Pathways to social–ecological indicators in the era of Digital Twins of the Earth

Effective ecosystem-based management strategies that prevent further degradation of the ecosystem state and negative impacts on societies are needed but require cross-disciplinary and complex understandings of social–ecological systems (SES). This study defines, evaluates, and operationalises the concept of social–ecological (SE) indicators using the butterfly model based on the DAPSI(W)R(M) framework, linking drivers, activities, pressures, state changes, impacts on human welfare, and responses as measures. A review of 148 indicators developed under the marine strategy framework directive (MSFD), including contributions from HELCOM, OSPAR, the European Environment Agency, and marine-relevant UN sustainable development goals (SDGs), was conducted. Indicators were assessed across four key dimensions: mapping along the DAPSI(W)R(M) cause-effect chain; indicator complexity; normative basis; and the relationship between the indicator and its indicandum. Based on the analysis and a literature study, an SE indicator definition was developed requiring SE indicators to be compound or explicitly connect the ecological and social dimensions. Only nine of the analysed indicators qualify as SE indicators by making traceable connexions across both dimensions of the butterfly model. Existing compound indices, such as the ocean health index (OHI) and its regional adaptation, the Baltic health index (BHI), illustrate how nested, goal-oriented indices can operationalise SE integration while also highlighting challenges related to transparency and aggregation due to their nested nature. Pathways for adapting existing non-SE indicators into SE indicators are outlined, drawing on ongoing work in the SEADITO project (2024–2027) that develops SE modelling tools for the European Digital Twin of the Ocean (EU DTO). To enable compound SE indicators and support implementation within emerging digital twins of the earth, such as the EU DTO, additional criteria for SE indicators are proposed, consisting of the addition of crosscutting interlinkages and digital readiness, respectively, in addition to existing criteria of measurability, sensitivity, specificity, scalability, transferability, and precision. By clarifying what makes an indicator SE and identifying concrete development pathways, this study advances the methodological foundation for integrating SES analysis into marine spatial planning and ecosystem-based management in an era in which digital twins of the earth are accelerating the digital infrastructure for SES.

Most ecosystems and biodiversity are globally under immense pressure, escalated by climate change and unsustainable human activities (Zhang et al., 2025; Nimma et al., 2025). Effective ecosystem-based management that prevents further degradation of the ecosystem state and negative impacts on society is needed but it requires a robust understanding of the interactions between people and nature, as well as of their feedback loops (O'Higgins et al., 2020; Vugteveen et al., 2015). Analyses of these complex interactions are challenging (Flood et al., 2020), but the notion of social–ecological systems (SES) thinking has improved understanding of them by providing a worldview which focuses on relational qualities and dynamic processes between actors and their environment, rather than just on the physical design of places (Preiser et al., 2022; Healey, 2007; Potschin-Young et al., 2018; Schlüter et al., 2019).

An SES approach, at its core, is based on the understanding that society is integrally and dynamically linked to the natural environment (Schlüter et al., 2019; Biggs et al., 2022). It enables the breaking down of the complex interactions into constituent parts for analytical purposes and defining clear boundaries to external forces (Healey, 2007; Ostrom, 2009). Requirements to implement a SES approach involve fit-for-purpose conceptual frameworks, indicators, and tools that capture the ecological, social, and economic dimensions of a changing society and environment (Ban et al., 2025; McKane et al., 2020), and which draw on robust data, geographic information systems, and HPC (high-performance computing) environments (Bryan, 2013).

The DAPSI(W)R(M) model— Drivers-Activities-Pressures-State-Impacts-on-human-Welfare-Responses-as-Measures —the butterfly model in short—is a SE cause-and-effect framework developed to inform decision-making by linking interacting social and ecological components, visualised as the two opposing wings of a butterfly (Elliott and O'Higgins, 2020; Patrício et al., 2016). One wing of the butterfly model makes up the ecological part of the system, while the other depicts elements mostly linked to the social dimension of a SES. In DAPSI(W)R(M), the only environmental elements are the “state,” consisting of ecosystem components and functions, and pressures affecting the ecosystem state. Its element “Impacts on human Welfare” derived from the societal use of ecosystem services and benefits can be viewed as an equivalent to the ecosystem service cascade framework (Haines-Young and Potschin, 2010), a view built into the butterfly model (Elliott and O'Higgins, 2020). Pressures on the ecosystem State, or structures, may result from human activities or environmental processes and may co-occur (Reckermann et al., 2022). Responses in measures can be triggered to control unwanted impacts on the social side of the butterfly, carry out management actions to improve the environmental state, and promote a more sustainable supply of services (Elliott and O'Higgins, 2020). In this way, societal responses can influence three types of impacts: (a) ecological impacts on ecosystem capacity, (b) social–ecological impacts on ecosystem services, and (c) social impacts on (dis)benefits and their value and distribution (Frederiksen et al., 2021). The former represents impacts on the ecosystem's State, while the latter two represent Impacts on Welfare, drawing attention to the impacts on distributional social justice (Frederiksen et al., 2021).

Ecosystem service supply or flow and benefits are only realised if they are demanded, consumed, and valued by people, which greatly depends on social actors, processes, and human drivers (Haines-Young and Potschin, 2010). Ecosystem services analysis is an SES approach aimed at measuring the final outputs from an ecosystem that people depend on for their wellbeing and livelihoods, and it serves as a way of interpreting our relationship with ecosystems in terms of biophysical structures, processes, and functions that underpin ecosystem service supply (Culhane et al., 2020; Potschin-Young et al., 2018). The central elements of the cyclical ecosystem services cascade framework are: (i) ecosystem components that make up the ecosystem, (ii) ecosystem functions performed by it, (iii) ecosystem services it supplies, (iv) the benefits it provides to people—often measured as changes in human wellbeing resulting from service consumption or experience, (v) the socially constructed value, representing the worth or importance of the change in wellbeing, and (vi) the pressures on the ecosystem, caused by various human activities (Potschin-Young et al., 2018). Ecosystem services are often categorised into groups such as provisioning (e.g., fish for food), cultural (e.g., water environments for recreation), and regulating and maintenance (e.g., nutrient regulation). Some of these groups have a much more direct and easily measurable link to benefits, with a clear distinction between ecosystem functions—such as the connexion between provisioning services like wild fish for food and increased economic welfare (Hyytiäinen et al., 2014; Armoskaite et al., 2020).

Measurability is essential for applying SES frameworks in practise, requiring the use of indicators that—when well-selected—are variables useful for assessing change (Murillas-Maza et al., 2020). The term “indicator” is widely used in science and policy, with its purpose varying across contexts, ranging from strictly biophysical indicators to performance indicators developed to track trends across economic, environmental, and social domains for policy evaluation (European Commission (EC), 2021; Asmala et al., 2024; Zhao et al., 2020; Weyrauch, 2012). Bousquin and Mazzotta (2020) stress that a desirable indicator (how the subject is studied) must have a strong relationship to the indicandum (the subject studied) to properly inform decision-making. However, it must also be simple enough to be modelled and measured with available data. To measure sustainable development, reduce the number of indicators, and reflect the integrative characteristics of systems, single indicators are sometimes combined into indices, which are compound bundles of combined indicators, naming the human development index by the United Nations Development Program (UNDP) as an example (Wu and Wu, 2012). Hattam et al. (2015) describe indicators and indices as proxies for complex phenomena in the context of ecosystem service analysis but note a tendency for ecosystem service indicators and indices to be mostly biophysical, e.g., focus on species or ecological, structural, and functional aspects related to provisioning and regulating services. Social variables and data have been given historically less attention by marine spatial planning modellers (Galparsoro et al., 2021; Leenhardt et al., 2015). Yet, it is increasingly acknowledged that social data and indicators can offer significant insights for successful environmental management and MSP. For example, indicators and indices in policymaking include instruments for measuring inputs, processes, products, results, and outcomes of interventions to monitor and evaluate the influence of policy (Weyrauch, 2012), and performance indicators and indices are often applied to track trends in economic, environmental, governance, and social domains such as the reduction in greenhouse gas (GHG) emissions in ports compared to a baseline (Asmala et al., 2024).

While single indicators are important for describing the different components of the butterfly individually, compound social–ecological (SE) indices—and indicators on the SE interactions—are much needed for sustainable development indicators and indices to “be more than environmental” in nature (Wu and Wu, 2012), and to strengthen SES understandings (Estoque and Murayama, 2017; Teixeira et al., 2015). This gap—the missing links between the environmental and the social side of a SES—is particularly noted in recent development of marine spatial planning, a legal planning framework introduced in recent decades, to broaden spatial planning from terrestrial areas to coasts and seas (Galparsoro et al., 2021; Leenhardt et al., 2015). Therefore, the objective of this paper is to define criteria for SE indicators, utilise the butterfly model to analyse and expand the scope of existing, well-known indicators and indices to reflect SE linkages, and present criteria for developing tools to analyse SE change in support of marine spatial planning and ecosystem management. The need for better representations of SE linkages is escalated by the recent developments of digital twins of the earth around the globe, which are digital systems or Earth replicas for modelling advanced earth and societal processes (Nativi et al., 2021), since they support SE approaches with digital environments to a degree not seen before. For example, the European Union Digital Twin Ocean (EU-DTO) was launched in 2022 to make advanced, real-time, and widely shared Findable-Accessible-Interoperable-Reusable (FAIR) data, models, and high-computing environments available to support action towards healthier waters (Tzachor et al., 2023). The central research questions this paper explores are: what makes an indicator or index SE? Do proper SE indicators already exist? How can non-SE indicators or indices be adapted to become SE, enabling their operationalisation in digital twins such as the EU DTO?

The second section presents the methods for data collection and analysis, which consist of the combination of a literature review of existing definitions of indicators in the context of ecology and environmental and marine spatial planning, and an analysis of a pool of existing potential SE indicators utilising the butterfly model and four key elements of the indicator concept. The results in the third section present the indicator analysis divided into the four key indicator elements, which involve (a) their application along the cause-and-effect chain, (b) complexity of the indicator, (c) normative basis, and (d) the relationship between the indicator and the indicandum, before a SE indicator definition is deduced and applied to the selected indicator pool. Pathways for developing SE indicators are discussed in the fourth section, with the Digital Twin of the Earth developments, such as the EU DTO, providing a digital setup to escalate the integration of SE indicators into decision-making when developed, by combining data spaces that have tended to be separated across sectors.

While the scope of highlighting a definition and road forward for indicator and index development to better integrate the two wings of the butterfly model is globally useful, the work presented here is part of the EU Horizon-funded Social–Ecological Analysis and Models for Digital Twin Oceans [2024–2027; Social-Ecological Analysis and Models for Digital Twin Ocean (Social-Ecological Analysis and Models for Digital Twin Ocean (SEADITO), 2025) project. Regional data spaces around the world—when combined into cross-sectoral data environments, as many data spaces tend to be limited to certain sectors—provide information that can be utilised to populate SE indicators (Miguez et al., 2019). For example, an important data space in Europe consists of the European Marine Observation and Data Network (EMODnet), providing sub-portals for data on bathymetry, geology, physics, chemistry, biology, seabed habitats, and human activities (Miguez et al., 2019), in the EU DTO integrated with other data spaces, e.g., Copernicus Marine Service (CMEMS). The Digital Twin of the Earth developments, such as the EU DTO, are still ongoing, and they require cross-sectoral what-if scenarios to better plan the future which highly depend on SES tools, services, indicators, indices, and data (Tzachor et al., 2023).

To answer our research questions on defining and exploring SE abilities of existing indicators, we develop an SE definition by combining a literature study of this topic with the application of the butterfly model. We use our SE definition to analyse indicators developed by regional sea conventions and EU organisations for the EU Marine Strategy Framework Directive [MSFD; European Commission (EC), 2017a] from the Baltic Marine Environment Protection Commission (HELCOM), the Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR), the European Environment Agency (EEA), as well as marine-context related United Nations Sustainable Development Goal (SDG) indicators. A pool of 148 indicators from OSPAR, HELCOM, the EEA, and the UN SGDs was compiled (Table 1; Supplementary Table S1) and analysed according to the butterfly model and SE definition, determining which of the selected existing widely applied indicators qualify as SE, and what it takes to turn existing indicators into SE. While we only analyse this specific set of indicators, we broaden the scope of indicators and indices analysed as part of the literature review, including the introduction of one existing compound index consisting of the Ocean or Baltic Health Index (OHI/BHI) (Halpern et al., 2012; cited in Blenckner et al., 2021), unpacked in the discussion as a compound index under development within the ongoing SEADITO project (2024–2027).

The butterfly model conceptually contains fewer environmental than social elements (Elliott and O'Higgins, 2020). For practical reasons, the environmental elements were chosen as the starting point of the analysis of existing indicators. The most recent criteria and methodological standards for the MSFD (European Commission (EC), 2017a) were used to structure the indicators as it is a common framework shared across the European seas. The MSFD descriptors and criteria encompass a wide range of themes that include all aspects of marine ecosystems and their interactions with society. HELCOM, OSPAR, and EEA indicators align with the MSFD requirements. Consequently, the MSFD descriptors and criteria were used to organise and synthesise the findings of HELCOM, OSPAR, and EEA indicators. The SDG goals and indicators were reviewed and connected to the MSFD descriptors and criteria to propose potential routes for developing non-SE indicators into SE indicators.

The development of the SE definition and the overall methodology followed an iterative, collaborative process involving three in-person SEADITO project meetings and bi-monthly online discussions held between September 2024 and 2025. The meetings brought together the authors and the wider SEADITO project team with experts in marine social, environmental and geospatial information science. Early discussions focused on establishing a shared understanding of SES analysis and the butterfly model. Later stages were focused on developing examples of SE indicators and moving towards developing and operationalising the concept of SE indicators. One concept validation workshop was held on 23 April 2025, with 16 SEADITO project partners to review a sample of indicators from the pool (see Supplementary Data sheet 1 for results and Table 2 for the template used). The participants were divided into groups of four and assigned five random indicators per group to review for SE characteristics. They were asked to discuss and record their answers to the following questions: (a) Is the indicator SE?

(b) Which element of the DAPSI(W)R(M) butterfly framework does it represent?

(c) What does it take for the indicator to be made SE, and how can this be done?

The comments informed a comprehensive analysis of the entire selected pool of indicators, based on the questions tested at the workshop. Source documents (Table 1) were analysed to identify variables measured by the indicator, and the potential links to other parts of the butterfly components. For example, which activities or impacts on welfare can be linked to the pressure and state indicators? The HELCOM, OSPAR, and EEA indicator protocols often provided this information.

Reviewing existing definitions of indicators in the context of ecology and environmental planning that are potential SE indicators highlighted four key elements of the indicator concept and its application across social and ecological domains that we utilise to structure our findings, following Heink and Kowarik (2010): a. DAPSI(W)R(M) mapping: application along the chain of cause-and-effect, i.e., the elements in the butterfly model, indicates whether an indicator or index is social, ecological, or both, and to what stages in dynamic causal chain it belongs.

b. Complexity of the indicator or index: complexity spans a scale from simpler measurements of just one parameter to a complex index which consists of complex, composite, or compound indicators.

c. Normative basis: the normative basis describes the tendency for indicators and indices to be either descriptive and illustrative, normative and prescriptive, potentially evaluative, or hybrid.

d. Relationship between the indicator and the indicandum: while the empirical correlation between the indicator and the indicandum is a defining criterion for descriptive indicators, it is not a requirement for evaluative and prescriptive indicators, although complex indicanda (e.g., sustainability or biodiversity) require integrating information over time and space.

The selected list of analysed indicators and the compound index OHI found in the literature review were assessed according to these four domains before the SE criteria definition was deduced.

The following section presents the analysis of existing indicator examples assessed through the four key elements of the indicator concept outlined above and uses the analysis to distil the core characteristics of an SE indicator.

DAPSI(W)R(M) mapping of indicators along the chain of cause-and-effect

Indicators can be applied at any stage of the causal chain, from human activities to changes in environmental state and societal responses (Vugteveen et al., 2015). Standalone social, economic, and environmental indicators play important roles in measuring separate parts of the butterfly model. Drivers are the basic physical and psychological needs of people, such as shelter, sustenance, and clean air (Elliott and O'Higgins, 2020). A driver, or driving force indicator measures social, demographic, and economic changes in societies, such as consumption, production, and lifestyles (Heink and Kowarik, 2010; Figure 1).

Activities refer to the ways in which people interact with the natural environment to meet their needs, such as developing infrastructure for energy production or fishing. Geospatial and temporal data on activity footprints—indicating where, when, and which activities take place—are used to measure the extent of human interaction with the natural and built environments. The EU MSFD (European Commission (EC), 2017b) lists 34 different activities (e.g., coastal defence and protection, urban uses, various types of fishing, and forestry) and 16 pressures to be addressed in marine strategies across EU member states.

Pressures can arise from environmental processes, such as changes in salinity or temperature, as well as from human activities. One example of a pressure indicator could be the introduction of non-indigenous/invasive species within a specific timeframe, which can be facilitated by climate change and measured through environmental and biodiversity indicators, as well as ballast tank wastewater discharge in ports, with underlying indicators closely linked to geospatial and temporal data on human activities.

Ecosystem state indicators are environmental and biodiversity indicators. The cumulative pressures change the ecosystem state, represented by state and effects on the state indicators. State indicators measure both biotic and abiotic factors, as well as the overall condition of the ecosystem and its functions (Murillas-Maza et al., 2020; von Thenen et al., 2020). Ecosystem state indicators include species abundance or distribution. In contrast, ecosystem functioning is assessed based on the processes that species or species assemblages perform, such as filtration of suspended matter or providing refuge and shelter for other species.

Changes in the ecosystem state can sometimes cause impacts on human welfare. Impact on human welfare indicators focus on the cumulative impacts of human-induced stresses cross certain politically and/or scientifically defined thresholds, relying on the sensitivity of natural resources. They include changes in the societal use of ecosystem services and their associated benefits, as well as changes in their value. Related indicators measure provisioning, cultural, regulating and maintenance ecosystem services and (dis)benefit supply and distribution. Benefit indicators measure the benefits and values of ecological changes on different groups within society and measure changes in human well-being in both economic and non-economic terms (Bousquin and Mazzotta, 2020). The System of National Accounts (SNA; economic market indicators) and the System of Economic Environmental Accounts Ecosystem Accounting (SEEA EA; non-market) are examples of economic indicators used to assess benefits and value. Valuation methods, such as willingness-to-pay (e.g., Pakalniete et al., 2021), provide metrics for measuring the value of non-market benefits. Indicators on responses through measures provide insights into how society addresses environmental concerns (HELCOM, 2025). Such measures can relate to, for example, mitigating negative effects or conserving natural resources, such as the number of protected areas in a country (Heink and Kowarik, 2010), or to tracking policy outcomes, such as the effectiveness of measures to combat eutrophication through river basin management plans.

DAPSI(W)R(M) mapping of selected indicators, grouped by MSFD descriptor

Of the 148 reviewed indicators, 87 were identified as state (59%), 20 as pressure (14%), 17 as response (11%; SDG and EEA indicators only), and only two as Impact (1%; Figure 2; see Supplementary Data sheet 1 and 2).

D8 (contaminants; 51 indicators), D1 (marine biodiversity; 47), and D4 (food webs; 41) had the most indicators, which may also serve as indicators for other descriptors. The vast majority of these were State indicators that describe species abundance, distribution, and productivity for D1 and D4, as well as contaminant concentrations and affected species for D8. This is followed by D5 (eutrophication; 29), most of which were also state indicators but focused on nutrient concentrations or biomass and community changes in phytoplankton and zooplankton. Several pressure–state indicators were identified for D5 (eutrophication), linking nutrient inputs to changes within the water column. Although fewer in number, D9 (contaminants in seafood; 26) were, like D8, primarily State indicators.

D3 (commercial fish and shellfish; 13), D10 (marine litter; 15), D11 (energy, including underwater noise; 10), D2 (non-indigenous species; 9), and D7 (hydrographical conditions; 5) had the fewest unique indicators per descriptor. D3 (commercial fish and shellfish) depicts healthy populations of all commercially exploited fish and shellfish stocks, and that commercial activities operate at a sustainable level. Among all descriptors, D3 had the widest range and a balanced number of indicators addressing the social and environmental dimensions of the butterfly model. D10 (marine litter) indicators can largely be allocated to the pressure category measuring the abundance and composition of litter. D11 (energy, including underwater noise) requires monitoring and action to control underwater noise and other energy inputs to prevent harm to marine species. Pressure indicators for the D11 monitor and estimate the distribution of impulsive sound, assess impact risk, and evaluate continuous/ambient noise. The distribution of pressure is often explicitly linked to an activity and its location, for instance, in impulsive sound locations associated with underwater explosions, active sonars, and pile driving, as well as continuous noise sources such as shipping, fishing, boating, and offshore renewable energy installations. The goal of D2 (non-indigenous species) is that non-indigenous/invasive species have no impact on the local ecosystem and are not introduced or spread in the future. Most D2 indicators are state, pressure, and response focused, monitoring the numbers of newly introduced non-indigenous species, their spatial distribution and abundance, and the proportion of non-indigenous to indigenous species (Figure 2). A single SDG indicator (15.8) is a response indicator for monitoring the proportion of countries that adopt legislation and allocate resources to prevent and control invasive alien species. D7 (hydrographical conditions) focuses on the physical and chemical characteristics of marine waters (i.e., the spatial extent and distribution of permanent alteration of hydrographical conditions) and the spatial extent of the habitats and species affected.

The SE integrative abilities of DAPSI(W)R(M)

To meet the information needs of SES analysis, multiple indicators must be selected and developed—potentially combined into compound indices—to represent all elements of the causal chain, linking human activities to their ultimate environmental impacts and societal responses (Vugteveen et al., 2015; Murillas-Maza et al., 2020). As an example of an applied SES framework inspired by the butterfly model, OSPAR has developed over 50 different indicators monitored and updated every 2, 3, or 6 years, useful in combination to describe changes in ecosystem state due to pressures such as underwater noise, physical disturbance, and loss created by human activities such as offshore wind farm development. They are further linked to impacts on ecosystem services (e.g., loss or gain in provisioning or regulating services), and the framework differentiates between policy, preventive, service, and benefit responses (OSPAR, 2023a). The impacts on ecosystem services include considerations by OSPAR for adapting natural capital accounting into the DAPSI(W)R(M) framework. From 2025, OSPAR will account for ecosystem services, natural capital, and human activities affecting the implementation of ecosystem-based management in accordance with the North-East Atlantic Environment Strategy 2030 (NEAES, 2025).

Exemplifying how the application of multiple indicators along the cascade supports ES stock-taking and scenario analysis assessments within maritime spatial planning (MSP), an immense indicator pool based on the ecosystem service cascade steps was collected by von Thenen et al. (2020). They analysed eleven scientific paper reviews of indicators linked to marine and coastal ecosystem services, including papers providing comprehensive reviews of indicators from the Millennium Ecosystem Assessment (MA) in 2005, the Economics of Ecosystems and Biodiversity (TEEB), the Common International Classification of Ecosystem Services (CICES) V3, and the European Commission (EC) Working Group on Mapping and Assessment of Ecosystems and their Services (WG MAES) pilot classification based on CICES V4.3. The developed pool of indicators, grouped according to themes and classes, shows that among all cascade steps, “value” and specifically economic value are most represented in the analysed marine and coastal ecosystem service studies. In addition to providing an overview of available indicators for each step of the ecosystem service cascade, von Thenen et al. (2020) propose operationalising indicators through the development of narratives to demonstrate how the underlying indicators link different parts of the cascade—ecosystem capacity, services, benefits, and values.

In a similar way, Murillas-Maza et al. (2020) applied a narrative approach to support the application of indicators in storylines or use cases linked to the effectiveness of nature-based solutions and nature-inclusive harvesting in protecting future biodiversity and ecosystem services under climate change across the European seas. A pool of 201 social, economic, and environmental indicators from scientific literature and EU and international initiatives were employed in a workshop by scientists to select relevant indicators for their storylines. These storylines focused on restoring coastal habitats and species, the effectiveness of marine protected areas, and the sustainable use of marine and coastal provisioning services. Among the indicators used in the storylines, 85% were environmental (state and pressure), 71% were economic, and 31% were social.

Indicator complexity

Indicator complexity is a scale that span from simpler measurements of just one parameter being an indicator to a complex index which consists of complex, composite, or compound indicators. The difference between a simple or one-dimensional indicator (e.g., a measurement of a single parameter such as phosphorus concentration) and a complex or composite index, is that the latter may be an indirect measurement and incorporate several types of information to represent the indicandum, although simple indicators can also be multipurpose and nested (Heink and Kowarik, 2010). Heink and Kowarik (2010) give the example of invertebrates as an indicator of soil quality, which in turn serve as an indicator for sustainable land management, making the indicator multipurpose and nested.

While the indicators analysed are relatively simple, an example of an existing compound, multidimensional index is the OHI (Halpern et al., 2012, cited in Blenckner et al., 2021). The OHI is designed to quantitatively and comprehensively assess ocean health from a social–ecological perspective, as progress towards a set of key societal “goals” that reflect the benefits and services people expect from healthy oceans. The definition of these goals and their reference points is tailored to best support critical management and policy objectives.

The Baltic health index (BHI) was adapted from the OHI approach to suit environmental management in the Baltic Sea. Each goal score is calculated based on four dimensions, which include ecological and social data (Blenckner et al., 2021, p. 362): (1) Present status (x) is a goal's current value compared to its reference point, the management target.

(2) Trend (T) is the average percentage change in a goal's status over the most recent 5 years.

(3) Pressures (P) are represented by the weighted sum (based on the relative effects of different pressures on the given goal) of the ecological and social pressures that negatively affect the status of the goal. The ecological and social overall pressures are equally weighted.

(4) Resilience (r) includes (i) ecological integrity (i.e., the status of the biodiversity goal), (ii) goal-specific regulations aimed at addressing ecological pressures, and (iii) social integrity that increases status by reducing or eliminating pressures.

The overall OHI and BHI scores are the average of the present (x) status and the likely future status calculated as the current status modified by the recent trend (T), cumulative pressures, and resilience (r) associated with the goal. The maximum score for each goal and the entire BHI is 100, which indicates that reference points are met rather than pristine conditions. In Blenckner et al. (2021), the BHI score was 76 out of 100, indicating that the health of the Baltic Sea is suboptimal and highlighting the need for stronger efforts to meet management objectives and associated targets. OHI and BHI are highly nested indices that incorporate several simple and complex indicators. Within these indices, social integrity is measured by political instability and corruption, as reported by the World Governance Indicator (WGI). The WGI evaluates six dimensions of governance (voice and accountability, political stability and absence of violence/terrorism, government effectiveness, regulatory quality, rule of law, and control of corruption) based on over 30 data sources, providing subjective perceptions of relevant governance dimensions to generate six aggregate indicators (World Bank, 2025).

Normative basis and the relationship between the indicator and the indicandum

Definitions of indicators encompass descriptive, normative, and hybrid concepts, and they can relate to the indicandum in different ways. Descriptive indicators aim to represent the indicandum (e.g., species abundance) objectively and are assessed based on an empirically measurable, direct link to what they represent (Heink and Kowarik, 2010). Normative indicators, on the other hand, are used to evaluate conditions (e.g., whether consumption is sustainable) or to establish goals (e.g., whether biodiversity targets have been achieved). Normative indicators extend beyond simple description and involve value judgments, goal setting, interpretation, and evaluation, often having a more indirect relationship to the indicandum (Heink and Kowarik, 2010).

Links between the UN SDG indicators and the EU MSFD descriptors and criteria

SDGs are per nature normative, and normative indicators are needed to evaluate societal goals and impacts, but the relationship to the indicandum needs to be tracible for indicators and indices to guide planning for improving ecosystem health (Frederiksen et al., 2021; von Thenen et al., 2020). Therefore, explicit relationships between the SDG indicators and EU MSFD descriptors and criteria can assist traceability across the SES system.

Figure 3 is a network diagram illustrating the relationships between the SDG indicators (specifically, Goals 14, 6, 15, 2, 12, and 11) and the EU MSFD descriptors and criteria, underpinned by the reviewed indicators developed by regional sea conventions and EU organisations for the EU MSFD (HELCOM, OSPAR and the EEA).

The analysis suggests that the highest number of links between the MSFD descriptors and SDGs are between D5 (Eutrophication) and SDGs 6 (Clean water and sanitation for all) and 14 (Life below water). Additionally, the results showed that D1 (marine biodiversity), D4 (food webs), D9 (contaminants in seafood), and D8 (Contaminants) have the broadest range of indicators from HELCOM, OSPAR, and the EEA. However, they are linked to only a few SDGs.

At the same time, the SDGs were often categorised as response and associated with the general theme of the descriptor rather than a specific criterion, which is further categorised as Other” in Figure 3. For instance, indicators for SDG goals 11 (sustainable cities and communities), 12 (responsible consumption and production), and 14 (life below water) were linked to D10 (marine litter).

Response indicators for the two contamination descriptors (D8 and D9) were SDG indicators for goals 12 (Responsible consumption and production) and 14 (Life below water). They focus on quantifying commitments to international multilateral environmental agreements on hazardous waste and other chemicals (SDG 12.4.1), as well as hazardous waste generated per capita, waste treated, and treatment applied (SDG 12.4.2). Additionally, SDG indicators for goal 6 (clean water and sanitation) focus on quantifying government-coordinated spending plans for water and sanitation (SDG 6.a.1) and local community participation incentives in water and sanitation management (SDG 6.b.1), which also apply to eutrophication (D5; 29). The SDG indicators quantify urban waste generated and discharged by cities (SDG 11.6.1), national recycling rate (SDG 12.5.1), levels of fine particulate matter in cities (SDG 11.6.2), sustainable tourism strategies (SDG 12.b.1), research budget marine technology (SDG 14.a.1), progress in ratifying, accepting, and implementing instruments for conservation and sustainable use of the oceans and their resources (SDG 14.c.1), and implementing ecosystem-based approaches (SDG 14.2.1).

Definition of a social–ecological indicator

One aspect common to all these different applications of indicators in SES analysis, including their links and similarities to the butterfly model and to the compound index of OHI and BHI, is the understanding that a SE indicator is not necessarily a single measurement. Instead, it explicitly captures one or more links between the ecological and social sides of the butterfly model. Accordingly, this type of compound indicator should represent several elements spanning both sides of the butterfly and make the links between these assessed elements explicit. One of these elements should describe ecosystem state (i.e., ecosystem structures, ecological processes, and ecosystem functions). Examples of both normative and descriptive indicators are found in the analysis of existing indicators in the assessed list and in other studies, and an SE indicator should span both. Thereby, a SE indicator can be a combination of simple social and ecological indicators or a compound index, as illustrated by the OHI, but with explicit links to what they/it measure(s). In this way, a SE indicator is a complex, compound indicator that explicitly links the ecological and social aspects of the butterfly model. It is not merely an index number due to the requirements of explicit links to the butterfly elements that rule out the choice to reduce the number of indicators, but it can be a compound index on the conditions of explicit links to the butterfly through sub-categories of interlinked indicators.

The development of a narrative which tells a storey of how different parts of a SES influence each other, and how these influences can be measured—such as illustrated by von Thenen et al. (2020) and Murillas-Maza et al. (2020) in the context of ecosystem service analysis—is one way to build a SE indicator. This SE indicator approach is nested and captures several interconnected components of the butterfly model. For example, interlinked indicators that depict the flow of services and benefits following the ecosystem service cascade—specifically, the condition of ecosystem elements, their functioning, the supply of services and benefits, and their distribution—are SE by nature.

Social–ecological indicator abilities of selected indicators, grouped by MSFD descriptor

In the analysis of existing indicators, nine were found to comply with the definition of a SE indicator (Table 2; see Supplementary Data sheet 1), eight of which addressed activity, pressure, state, and one addressed pressure, state, impact, and response, due to their nature of making SE linkages explicit.

Three D1 (marine biodiversity) related SE indicators (Table 2) address human activities, pressures, and ecosystem state. These include OSPAR's indicators for marine mammal bycatch, pilot assessment of marine bird bycatch, and pilot assessment of marine bird habitat quality. The goal of D1 is that marine ecosystems (benthic and pelagic species and habitats) are healthy, diverse, and resilient, and that human activities do not adversely affect them. All three indicators measure bycatch, the incidental capture and entanglement in fishing gear. Bycatch assessments count the number of drowned mammals (cetaceans and seals) and waterbirds, reporting on additional incidentally occurring mortality, and the effects of bycatch on the population.

D6 (seabed integrity; 21) has the most SE indicators identified, including those linking activity, pressure, and state. They combine data on habitat sensitivity maps (which consider resilience and resistance to individual pressures) with physical pressure maps. D6 aims to secure monitoring and action to protect the seafloor structure and the functioning of benthic communities. Four of the five OSPAR SE indicators for D6 focus on monitoring the extent of physical disturbance to benthic habitats from aggregate extraction, oil and gas exploration, production, and decommissioning activities, as well as bottom trawling (e.g., soft-bottom macrofauna). Sentinel benthic species, also referred to as keystone species in ecology, are organisms that live on or within the seabed and serve as early-warning indicators of ecological change or human-induced pressures such as pollution, sedimentation, nutrient enrichment, or, in this case, bottom trawling. OSPAR's sentinels of the seabed indicator provides a geospatial prediction of sentinel species proportions and habitat environmental status (low, moderate, and high disturbance). Benthic habitat distribution, pressure distribution and intensity, and biological data on the abundance and biomass of benthic species from each habitat are needed for this assessment. This data is combined with sensitivity indices to identify sentinel species and calculate pressure-state gradients. The focus on trawling pressures explicitly links the results to a single human activity.

The fifth D6 SE indicator is HELCOM's Cumulative impact from physical pressures on benthic biotopes assesses the additive, cumulative impacts from physical pressures on benthic habitats of several physical pressures in the Baltic Sea. In contrast to OSPAR's SE indicators for D6, HELCOM's cumulative impact from physical pressures on benthic biotopes indicator combines multiple pressures from bottom trawling, aquaculture, sediment extraction/disposal, pipelines and cables, platforms and wind farms, coastal protection, and shipping. Each pressure is analysed separately for magnitude, based on frequency, intensity, and range, as well as the areas where human activities impact the seabed. The impact of physical pressures on benthic biotopes is assessed by combining pressure magnitude with sensitivity data in a matrix that classifies impacts as very low, low, moderate, or high.

A single SE indicator was identified in relation to D3 (Commercial fish and shellfish)—HELCOM's total allowable catch (TAC). TAC is a catch limit set for a particular fish stock for a year or a fishing season. It is defined annually for commercially important Baltic fish stocks (e.g., cod, herring, sprat, salmon, and plaice) using data on fishing opportunities, catch, and effort for the Baltic Sea ecoregion. It informs regulations at both the Baltic and EU levels. TAC links observations of ecosystem state (e.g., fish species abundance) and fishing pressures (e.g., species removal) to the availability of provisioning ecosystem services (e.g., fish for food) for society and facilitates responses through measures such as quotas. Therefore, TAC addresses Pressures, State, Impacts and Response in the butterfly model.

This study provides a definition of SE indicators based on a literature study and the applied DAPSI(W)R(M) butterfly model to explicitly link and assess changes across environmental and social domains. It applies the definition as part of an extensive indicator analysis to assess the SE indicator abilities of existing marine planning-related indicators. More specifically, it provides a thematically organised overview of the social and environmental indicators explicitly linked to the EU MSFD and the UN SDGs. Of the 148 reviewed indicators, nine were identified to fulfil the SE indicator definition, linking activities, pressures, state, pressures, impact, and response components of the butterfly model. In addition to the existing SE indicators, the results of the indicator analysis can inform approaches to transforming non-SE indicators into SE indicators, as required by Digital Twin of the Earth initiatives such as the EU DTO, to better simulate the future state (Tzachor et al., 2023). In practise, this could involve the development of narratives or storylines to demonstrate how the underlying indicators link different parts of the cascade—ecosystem capacity, services, benefits, and values—as exemplified by von Thenen et al. (2020) and Murillas-Maza et al. (2020). Pathways for developing links between indicators are described in the following sections.

Pathways for the development of SE indicators

Similarly to the OSPAR's bird and mammal bycatch indicators, HELCOM's bycatch indicator addresses D1 (marine biodiversity) and D4 (food webs). It measures the number of drowned mammals and waterbirds in fishing gear and has a direct link between human activities (fishing) and pressures (species extraction). However, it does not directly capture the effects of bycatch on the ecosystem state. With a stronger link to, e.g., the effects of activities and pressures on population dynamics, species abundance, or the extent of habitats and species adversely affected by bycatch, it could serve as a SE indicator. Similarly, the fishery operations indicator (D3 commercial fish and shellfish) addresses pressures from fishing and impacts on human welfare; however, it lacks an explicit link to ecosystem state (e.g., evaluation against population thresholds based on abundance data to assess long-term sustainability). Fishery operations assessment provides information on the structure and economic performance of fishing fleets, specifically the number of fishing vessels, total vessel tonnage, full-time equivalent employees, gross value of landings, and weight of landings, for the fleets of Baltic Sea countries. This indicator connects pressures caused by fishing (e.g., species removal assessed by the number of fishing vessels and total vessel tonnage) to the availability of provisioning ecosystem services (weight of landings—fish for food) and the associated economic (dis)benefits to society (e.g., full-time equivalent employees, gross value of landings) within the framework of the common fisheries policy (CFP). The outcome of this assessment is reported in the Annual Economic Report on the EU Fishing Fleet (STECF 21-08). A more explicit link to the ecosystem state is necessary to make the current indicator a SE one.

D5 (eutrophication) ensures EU member states monitor and act to achieve a more nutrient-balanced marine environment by reducing nutrient levels entering marine waters from land-based human activities. It involves measuring nutrient concentrations in the marine environment and assessing both the direct and indirect effects of nutrient and organic matter inputs on marine ecosystem components. The indicators for D5 are predominantly state (e.g., chlorophyll-a concentrations, water transparency, harmful cyanobacterial bloom events, dissolved oxygen) and pressure (nutrient inputs and complex indicators/indices for assessing eutrophication levels). To further develop the D5 state and pressure indicators into SE indicators, links to the social side would include activities related to nutrient inputs, such as agriculture and aquaculture, and land-use changes that affect diffuse nutrient inflows, for example, as monitored using the agricultural nutrient balance indicator. This indicator measures whether excess or insufficient nitrogen and phosphorus are being applied to agricultural land, assessing the sustainability of nutrient management and its connexion to excess nutrients entering marine waters. Nutrient balance is calculated as the total nutrient inputs minus the total nutrient outputs, where input refers to nutrients added to soil (manufactured fertiliser or manure), and output includes nutrients removed by crops or livestock. Additionally, links to impacts on human welfare, for instance, an existing indicator of harmful cyanobacterial bloom events, could be established. These adverse effects of nutrient enrichment and significant impacts on human welfare (e.g., disbenefits to human health and tourism) can be measured in socio-economic terms. The European Environment Agency indicator for nutrients in Europe's transitional, coastal, and marine waters, which measures ecosystem state, along with HELCOM's wastewater treatment and SDG 6.3.1 proportion of wastewater safely treated (impact on welfare and response indicators), could give a more explicit link between the social and environmental sides of the butterfly model. Additionally, costs associated with wastewater treatment can serve as an indicator for estimating the loss of nutrient regulation services resulting from habitat and ecosystem function degradation and the associated benefits (see HELCOM, 2023; p. 58 onwards).

Many D6 (seabed integrity) indicators assess the extent of the area affected by pressures and state of habitats separately. However, several indicators combine the assessments to evaluate the extent of adverse effects from human activities and pressures on the state of different benthic habitat types and ecosystem functions. In some cases, the link to activities is implicit or instrumental in mapping the magnitude and location of pressure. For instance, OSPAR's condition of benthic habitat communities: the common conceptual approach indicator quantifies the relationship between the state of benthic habitats (characterised by particularly vulnerable, sessile species) and pressures. The indicator requires data on benthic community species abundance, substrate and water characteristics, and type, intensity, and frequency of pressures to measure and compare benthic species richness/diversity over time and space. The indicator is an index value, which, if low, indicates stress stemming from pressure affecting a benthic community and a decline in the quality of the community state.

Human activities that contribute to the spread of non-indigenous species (addressed by the MSFD D2 non-indigenous species) include shipping, fisheries, aquaculture, and recreational activities, among others. Determining and analysing the cause of non-indigenous species spread (e.g., climate change or ballast water) and the ecological and social effects (e.g., impacts on the welfare of fishing communities) would make these SE indicators.

D8 (contaminants) focuses on monitoring chemical pollution in marine waters and taking action to reduce contamination to levels that are safe for both ecosystems and people. Most D8 indicators are state (e.g., concentrations of heavy metals in sediment, shellfish, fish, and mammals) and pressure (e.g., liquid discharges from nuclear installations and oil spill events) indicators. Some, such as the EEA Bathing waters with excellent water quality in Europe and the UN SDG indicator 6.3.2 proportion of bodies of water with good ambient water quality, monitor water quality through observations of microbiological and hazardous substance pollution but are regarded as response indicators because of their close link to EU directives such as the urban wastewater treatment directive and their role as policy-effectiveness indicators. Establishing more explicit links to wastewater treatment and collection activities, as well as their impacts on human welfare, for instance, the disbenefits of poor water quality on human health, would make these indicators more SE in nature.

Relatedly, D9 (contaminants in seafood) supplements D8 by focusing on public health and contaminant levels in seafood. All D9-related indicators implicitly address impacts on human welfare. However, the focus remains on monitoring the species' condition. The EEA Hazardous Substances in Marine Organisms in Europe's Seas indicator, for example, monitors the concentrations of nine hazardous substances in eight mussel and oyster species. There is potential to establish links to sources of contaminants, such as point-source locations and contaminant flow rates. Moreover, it is beneficial to establish links to the effects of contamination in biota (e.g., fish, shellfish) on provisioning and regulating and maintenance ecosystem services, as well as the changes in human welfare related to disbenefits (e.g., health issues). Links to indices used in ecotoxicology, such as pharmaceutical and heavy metal concentration analysis and their adverse effects on human health through seafood consumption, could be used to expand State and Pressure indicators to account for the social dimensions of the butterfly model.

D10 (marine litter) requires monitoring and action to reduce and prevent marine litter. D10 indicators address micro and macro litter on the coast, seabed, the water column, and litter ingested by animals. D10 indicators are all pressure indicators that assess the amount of litter in the environment, as well as its ingestion by mammals and other species affected by litter. To develop the existing MSFD D10 indicators into SE indicators, an explicit link to activities resulting in litter should be established (e.g., fishing, shipping, recreation, and tourism), as well as links to the adverse effects of litter on ecosystem services and the (dis)benefits of the littered marine and coastal environment. These indicators, particularly those for litter ingested by mammals and for individuals with negative health impacts, can become SE once the affected species have measurable economic consequences and long-term effects on human welfare.

Lastly, D11 (energy, including underwater noise) requires monitoring and action to control underwater noise and other energy input levels to ensure they do not harm marine species. Pressure indicators for the D11 monitor and estimate the distribution of impulsive sound, assess impact risk, and evaluate continuous/ambient noise. The distribution of pressure is often explicitly linked to an activity and its location, for instance, in impulsive sound locations associated with underwater explosions, active sonars, and pile driving, as well as continuous noise sources such as shipping, fishing, boating, and offshore renewable energy installations. To develop these pressure indicators into SE indicators, a more explicit link to ecosystem state, e.g., spatial overlap with species sensitive to noise, would be necessary.

Social–ecological indicators under development to support SES approaches

In the near future ongoing DTO initiatives such as SEADITO provide opportunity to operationalize and demonstrate the applicability of DAPSI(W)R(M) frameworks under a pool of different scenario-based socio-ecological modelling approaches for coastal and marine environments.

A cumulative impact index explicitly including human activities and temporal dimensions

Activities are often a source of not one or two, but a whole range of pressures on the ecosystem, which can have immediate effects or become apparent only over time (Reckermann et al., 2022). Moreover, these activities rarely occur in isolation. The multitude of pressures on ecosystem activities is known to produce cumulative impacts, that is, impacts greater than the sum of individual effects (Reckermann et al., 2022). A key limitation of the widely used cumulative impact assessment (CIA) approach developed by Halpern et al. (2008) and applied in later assessments Halpern et al. (2025) is that it is not a truly socio-ecological framework. The method considers social dimensions only implicitly, as it focuses primarily on human-derived (for example, light pollution, nutrient load) and climate change induced pressures (for example ocean warming, sea level rise) and their biophysical effects on ecosystem state, without explicitly representing social processes, dynamics, or outcomes. This makes the traditional CIA index challenging to apply in a scenario context, as planning regulations usually relate to human activities rather than pressures directly, and because the links between human activities and pressures are complex rather than one-to-one. Nevertheless, recent efforts have sought to adapt the traditional approach to measure cumulative impacts and assess management effectiveness (Halpern et al., 2015), but without explicitly linking to the butterfly model or separating human activities from pressures.

Expanding the CIA approach to encompass human activities and linking them to the pressures they create will directly incorporate the social aspect of the butterfly model by enabling scenario analysis based on future or potential changes to human activities. The current CIA implementations also lack the temporal dimension and support for what-if scenarios. An updated CIA implementation is being developed as part of a suite of SE models for the EU-DTO (Hansen, 2025). This new approach to CIA, called MYTILUS-R, will support MSP as a forward-looking process through what-if scenarios supporting explorative stakeholder involvement. It is developed in R and will be available on Zenodo and GitHub.

The DAPSI(W)R(M) concept is applied to facilitate scenario-building and add a temporal dimension. The cumulative impact is not directly part of DAPSI(W)R(M) but can be considered an indicator of the state of the environment and the impact on ecosystem services. Similarly, the cumulative pressure index shows the spatial distribution of the aggregated pressures towards the ecosystem. The MYTILUS-R suite will provide several analytical diagnostics to support the identification of adverse impacts on the ecosystem, the development of suitable measures to mitigate those impacts, and the application of a scenario-based approach to compare various options.

Marine cumulative socio-ecological vulnerability assessment (CMSEV)

Marine socio-ecological vulnerability assessment (MSEV) frameworks describe how coastal and marine systems—including ecosystems, communities, and economies—are exposed and sensitive to environmental, social, and economic stressors (Cabanillas et al., in preparation). MSEV is a compound index that reflects both ecological vulnerability (e.g., habitat degradation, declining fish stocks, or visual blight) and social vulnerability (e.g., reliance on commercial fish stocks) and the adaptive capacity of coastal communities to change (e.g., level of education, employment, GDP/capita). Cumulative MSEV (CMSEV) occurs when multiple stressors interact over time and space (e.g., anthropogenic pressures, climate change, and infrastructure projects), intensifying impacts on both people and ecosystems, reducing the system's overall resilience. For instance, in the fisheries sector, where offshore wind farm development creates competition for spatial use, acting concurrently with climate change induced range shrinkage of fish stocks. Studies on socio-ecological vulnerability assessment (SEV) in marine realms have found application in diverse sectors of the Blue Economy across the globe, but especially in the fishery sector: applications of socio-ecological vulnerability assessments have been carried out along the U.S. West Coast (Koehn et al., 2022) and in small-scale fisheries in Baja California, Mexico (Micheli et al., 2024). These studies examine the ecological and social impacts of climate change on commercially important fish stocks and highlight the vulnerabilities of coastal fishing communities to species range shifts driven by extreme events. Similarly, research in Moorea, French Polynesia (Thiault et al., 2018), focuses on small-scale fisheries, emphasising how interactions among ocean users can lead to risks such as the depletion of ecosystem services tied to food resources. Other applications include event-based environmental and social vulnerability assessment in the context of oil spill risk modelling (Nelson and Grubesic, 2018), or applications to address ecological and social risks to beach-related recreational ecosystem services applied in the Italian Adriatic Sea (Depellegrin et al., 2023). On a Pan-European scale, spatially explicit vulnerability models were developed primarily for territorial planning. These includes, for example, the direct and indirect effects of the Green Deal on the European population, resulting in the development of a Regional Green Transition Vulnerability Index (Rodríguez-Pose and Bartalucci, 2024).

Future development of goal-specific social–ecological indices

The OHI/BHI is an example of a goal-specific SE indicator composed of multiple metrics aligned with a set of management goals, that is, statements about desired ocean conditions (Blenckner et al., 2021). OHI/BHI are highly adaptive—goals and weights can be selected to suit the context. However, they are also highly nested and challenging to apply in practise, as it is unclear how the goals are selected and how they relate to a mix of elements from the butterfly model (e.g., some goals are pressure-related, some are state-related, some are benefit-related). To make the goal-specific SE indicator more transparent, rather than aggregating it into a single health index, it could be measured as separate, interconnected indices that can be visually explored, thereby revealing the contributions from the present status, trend, social pressures, ecological pressures, and resilience, making the links to the different DAPSI(W)R(M) components clearly visible. Selecting one-at-a-time use-case SE goals to measure and assess their capability to reach them could also improve transparency. Additionally, common goals applicable across different use cases could be identified, e.g., achieving a favourable eutrophication status, and assessment results could be compared.

Criteria for SE indicator evaluation

In this paper, we focus on examining the conceptual link between indicators for social and ecological components of the butterfly model which has a general relevance for SES approaches in all planning and management environment. Nevertheless, the ongoing developments of digital twins of the earth such as the EU DTO are escalating fast, supporting findable–accessible–interoperable–reusable (FAIR) data and models that can benefit from better SE integration. The EU DTO has already entered phase two of development (2025–2028), with the aim to scale up and increase the range of tools available for decision-making support for researchers, policymakers, and other stakeholder on the platform (CORDIS, 2025).

SE indicator applicability has been evaluated for ecosystem service indicators by Hattam et al. (2015) and von Thenen et al. (2020) based on measurability, sensitivity, specificity, scalability, transferability, precision as the key criteria, but the list of criteria can be further extended to ensure (i) proper coverage of compound indicators and (ii) compatibility with the Digital Twin of the Earth (see Table 3 for the full list). The two further requirements for SE indicators are: a. Proper coverage of compound indicators: the SE indicators should explicitly demonstrate cross-cutting interlinkages across the social and environmental parts of the butterfly model. As argued by (HELCOM 2025, p. 24), the currently employed indicators, which measure a narrow range of parameters, do not address cross-cutting or horizontal issues. However, highlighting the connexions between different DAPSI(W)R(M) indicators would support the analysis of multiple pressures and impacts, and furthermore, provide insight into how well calibrated the state and pressure indicators and thresholds are, and if they are appropriate for measuring the effectiveness of responses or measures.

b. Digital Twin of the Earth compatibility: to ensure the SE indicators are ready for implementation in models, workflows, and tools into digital twins of the earth such as EU DTO, the readiness level of the data sources and their repositories for integration is crucial, as documented during the development of the workflows for the AquaINFRA virtual research environment and data discovery and access service (Rönneberg and Latvala, 2023). This is particularly important, as uneven spatial and temporal resolution of social data compared to biophysical data may impact readiness levels for analysis and integration into tools. Therefore, operationalised SE indicators must be empirically measurable, scale appropriate, scientifically sound and relationships tested for cause-and-effect to be operationalised into the EU DTO (Moldan et al., 2012).

Methodological reflections

Due to the focus on the MSFD and the UN SDGs for their regional and global implementation across several seas and the practical limitations of the study, most of the indicators reviewed in this paper are environmental. To expand the number of social and socio-economic indicators it would be important to also review scientific and grey literature, such as the United Nations Economic Commission for Europe (UNECE), System of Environmental-Economic Accounting (SEEA), the Biodiversity Finance Initiative indicators (BIOFIN), the EU SDG Indicators set developed by the EC in 2017 to monitor the SDGs in an EU context (see Hametner et al., 2023), the Economics of Ecosystems and Biodiversity (TEEB) Database, EUROSTAT, OECD, the IPBES framework for Nature's Contributions to People social indicators (e.g., Díaz et al., 2015), the System of National Accounts (SNA) providing economic market indicators, and the SEEA Ecosystem Accounting (SEEA EA) for non-market indicators. Additionally, further environmental indicators from the Global Biodiversity Outlook and the EU Biodiversity Strategy 2020 and 2030 can be considered.

This study follows a SES approach by applying the butterfly model to make a comprehensive analysis of an existing pool of indicators in relation to the DAPSI(W)R(M) elements. The findings argue for combining indicators into compound SE indicators that explicitly highlight SE linkages. Compound indices can act as SE indicators if they are more than just aggregated numbers, making the links to social as well as ecological elements in the butterfly model traceable. Thereby, a SE indicator is not simply a composite index, nor merely a combination of social and ecological metrics, but a compound, explicitly linked representation of interactions across both wings of the butterfly model. At a minimum, it must connect ecological state with one or more social dimensions such as activities, impacts on welfare, or responses.

The analysis of the indicator pool and indicators in literature illustrates that most existing marine indicators—particularly those developed under the MSFD—remain predominantly ecological, with a strong emphasis on state and pressure metrics and limited explicit treatment of impacts on human welfare and societal responses. Nevertheless, nine out of the 148 analysed indicators from HELCOM, OSPAR, the EEA, and UN SDGs qualified as SE indicators due to their explicit linkages to the SE butterfly sides. Linking MSFD indicators with UN SDG indicators enables a hybrid, normative basis, as SE indicators cannot remain purely descriptive, since they need to address impacts on welfare as well as on the ecological state. Besides requiring a normative basis, relationships with the indicandum need to be as explicit as possible for SE indicators to ensure transparency and real-world relevance.

The analysis sets the pathways for future SE indicators to be developed and assessed. In the SEADITO (2024–2027) project, as part of developing tools for the EU DTO, ongoing work is underway to develop SE indicators based on improvements to the CIA method by Halpern et al. (2008) and to define a pan-European CMSEV for the marine domain. Existing compound indices such as the OHI/BHI illustrate the feasibility of nested, goal-oriented SE indicators, but also highlight challenges related to transparency and traceability. Therefore, future compound SE indicators should highlight the relationships to individual components in the butterfly model in addition to acknowledging the dynamic SE links between these components, rather than collapsing complexity into opaque aggregate scores such as the OHI/BHI, which is somewhat transparent, although its strengths are in integrating SE perspectives. Digital Twin Oceans (DTOs) strengthen socio-ecological modelling by integrating ecological, physical, socio-economic, and behavioural data, helping address gaps in current SES indicators and modelling approaches. A particular strength of DTO initiatives like SEADITO is the operationalisation of scalable, spatially explicit models aligned with DTO architectures, providing pools of SES tools, such as OHI/BHI, CEA, MSEV or coastal vulnerability assessments. DTOs also enable dynamic “what-if” simulations for evaluating policy impacts and cumulative pressures, as highlighted in SEADITO. Through interoperable and FAIR data standards, DTOs increase the transparency and robustness of social–ecological assessments and their usefulness for marine planning.

The paper situates the indicator within the analysis of the social–ecological chain of cause-and-effect and recognises the unprecedented opportunities for integrating cross-sectoral data, scenario modelling, and what-if analyses provided by DTOs while also considering the specific requirements implemented by such platforms. To be compatible with such platforms, SE indicators must meet extended criteria beyond classical measurability and sensitivity, including crosscutting interlinkages and digital readiness. This includes harmonised spatial and temporal scales, tested cause-and-effect relationships, and FAIR-compliant data structures.

Overall, this research provides a conceptual definition of SE indicators, a diagnostic review of current practise, and a set of operational criteria for future development. It concludes that meaningful support for marine spatial planning and ecosystem-based management depends on moving beyond predominantly ecological monitoring frameworks towards explicitly integrated, traceable, and digitally operational SE indicators. Such indicators are essential for enabling advanced decision-support systems and scenario modelling within next-generation digital environments such as the EU DTO, thereby strengthening the capacity to manage marine ecosystems sustainably while safeguarding human well-being.