Exploring the relationships between coastal vulnerability and socio-economic variables in the context of climate change using geoinformatics for the region of Longyearbyen, Svalbard

Exploring the relationships between coastal vulnerability and socio-economic variables in the context of climate change using geoinformatics for the region of Longyearbyen, Svalbard Evangelia Tita, George P. Petropoulos * , Danai Varvatsouli Department of Geography, Harokopio University of Athens, El. Venizelou 70, Kallithea, Athens, 17676, Greece ARTICLE INFO Keywords: Coastal vulnerability index Socio-economic vulnerability index Integrated coastal vulnerability index Coastal vulnerability Longyearbyen ABSTRACT The continuous monitoring of coastline vulnerability is of major importance, due to the significant human activity and vital ecosystems, like, permafrost wetlands, and sea ice-habitats in these areas. The Arctic Circle, in particular - where climate change effects are more pronounced - faces accelerated sea ice melting and permafrost thawing, altering coastal erosion. One widely applied method for monitoring coastal vulnerability is the Coastal Vulnerability Index (CVI). Another index that assesses the socio-economic impact on these areas is the Social Vulnerability Index (SVI). Combined, they form the Integrated Coastal Vulnerability Index (ICVI), identifying the most vulnerable areas. This study explores the use of the above indices for the area of Longyearbyen, a town on Svalbard located in the Arctic Circle. In particular, two methods were applied to calculate the indices: the classical approach, using established equations and equal weighting to all factors, and the Analytic Hierarchy Process (AHP) which calculates statistical weights for variables involved in the assessment. Since research linked to integrating socio-economic factors into coastal vulnerability assessments remains limited, particularly in the Arctic, the results of this study provide an important contribution to filling this gap and to better understand the impacts of climate change in relation to the socio-economic variables for Arctic coastal regions. Also, by integrating the ICVI, this study contributes towards efforts to develop informed and more robust coastal management strategies, which in turn facilitate sustainable adaptation to climate change impacts in the Arctic. It aims to foster more informed coastal management strategies and climate change adaptation in vulnerable Arctic coastal areas. 1. Introduction Arctic regions serve as vital reference points for understanding the extensive impacts of climate change. A key focus in these regions is the monitoring of coastlines, particularly in relation to their vulnerability to erosion (Jang et al., 2021). Monitoring coastline vulnerability to erosion has become increasingly crucial over the years, as the effects of climate change become not only more evident but also catastrophic, often resulting in human losses (Calvin et al., 2023). Coastal areas also host invaluable ecosystems and cultural This article is part of a special issue entitled: Earth Observation of the Cryosphere published in Remote Sensing Applications: Society and Environment. * Corresponding author. E-mail address: [email protected] (G.P. Petropoulos). Contents lists available at ScienceDirect Remote Sensing Applications: Society and Environment journal homepage: www.elsevier.com/locate/rsase https://doi.org/10.1016/j.rsase.2025.101642 Received 4 May 2025; Received in revised form 2 June 2025; Accepted 22 June 2025 Remote Sensing Applications: Society and Environment 39 (2025) 101642 Available online 24 June 2025 2352-9385/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). heritage sites, highlighting the importance of socioeconomic consequences of climate change with physical processes (Cabana et al., 2023). However, despite its importance in becoming more widely acknowledged, limited research has been conducted on Arctic coastal vulnerability. That research gap presents an opportunity for further investigation and ongoing monitoring needs to be conducted to address coastal vulnerability risks. It also contributes to the decision-making of local authorities, enabling them to implement necessary measures to minimize and prevent the impacts of various factors (Cruz Ramírez et al., 2024). Depending on the ecological, geomorphologic, and socioeconomic features of a place, coastal erosion can vary greatly in intensity and influence. The scale at which vulnerability assessments are carried out is divided into global, regional and local scales (Nguyen et al., 2021; Nath et al., 2021; Cruz-Ramírez et al., 2024). At a global scale, an overview of relative vulnerability levels is obtained (Dada et al., 2024). Regional assessments such as those discussed by Nguyen et al. (2021), integrate both physical and socio-economic features to assess erosion risks, in regional coasts in Vietnam. On the contrary, Cruz-Ramírez et al. (2024), underlined a more detailed investigation on local-scale assessments, focusing on key elements such as infrastructure exposure and social vulnerability, to inform targeted coastal management strategies, emphasizing the importance of coastal management. The advancement of coastal vulnerability reflects the growing imperative to assess and mitigate erosion risks on coastal environments (Nath et al., 2024). One widely accepted method for assessing coastal vulnerability is the use of indices, which are predicated on the idea of reducing multivariate, complex environmental data into more easily interpreted formats, according to Boumpoulis et al. (2025). The Coastal Vulnerability Index (CVI) stands out as a prominent tool for evaluating coastal vulnerability, which combines coastal variables. Subsequent refinements have expanded its applicability by incorporating diverse environmental variables. Various studies like Nath et al. (2024), Widura and Mardiatno (2022), Hzami et al. (2021), have highlighted the need to incorporate socio-economic variables leading to the development of the Socio-economic Vulnerability Index (SVI) and, more recently, the Integrated Coastal Vulnerability Index (ICVI). While these indices have significantly contributed to the overall assessment of coastal vulnerability, they have not yet been applied in studies concerning the Arctic region. Coastal variables like geomorphology, coastal slope, and sea level rise are mainly integrated into CVI. Bera, R. and Maiti, R. (2021) highlighted the importance of choosing appropriate variables for CVI based on local characteristics, while Pantusa et al. (2022), included additional variables such as hydrology for vulnerability assessments. However, despite the importance of socio-economic factors, CVI does not take them into consideration. To address this limitation, the ICVI, which integrates socio-economic factors and CVI, has been used by Cruz-Ramírez et al. (2024) and Hzami et al. (2021), with the second incorporating socio-economic factors into the assessment of coastal vulnerability. While studies such as those by Jang et al. (2021) and Toumasi et al. (2024), have utilised the CVI in assessing coastal vulnerability in the Arctic regions, no previous studies have incorporated socio-economic factors in their analysis of coastal vulnerability. The ICVI, which combines CVI and SVI, provides an additional thorough approach to identifying vulnerable areas to erosion based on both environmental and socio-economic perspectives. Research in Arctic coastal regions highlights unique environmental dynamics — including permafrost thaw, glacial retreat, and sealevel rise — all contributing to accelerated shoreline changes (Jang et al., 2021; Nath et al., 2023). These conditions, coupled with limited accessibility, have led to a growing reliance on satellite remote sensing and Geographical Information Systems (GIS)-based methodologies to analyse and visualize coastal risks with greater efficiency and scale-adaptability (Tsai, 2024). Remote sensing, such as satellite images, provides valuable information for monitoring coastal changes like shoreline erosion, permafrost thaw, and glacial retreat, as they offer both coverage and spatial resolution, contrary to ground-based technologies (Bera R. & Maiti R., 2021; Jang et al., 2021; Rehman et al., 2022). Moreover, GIS allows the spatial analysis, modelling, and integration of several datasets, improving remote sensing utilization and presenting a comprehensive understanding of coastal vulnerability (Liu et al., 2024). This integration allows monitoring extended environmental changes with higher accuracy, essential for coastal dynamics in Arctic regions (Tsai, 2024), and coastal management (Cruz-Ramírez et al., 2024). In Arctic regions, vulnerability is notably driven by the interplay between geomorphological features and the retreat of permafrost in conjunction with sea-level rise. As more coastlines previously covered with ice are now exposed to storm wave energy, the thawing of permafrost is accelerating, presenting rapidly increasing erosion rates (Hzami et al., 2021; Jang et al., 2021). A significant example of these processes can be observed in Svalbard, a significant area of the European Arctic that offers a valuable tool for studying environmental and climate change. The coastline of the area is particularly vulnerable, making it an ideal case for addressing the effects of glacial retreat, sea ice loss, and the overall coastal morphology and sea level dynamics (Jang et al., 2021; Boumpoulis et al., 2025). Previous studies at local and regional scales have demonstrated that Svalbard’s coastline morphology is highly susceptible to these factors, contributing altogether to the region’s vulnerability to climate changes (Jang et al., 2021; Boumpoulis et al., 2025). To better understand and measure these changes, more recent studies have emphasized the use of satellite imagery and GIS to acquire geological and geomorphological data (Hzami et al., 2021; Toumasi et al., 2024). These tools assist in the creation of coastal vulnerability indicators that contain three main types of variables: coastal characteristics, such as solid geology, shoreline type, elevation and river mouths, coastal forcing factors such as wave height, tidal range and storm frequency, and into socio-economic factors such as population density, road network, and land use. While there are previous studies that have applied either CVI or SVI, there are no previous studies integrating both into ICVI, using satellite images. Additionally, there is also a research gap in the comparison between these two indices. In purview of the above, the present study aims at examining the coastal vulnerability in the Arctic regions using satellite imagery from Sentinel-2 (2017–2023), and GIS, the region of Longyearbyen, Svalbard. This was achieved by integrating the CVI and SVI. The latter allows creating the Integrated Coastal Vulnerability Index (ICVI), which presents an integrated approach that merges physical and socio-economic factors into a single integrated indicator to assess overall vulnerability (Campos et al., 2024). The purpose of this approach is to further understand the climate risks, while it was the first time applied in this area. In this context, to highlight more effectively areas at higher risk under the influence of climate change, two different methodologies were applied to the indices, a E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 2 classical approach and an Analytical Hierarchy Process (AHP) method. The comparison between the classical and AHP approach serves, also, as a novel contribution to Arctic coastal vulnerability research, as no prior studies, to our knowledge, have conducted this comparison. Indeed, although studies like Araujo and Dias (2021), Nath et al. (2025) and Arda et al. (2025), have conducted comparisons, no previous studies regarding this comparison in the Arctic have been computed before. The first provides a more straightforward framework, while the second offers a more structured, weighted comparison of criteria based on expert judgement, providing a more comprehensive objective vulnerability assessment. Therefore, the present study contributes significantly to Arctic coastal vulnerability ongoing research by providing an innovative assessment of coastline impacts, integrating both physical and socioeconomic factors into a single indicator for a more comprehensive and detailed assessment. 2. Experimental setup 2.1. Study area Longyearbyen, located on the western coast of Spitsbergen in the Svalbard archipelago, was chosen as the study area due to its strategic significance within the European Arctic region and its high vulnerability to climate change (Fig. 1). The environmental sensibility and its socio-economic activity make it an ideal case for applying an integrated vulnerability approach. As one of the northernmost permanently inhabited settlements, Longyearbyen offers a rare opportunity to assess the interaction between natural coastal processes and human presence under rapidly changing climatic conditions (Winther and Gudmestad, 2023). The area is already experiencing an increase in temperature and precipitation, with shifts becoming more noticeable during winter (Isaksen et al., 2007; Hanssen-Bauer et al., 2019). Anticipated thawing of permafrost would accelerate, increasing the rates of sediment movement and erosion, with significant impacts on human activities and coastal zone (Jask´ olski et al., 2018). Thus, the area is a crucial example of research and how human activities and coastal vulnerability interact with climatic changes. Longyearbyen’s population rises to 2600 (Statistics Norway, 2024), with most residents employed in the coal mining industry, and research. The local economy depends heavily on tourism, which has been growing consistently since 1990. A limited coastal zone and the Longyearelva River are home to the town’s infrastructure, which includes roads, buildings, and the airport. These areas are vulnerable to geo-hazards such as floods, avalanches, rock-falls, and debris flows (Etzelmüller et al., 2000; Bogen and Bønsnes, 2003). Svalbard’s geomorphology is characterized by widespread glacial cover, covering around 60 % of the land area. The region’s Fig. 1. Map of study area. (Basemap used: Sentinel 2 b satellite imagery, August 2, 2017). E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 3 topography is characterized by rocky mountains and deep fjords, shaped by Longyearbreen and Larsbreen, which together constitute a major system in charge of erosion, sediment, and deposition, feeding the town’s main river, Longyearelva (Etzelmüller et al., 2000). With a thickness of 200–450 m (Etzelmüller et al., 2000), permafrost is a prominent geophysical feature made up of frozen marine sediments (Hanssen-Bauer et al., 2019). Also, the flat-lying sedimentary bedrock, particularly sandstone, siltstone, and shale, which are soft and fine-grained and therefore more prone to erosion, makes up the majority of the lithological composition (Etzelmüller et al., 2000; Jang et al., 2021). 2.2. Datasets The datasets for this study include spatial and thematic data sources, valuable for the estimation of coastal vulnerability in Longyearbyen area. Spatial data includes the Sentinel-2 imagery used for the extraction of the area’s coastline. Accordingly, thematic data includes environmental variables, which have been used for calculating the (CVI), and socio-economic variables for calculating the (SVI). For each data category, the appropriate pre-processing steps were performed to ensure proper spatial analysis. The combination of these datasets formed the framework for assessing physical and socio-economic variables determining the coastal vulnerability of the area. More details about the datasets used in the computation of each index are provided below. 2.2.1. Coastal vulnerability index (CVI): datasets A total of six variables were selected for the calculation of the Coastal Vulnerability Index (CVI), enabling a thorough examination of its coastal vulnerability and forming the basis for the analysis (Roukounis and Tsihrintzis, 2022; Loinenak et al., 2015). More specifically, they were divided into coastal characteristics and coastal forcing. Coastal characteristics include geomorphology, coastal slope, and shoreline erosion or accretion, while the coastal forcing category includes mean wave height, relative mean sea level change, and mean tidal range. The lithological categories included marine and glaci-fluvial deposits, shale, siltstone, and sandstone, which are prone to erosion. Geomorphological data for Longyearbyen were obtained from the Norwegian Polar Institute dataset (accessed March 3, 2024), with a spatial resolution of 1: 250,000. Although this dataset offered the most relevant available data, it was recognised that its relatively coarse scale might not fully match the detail needed for a regional analysis. For that reason, lithological information and erosion-prone classifications were validated through comparisons with previous studies (Etzelmüller et al., 2000; Jask´ olski et al., 2018). Coastal slope, which indicates a coastline’s resistance to erosion, is another crucial factor. According to L´ opez Royo et al. (2016), softer slopes are more fragile, whereas steeper slopes are often less so. The slope data used in this research were derived from the Terrengmodell–Svalbard Digital Elevation Model (DEM), with a 20-m resolution, provided by the Norwegian Polar Institute. The DEM was generated using stereo models from aerial imagery and is periodically updated. An alteration may occur in glacier regions, but generally the standard error is 3–5 m. In addition to the above, in this study, Sentinel-2 images from the Copernicus program was exploited between 2017 and 2023. More specifically, seven satellite images were evaluated during the winter season, as then there is minimal snow and cloud cover. Due to insufficient quality for accurate shoreline extraction, images from 2018 were excluded. Level-1C Top Of the Atmosphere (TOA) reflectance images from the MultiSpectral Instrument (MSI) were used. The shoreline from 2017 was set as the reference baseline to compute erosion and accretion rates, which were processed using the Digital Shoreline Analysis System (DSAS), a USGS extension tool in ArcMap. Other factors contributing to the coastline’s susceptibility are under the category of coastal forcing. According to Widura & Mardiatno (2022), geomorphological variables, significant wave height, and sea level rise have the most significant influence on vulnerability. Thus, the three variables chosen for Longyearbyen were mean tidal range, mean wave height, and relative sea level change. The numbers chosen for this analysis, according to Jask´ olski et al. (2018b), were 0.8 m for mean wave height, −1.76 mm/year for sea level change, and 1.5 m for tidal range. The CVI equation consistss of six variables, and the proper choice of them is often related to the study area and the spatial scale. However, more variables can be taken into consideration, such as dunes and emerged beaches, river discharge, and vegetation, which played a crucial role in coastal vulnerability in Pantusa et al. (2018 ; 2020). 2.2.2. Socio-economic vulnerability index (SVI): datasets Socioeconomic factors also play an essential role in understanding coastal vulnerability in the area. Given Longyearbyen’s economic shift and growing exposure, four variables were included in the Socio-economic Vulnerability Index (SVI): land use and land cover, road network, settlements, and population density (Jask´ olski et al., 2018; Sokolickova et al., 2022). Regarding this, land use and land cover are key factors in managing responses to hazard exposure, as they showcase environmental, cultural, and economic patterns (McLaughlin et al., 2002). For this purpose, the ESA WorldCover 2021 raster dataset was used, providing a 10-m resolution and covering 11 distinct land cover classes, which was accessed on March 3, 2024. Roads and settlements were also used as indicators of human presence and infrastructure, as the first one is necessary for supply chains and transportation, and the second one to describe the size and absence of a place, two factors crucial for determining how much human exposure there is to coastal dangers (McLaughlin et al., 2010). These data were also accessed on March 3, 2024. A final factor studied in this research is population density, which measures population per square kilometer, to reflect the human aspect of vulnerability. The population statistics of Longyearbyen that were relevant to the study period, were acquired from Statistics Norway’s official records, which are freely accessible at https://www.ssb.no/en. The contribution of SVI is an additional and crucial factor for better understanding the processes that lead to the coastline’s higher E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 4 vulnerability, recognised by many previous studies as those of Mani Murali et al. (2013) & Mahapatra et al. (2015). Moreover, anthropogenic factors have an instant impact on coastal vulnerability, and sometimes even to a greater extent that sea level rise due to climate change (Hzami et al., 2021). 2.2.3. Sentinel imagery Lastly, for the extraction of the coastline, Sentinel-2 images from the Copernicus program were used. Specifically, seven images were used for the period 2017–2023, focusing on data obtained during summer months, as weather conditions (snow coverage and cloud presence) are suitable to monitor Arctic areas. Due to insufficient quality for accurate shoreline extraction, images from 2018 were excluded. Level-1C Top of the Atmosphere (TOA) reflectance images from the MultiSpectral Instrument (MSI) were used. Images were cropped at the desired boundary of the study area, and then resampled, excluding thermal bands from the analysis, and only SWIR and Green bands were used. After the extraction of shorelines for each year, shoreline of 2017 was chosen as the baseline reference to calculate the rate of erosion/accretion with Digital Shoreline Analysis System (DSAS) of USGS extension tool in ArcMap. 3. Methodology In this section, the overall methodology is presented following the data pre-processing steps and main processing. The methodology aims to assess the coastal vulnerability in Longyearbyen through the CVI, SVI, and the resulting ICVI. The procedure consisted of two steps: data pre-processing, and index calculation with two different approaches. Jask´ olski et al. (2018), who computed CVI for the years 1990–2009 which provided a basis for the current study. Fig. 2 presents a flowchart with the methodological steps for mapping coastal vulnerability in Longyearbyen, Svalbard. The methodology integrates CVI and SVI variables processed and ranked before their estimation is conducted by two methods. The AHP is applied as an additional weighting method to refine CVI and SVI estimation. The final ICVI is derived from both classic approach estimations and AHPweighted results, providing a comprehensive assessment of coastal vulnerability. Fig. 2. Flowchart presents the methodological steps for mapping coastal vulnerability in Longyearbyen, Svalbard. E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 5 Fig. 3. Presenting initial satellite imagery, in the preprocessing step, used to extract water bodies and land. Images of Sentinel 2 in RGB color, after the application of processing steps i.e. registration, resampling, and subset. The referring chronological order is (a) 2017, (b) 2019, (c) 2020, (d) 2021, (e) 2022, and (f) 2023. E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 6 3.1. Pre-processing At the pre-processing stage, geospatial datasets were collected and analyzed to produce the variables for CVI and SVI calculations. Firstly, multispectral imagery and elevation models were acquired and processed using ESA’s SNAP software and ESRI’s ArcGIS Pro, to extract shoreline of the area. As mentioned, in the dataset section, 7 Sentinel-2 images from the Copernicus mission were used, from 2017 to 2023, focusing on the summer months to avoid snow and cloud interference. All data were transformed to the WGS 84 UTM Zone 33 N projection to ensure uniformity across the region. Only Level-1C TOA reflectance data from the MSI instrument were selected. The satellite data were processed in SNAP which consisted of geolocation or georeferencing by GeFolki Co-registration Processor GeFolki Co-registration Processor was utilised in bounding box registration to execute alignment. Then, all bands were resampled to a 5 m resolution using bicubic interpolation. Next, a subset to each image was cropped to the area of interest (Fig. 3). The MNDWI Calculation has been used, which is the process of calculating the Modified Normalized Difference Water Index, with the following equation (Equation (1)): MNDWI = (Green –SWIR)/(Green+SWIR)(1) Next, six physical variables regarding CVI were collected. Regarding geomorphology variables, data was obtained from the Norwegian Polar Institute (scale 1:250,000), while suitable for regional studies, the resolution is less detailed. Lithological classifications and boundaries were checked with Sommerfeld et al. (2000) and Beaumont et al. (2018) publications. Soft materials consisting of shale, siltstone, sandstone, and glaciofluvial deposits dominated the lithology, presenting towards higher erosion rates. For the coastal slope, a DEM of 20 m resolution was extracted from the ‘Terrengmodell – Svalbard’ by the Norwegian Polar Institute. After the extraction and reclassification, slope values were calculated to represent erosion potential. The steeper slopes are less susceptible, while gentle slopes show increasing erosion susceptibility. To calculate the coastal slope %, the following equation was used (Equation (2)): Coastal Slope(%) = (Elevation Distance)(2) Coastal forcing is a category implemented in CVI estimation from the aspect of the dominant oceanographic conditions of an area. Three coastal forcing factors are implemented in CVI calculation, and they consist of a) Relative Sea Level Change (RSLC), b) Mean Tide Range (MTR), and c) Mean Wave Height (MWH), according to (Klose, 2003). The data of these coastal forcing parameters were obtained both from literature and Jask´ olski et al. (2018). The numbers which were used for the calculations are for Mean Wave Height 0.8 m, for Relative change of Sea Level: −1.76 mm/year, and for Average tidal range 1.5 m. Lastly, four socio-economic variables were selected for SVI: land use/land cover, road network, settlements, and population density. For the first one, the data were obtained from ESA WorldCover 2021, which offers 11 thematic classes with 10 m resolution. These were reclassified into vulnerability categories according to McLaughlin et al. (2010). Road network and settlements data were extracted from the Geofabrik open-source platform. Roads were categorized by size and type, while settlements were mapped and assigned values indicating size and human presence. Lastly, regarding population density, data derived from Statistics Norway for 2023. The value of 246 inhabitants/km 2 was used for the baseline zone of Longyearbyen. 3.2. Main processing - indices computation The next stage of the study includes the computation of CVI, SVI, and ICVI using two methods, the classical approach as presented by McLaughlin et al. (2010), and the Analytical Hierarchy Process (AHP). In this stage, data were calculated using specific equations and multi-criteria decision analysis to generate indices for coastal vulnerability. Firstly, the classical approach was conducted, with the use of variables, ranked based on semi-quantitative scale by McLaughlin et al. (2010), were values ranging from 1 (Very Low) to 5 (Very High), for different sections of the coastline. Secondly, the Analytical Hierarchy Process (AHP) method, which assigns weights relevant to the importance of these variables. Both methods follow a quartile classification according to V. M. Gornitz et al. (1994), providing insight into the vulnerability of coastal sections to key physical parameters. By applying the quartile classification to both calculations, the coastline was classified into four vulnerability categories: Low, Moderate, High, and Very High. Towards this purpose, every element was given a score based on its exposure to coastal dangers. CVI was calculated using the classical method (Equation (3)): CVI = a+b+c+d+e+f 6 √(3) where: a: Geomorphology, b: Shoreline Change Rate, c: Coastal Slope, d: Relative Sea Level Change, e: Mean Wave Height, and f: Mean Tidal Range. Similarly, SVI was computed using Equation (4): SVI = a+b+c+d 4 √(4) E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 7 where: a: Land Use/Land Cover, b: Road Network, c: Settlements, d: Population Density. The integration of physical and socio-economic vulnerability ICVI was achieved via Equation (5): ICVI =(CVI +SVI 2)(5) This same formulation was applied in both the classical and AHP-based methods (see Table 1). The weights under the CVI and SVI variables were determined using the expert-based Analytical Hierarchy Process (AHP) method. This multi-criteria decision analysis method is constructed by Saaty (2001) who divides each particular issue into composite factors and constructs a weighted sum of the factors for the case at hand using self-defined matrices and value assignments based on its importance (from 1 to 9, Table 2). Firstly, a creation of pairwise matrix, for СV1 and SVI, is conducted, followed by weight calculation, where weights were derived from the eigenvectors of the matrix. Next, a consistency check is essential for calculating the consistency ratio (CR) to ensure matrix reliability. A CR ≤0.1 indicates acceptable consistency. Finally, CVI and SVI variables scores were multiplied by their weight, and indices were recalculated, according to the AHP method. For the geological and physical variables, weighting factors for CVI were calculated by the AHP method. Specifically, for the estimation of weighting factors for CVI with the AHP method, a 6 by 6 comparison matrix is formulated with consistency tested via the Consistency Ratio (CR). Geomorphology variable was assigned the highest weight, due to substantial impact on coastal vulnerability, and relative sea level change (RSLC) with the lowest one (Hzami et al., 2021; Pantusa et al., 2022). For socio-economic variables, weighting factors for SVI were calculated by the AHP method, with a 4 by 4 comparison matrix, where land use/land cover variable was assigned the highest weight, and settlements the lowest one. Weights for each CVI and SVI variables are presented below in Table 2. 4. Results In the present study, the Coastal Vulnerability Index (CVI), and Socioeconomic Vulnerability Indices (SVI) were calculated independently, to evaluate the coastline vulnerability to erosion and the impact of coastal vulnerability erosion on socioeconomic factors, respectively. For both methods, the variables that have been used for CVI are geomorphology, coastal slope, shoreline change rate, mean wave height (MWH), mean tidal range (MTR), and relative sea-level rise (RSLC), which are key physical parameters on coastline erosion. The coastline of 2017, used as the baseline, is divided into 21 sections where all the corresponding values were assigned for variable ranking and CVI, SVI, and ICVI estimation (Fig. 4). Results of the fluctuation of each variable used for the indices calculation as the coastline percentage, are shown in Fig. 5. Both CVI and SVI have a classical method of calculation which relies on preset specific criteria focusing on physical parameters and socio-economic factors, respectively. Unlike these two methods, the AHP approach is more sophisticated because it emphasizes the importance of variables such as geomorphology, coastal slope, land use/land cover, and population density by assigning weight factors to them. This shift allows for enhanced evaluation of vulnerability relating to coastal erosion because the purpose of assigning weights is to stratify the importance of some contours in relation to others. Both approaches identify vulnerable areas along the coastline, including River Delta, Ports, and Main Road. However, the AHP method improves the dynamic evaluation of the coastal vulnerability in the region where socio-economics exacerbate the underlying physical vulnerabilities. Regarding the geomorphology vulnerability fluctuates between two values, Very High (28.57 %), concentrated in the river Delta, Table 1 Detailed dataset along with their data sources. Index Data Type Data Source Link Year CVI Geomorphology Norwegian Polar Institute https://data.npolar.no/dataset/ea550619-2854-45fbbeb9-e453ceeb293c – Coastal Slope S0 Terrengmodell – Svalbard https://data.npolar.no/dataset/dce53a47-c726-4845B5c3-a65b46fe2fea – Shoreline Erosion/ Accretion (m/yr) Sentinel-2 MSI, Level 1C (Copernicus) S2B_MSI1C_20170802T123659 (August 2, 2017) 2017–2023 S2B_MSI1C_20190806T121659 (August 6, 2019) S2B_MSI1C_20200807T120649 (August 7, 2020) S2A_MSI1C_20210706T130711 (July 6, 2021) S2A_MSI1C_20220709T122711 (July 9, 2022) S2B_MSI1C_20230808T122659 (August 8, 2023) Mean Wave Height Based on literature –2017–2023 Relative Sea Level Change (mm/year) Permanent Service for Mean Sea Level (PSMSL), Station: Barentsburg & Literature https://psmsl.org/data/obtaining/stations/541.php 2017–2023 Mean Tide Range Kartverket (Norwegian Mapping Authority) & Literature https://kartverket.no/en/at-sea/sehavniva/result? id=96647&location=Longyearbyen 2017–2023 SVI Lang Use/Land Cover ESA WorldCover 10 m V200 https://worldcover2021.esa.int/download 2021 Population Density Statistics Norway https://www.ssb.no/en/befolkning/folketall/statistikk/ befolkningen-pa-svalbard 2023 Road Network OpenStreetMap - Geofabrik https://download.geofabrik.de/europe/norway.html 2023 Settlements OpenStreetMap - Geofabrik https://download.geofabrik.de/europe/norway.html 2023 E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 8 coastal city area and parts of the main road (Sections 1-6), while Moderate (71.43 %) values, identified mainly in the west and largest part of the coastline (Sections 7-21) (Fig. 6). Coastal slope, on the other hand, varies significantly. Very Low values (33.33 %) with the highest percentage and focusing on Airport area (Sections 16-21) and River Delta (Section 2), while Low values (33 %), are observed along the Ports (Port 1: sections 3-7, Port 2: sections 9-10), and Main Road area (Sections 3-7). Moderate values (14.29 %) prevail in River Delta, Port 3, and Airport area (Sections 1, 11, 15), while High vulnerability values (19.05 %) in Port 2 and Airport area (Sections 8, 12-14) (Fig. 6). For shoreline changes, rates range from very high to very low. Very High vulnerability is observed in Section 1 of River Delta, aligning with geomorphological dynamics, while Very Low and Moderate values, are comprising 85.71 % of coastline. Regarding coastal forcing factors, they were pre-assigned during pre-processing. Relative Sea Level Change (RSLC) presents a Very Low vulnerability, and Mean Wave Height (MWH) a Low vulnerability. On the contrary, Mean Tide Range (MTR), has a major impact on vulnerability, as it presents High values that are recorded. The results of the vulnerability were calculated using Equation (3) and classified accordingly to quartile method (Gornitz et al, Table 2 Weighting factors for CVI and SVI Variables as calculated by AHP method. Variables Weighting factor CVI Geomorphology 0.4191 Coastal Slope 0.2867 Shoreline Rate 0.1590 MWH 0.0615 MTR 0.0423 RSLC 0.0314 SVI Landuse/Landcover 0.5358 Population Density 0.3179 Road Network 0.1047 Settlements 0.0416 Fig. 4. Coastline of 2017 divided into 21 sections to serve as the baseline for the proper assigning of variable values and CVI, SVI, and ICVI estimation. E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 9 Loinenak, F.A., Hartoko, A., Muskananfola, M.R., 2015. Mapping of coastal vulnerability using the Coastal Vulnerability Index and Geographic Information System. IJtech 6 (5). https://doi.org/10.14716/ijtech.v6i5.1361. L´ opez Royo, M., Ranasinghe, R., Jim´ enez, J.A., 2016. A Rapid, Low-Cost Approach to Coastal Vulnerability Assessment at a National Level. J. Coast Res. 32 (4), 932–945. https://doi.org/10.2112/JCOASTRES-D-14-00217.1. Major, D.C., Lehmann, M., Fitton, J., 2018. Linking the management of climate change adaptation in small coastal towns and cities to the Sustainable Development Goals. Ocean Coast Manag. 163, 205–208. https://doi.org/10.1016/j.ocecoaman.2018.06.010. Mahapatra, M., Ramakrishnan, R., Rajawat, A.S., 2015. Coastal vulnerability assessment using analytical hierarchical process for South Gujarat coast, India. Nat. Hazards 76 (1), 139–159. https://doi.org/10.1007/s11069-014-1491-y. Mani Murali, R., Ankita, M., Amrita, S., Vethamony, P., 2013. Coastal vulnerability assessment of Puducherry coast, India, using the analytical hierarchical process. Nat. Hazards Earth Syst. Sci. 13 (12), 3291–3311. https://doi.org/10.5194/nhess-13-3291-2013. McLaughlin, S., Andrew, J., Cooper, G., 2010. A multi-scale coastal vulnerability index: A tool for coastal managers? Environ. Hazards 9 (3), 233–248. https://doi. org/10.3763/ehaz.2010.0052. McLaughlin, S., McKenna, J., Cooper, J.A.G., 2002. Socio-economic data in coastal vulnerability indices: Constraints and opportunities. J. Coast Res. 36, 487–497. https://doi.org/10.2112/1551-5036-36.sp1.487. Nath, A., Koley, B., Choudhury, T., 2024. Coastal vulnerability differentiated according to geospatial quartiles method: Rasulpur to Subarnarekha estuary, east coast of India. Spatial Inform. Res. 32, 383–398. https://doi.org/10.1007/s41324-023-00554-w. Nath, A., Koley, B., Choudhury, T., others, 2025. Prioritizing flood drivers: an AHP-based study of physical factors in Digha’s coastal belt, East Coast, India. Spatial Inform. Res. 33, 12. https://doi.org/10.1007/s41324-025-00615-2. Nath, A., Koley, B., Saraswati, S., 2023. Geospatial analysis of short term shoreline change behavior between Subarnarekha and Rasulpur estuary, east coast of India using intelligent techniques (DSAS). Geojournal 88 (Suppl. 1), 255–275. https://doi.org/10.1007/s10708-022-10683-8. Nath, A., Koley, B., Saraswati, S., 2021. Identification of the coastal hazard zone between the areas of Rasulpur and Subarnarekha estuary, east coast of India using multi-criteria evaluation method. Model. Earth Syst. Environ. 7, 2251–2265. https://doi.org/10.1007/s40808-020-00986-5. Nguyen, C.H., Dang, K.B., Ngo, V.L., Dang, V.B., Truong, Q.H., Nguyen, D.H., Giang, T.L., Pham, T.P.N., Ngo, C.C., Hoang, T.T., Dang, T.N., 2021. New approach to assess multi-scale coastal landscape vulnerability to erosion in tropical storms in Vietnam. Sustainability 13 (2), 1004. https://doi.org/10.3390/su13021004. Nitze, I., Grosse, G., Jones, B.M., Romanovsky, V.E., Boike, J., 2024. Pan-Arctic assessment of coastal settlements and infrastructure exposure to permafrost degradation. Earths Future. https://doi.org/10.1029/2024EF005013. Obura, D.O., 2021. Integrated Assessment of Coastal Exposure and Social Vulnerability to Coastal Hazards in East Africa. Estuaries Coasts 44, 1234–1249. https://doi. org/10.1007/s12237-021-00930-5. O’Higgins, T., O’Dwyer, B., 2019. A vulnerability framework to protect coastal social ecological systems. Anthropocene Coasts 2 (1), 279–292. https://doi.org/ 10.1139/anc-2018-0023. ¨ Ozyurt, G., Ergin, A., 2010. Improving coastal vulnerability assessments to sea-level rise: A new indicator-based methodology for decision makers. J. Coast Res. 26 (2), 265–273. https://doi.org/10.2112/08-1055, 1. Pantusa, D., D’Alessandro, F., Riefolo, L., Principato, F., Tomasicchio, G.R., 2018. Application of a coastal vulnerability index. A case study along the Apulian Coastline, Italy. Water (Switzerland) 10 (9). https://doi.org/10.3390/w10091218. Pantusa, D., D’Alessandro, F., Frega, F., Francone, A., Tomasicchio, G.R., 2022. Improvement of a coastal vulnerability index and its application along the Calabria Coastline, Italy. Sci. Rep. 12 (1). https://doi.org/10.1038/s41598-022-26374-w. Pendleton, E.A., Theiler, E.R., Williams, S.J., 2004. Coastal vulnerability assessment of Cape Hatteras National Seashore (CAHA) to sea-level rise. U.S. Geological Survey. Open-File Report 2004-1064. https://pubs.usgs.gov/of/2004/1064/. Powell, E.J., Tyrrell, M.C., Milliken, A., Tirpak, J.M., Staudinger, M.D., 2019. A review of coastal management approaches to support the integration of ecological and human community planning for climate change. J. Coast Conserv. 23 (1), 1–18. https://doi.org/10.1007/s11852-018-0632-y. Rehman, S., Jahangir, S., Azhoni, A., 2022. GIS based coastal vulnerability assessment and adaptation barriers to coastal regulations in Dakshina Kannada district, India. Reg. Stud. Marine Sci. 55, 102509. https://doi.org/10.1016/j.rsma.2022.102509. Rocha, C., Antunes, C., Catita, C., 2020. Coastal Vulnerability Assessment Due to Sea Level Rise: The Case Study of the Atlantic Coast of Mainland Portugal. Water 12 (2), 360. https://doi.org/10.3390/w12020360. Roukounis, C.N., Tsihrintzis, V.A., 2022. Indices of coastal vulnerability to climate change: a review. Environ. Process. 9 (2), 29. https://doi.org/10.1007/s40710022-00577-9. Saaty, T.L., 2001. Fundamentals of the Analytic Hierarchy Process. In: J, A, M.G., M, P., Schmoldt Daniel, L., Kangas (Eds.), The Analytic Hierarchy Process in Natural Resource and Environmental Decision Making. Springer, Netherlands, pp. 15–35. https://doi.org/10.1007/978-94-015-9799-9_2. Sallaye, M., Khoudir, M., Abdelalim, D., Cherif, Y.S., 2022. Coastal vulnerability assessment and identification of adaptation measures to climate change between Cape Matifou and Cape Djinet. Algeria. Geo-Eco-Marina 28, 181–193. https://doi.org/10.5281/zenodo.7493268. Sokolickova, Z., Meyer, A., Vlakhov, A.V., 2022. Changing Svalbard: Tracing interrelated socio-economic and environmental change in remote Arctic settlements. Polar Rec. 58, e23. https://doi.org/10.1017/S0032247422000213. Theocharidis, C., Prodromou, M., Doukanari, M., Kalogirou, E., Eliades, M., Kontoes, C., Hadjimitsis, D., Neocleous, K., 2025. Integrated Coastal Vulnerability Index (ICVI) Assessment of Protaras Coast in Cyprus: Balancing Tourism and Coastal Risks. Geographies 5 (1), 12. https://doi.org/10.3390/geographies5010012. Toumasi, P., Petropoulos, G.P., Detsikas, S.E., Kalogeropoulos, K., Tselos, N.G., 2024. Coastal Vulnerability Impact Assessment under Climate Change in the Arctic Coasts of Tromsø, Norway. Earth 5 (4), 640–653. https://doi.org/10.3390/earth5040033. Tsai, Y.L.S., 2024. Monitoring Arctic permafrost coastal erosion dynamics using a multidecadal cross-mission SAR dataset along an Alaskan Beaufort Sea coastline. Sci. Total. https://doi.org/10.1016/j.scitotenv.2024.170389. Wang, X., Liu, B., He, S., Yuan, H., Ji, D., Qi, L., Song, Y., Xu, W., 2024. An Analytic Hierarchy Process method to evaluate environmental impacts of in situ oil shale mining. Sustainability 16 (4), 1363. https://doi.org/10.3390/su16041363. Widura, E., Mardiatno, D., 2022. Assessment of the Coastal Vulnerability Index (CVI) for disaster mitigation strategies in some coastal tourism areas in Gunungkidul, Yogyakarta-Indonesia. IOP Conf. Ser. Earth Environ. Sci. 1315/989/1/012014 989 (1). https://doi.org/10.1088/1755-1315/989/1/012014. Winther, S.N., Gudmestad, O.T., 2023. Impact of and solutions to effects of climate changes for Longyearbyen, Svalbard, Norway. IOP Conf. Ser. Mater. Sci. Eng. 1294 (1), 012036. https://doi.org/10.1088/1757-899X/1294/1/012036. Zag´ orski, P., Jarosz, K., Superson, J., 2020. Integrated Assessment of Shoreline Change along the Calypsostranda (Svalbard) from Remote Sensing, Field Survey and GIS. Mar. Geod. https://doi.org/10.1080/01490419.2020.1715516. E. Tita et al. Remote Sensing Applications: Society and Environment 39 (2025) 101642 16