benefitting from exposure reduction by coastal habitats. However, significant portions of Fingal’s coastline are experiencing rapid erosion rates (Caloca-Casado 2018; Stori and O’Mahony 2021), and with coastal habitats nationwide facing increasing anthropogenic and climatic pressures, the conservation and restoration of these habitats, alongside appropriately designed engineered defences, represents a critical opportunity to reduce future risk in this region. Overall, this screening assessment delivers a robust and detailed resource for coastal managers and planners in Ireland. It also demonstrates the wider applicability of the InVEST® CVM to other coastal regions, despite the model's potential limitations. Integrating the CEI with demographic, socioeconomic, and governance factors would enable future studies to progress beyond relative hazard exposure towards a more comprehensive assessment of coastal vulnerability and risk. As climate-related hazards are projected to intensify, such a resource will be increasingly valuable in supporting actionable, equitable, and sustainable responses to ongoing coastal change. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1