A goal-based approach to the identification of essential transformation variables in support of the implementation of the 2030 agenda for sustainable development

ABSTRACT

The United Nations' 2030 Agenda for Sustainable Development sets seventeen Sustainable Development Goals (SDGs) to be achieved by 2030. Earth observation are needed that can support the development and validation of transformation policies to make progress towards the SDGs. A participatory and inclusive goal-based approach (GBA) is introduced that links societal goals, targets and indicators to Essential Transformation Variables (ETVs) of the human and non-human environment. The GBA is complementary to the widely used expert-based approach. The GBA is applied to the SDGs at the goal, target and indicator levels. The high-level conceptual model used for the SDGs is humanity embedded in the Earth's life-support system (ELSS). At the goal level, very few of the SDGs are directly focusing on the ELSS and its physiology. Most of the SDG Targets focus on transformations in society and the built environment. Having targets that explicitly focus on the physiology of the ELSS would be important for sustainability. Most of the current indicator measure the built environment and the embedded social fabric. Sustainable development requires a functioning ELSS, and to ensure this, complementary indicators that bring environmental aspects to the monitoring of SDG targets are needed.

KEYWORDS:

• 2030 agenda for sustainable development
• sustainable development goals
• targets
• indicators
• essential variables
• policy making
• decision making
• tranformation science

1. Introduction

The implementation of the seventeen Sustainable Development Goals (SDGs) of the 2030 Agenda for Sustainable Development (United Nations 2015) pose wicked problems to society. Wicked problems are social or cultural problems that are difficult or impossible to solve because of incomplete or contradictory knowledge, the number of people and opinions involved, the large economic burden associated with progress towards a solution, and the interconnected nature of these problems with other problems (Rittel and Webber 1973). All of this applies to the SDGs: Knowledge on how to make progress towards the SDGs is incomplete and contradicting, reaching the SDGs even on a local level involves the whole of society, making progress requires a rethinking of economy (UNRISD 2016), and the goals are strongly interconnected (e.g. Nilsson, Griggs, and Visbeck 2016; ICSU 2017; Singh et al. 2018) and there are many interactions between the individual goals that are variable across different economic, social, and cultural settings (Jules-Plag and Plag 2016a). For example, poverty (SDG 1) is linked with education (SDG 4), nutrition (SDG 2) with poverty, the economy (SDG 8) with nutrition, and so on. While wicked problems related to the same issue often are grossly similar, they are discretely different, which necessitates each problem to be addressed individually. Solutions cannot be generalized. Poverty in, e.g. California is grossly similar but discretely different from poverty, e.g. in Angola, and there is no practical set of characteristics that defines poverty.

Moreover, many aspects of the SDGs constitute super-wicked problems. Super-wicked problems have four additional characteristics: (1) time is running out; (2) there is no central authority to address the problem; (3) those seeking to solve the problem are also causing it; (4) policies discount the future irrationally (Levin et al. 2012). While the characteristics that define a wicked problem relate to the problem, those that define a super-wicked problem relate to the agent trying to solve it.

Wicked and super-wicked problems can hardly be addressed in the framework of traditional discipline-based science, and a transdisciplinary approach is needed to tackle these problems (Brown, Harris, and Russell 2010). The emerging fields of adaptation science (Moss et al. 2013) and sustainability science (Clark and Dickson 2002; Miller et al. 2014) are therefore inherently transdisciplinary.

A particular challenge of the quest for sustainability and the SDGs arises from the need to create transformation knowledge guiding the development of policies and means to make progress towards the SDGs. Science needs to support society and interact with societal agents in effort to work out this transformation knowledge. The SDGs present policy makers with a complexity individually and through many interconnections (e.g. Jules-Plag and Plag 2016a 2016b; Obersteiner et al. 2016; Nilsson, Griggs, and Visbeck 2016). At the same time, the unsustainability of the current global trajectories of society and the Earth's life support system (ELSS) (e.g. Rockström and Klum 2015; Steffen et al. 2015) and the global consensus of reaching these goals introduces an unparalleled urgency to develop the necessary transformation knowledge. A major gap exists in the absence of an epistemology for the creation of transformation knowledge (Miller et al. 2014). While there are increasingly efforts to carry out transformation research in ‘real-world laboratories’ (e.g. Evans and Karvonen 2011; Karvonen and van Heur 2014; Sengers et al. 2016), there is no thorough epistemological approach available for this new type of research.

Sustainability science is a developing research field addressing complex socio-ecological problems of our time ranging from climate change and mass extinction to pandemics and rapid urbanization (Kates et al. 2001; Clark and Dickson 2002). A major goal for sustainability science is to provide knowledge in support of societal transformations towards sustainability. To achieve this, sustainability science generates, tests, and integrates (a) system knowledge about sustainability problems, (b) goal knowledge about desirable futures, and (c) transformation knowledge about disturbances and interventions that can lead from the current state to the desired futures (Figure 1). In the past, epistemological work primarily has focused on the creation of system and goal knowledge (Grunwald 2007, 2015). System knowledge is associated with the analysis of complex systems across different domains (e.g. society, environment, economy) and local to global scales and is essential for understanding and detecting complex sustainability problems. Transdisciplinary approaches integrating scientific and societal knowledge present challenges and opportunities for our understanding of sustainability problems (Wiek et al. 2012). Both the assessment of (un)sustainability and the envisioning of a sustainable world that we aim for involve normative claims. Goal knowledge has attracted the attention of epistemologists because of the importance of normative components requiring a rethinking of the standards used in producing and evaluating scientific knowledge in sustainability science (Miller 2013). It also has attracted the interest of ethicists because of the links of goals to normative ethics and the potential to reach the goals to descriptive ethics (e.g. Rieder 2016).

Figure 1. The three main parts of sustainability science. Sustainability science relies on three main kinds of knowledge: system knowledge, goal knowledge, and transformation knowledge. While the epistemology of creating system and goal knowledge is well developed, the epistemology of creating transformation knowledge is in its beginning.

Because of its transformational and transdisciplinary character, sustainability science differs from traditional modes of knowledge production. Sustainability science links system knowledge and goal knowledge through transformation knowledge (Figure 1). System knowledge informs about what might happen, the possible threats and hazards, and the past, current and potential future system trajectories. Natural sciences have focused on system knowledge and created a broad basis of that knowledge. Goal knowledge describes what we want to happen and what desirable futures we want to realize. Transformation knowledge identifies the interventions required to change the system trajectory and to facilitate pathways to desirable futures. Over the last few decades, social sciences have developed both the epistemology and methodology for the creation of goal knowledge. The elaborate process that led to the agreement on the seventeen SDGs exemplifies the level of goal knowledge that can be reached today, and a transition to global governance by goal-setting appears feasible (Biermann, Kanie, and Kim 2017). What is currently lacking is a fully developed transformation science (Grunwald 2007) that links the system and goal knowledge through the disturbances and interventions needed to ensure a progress towards desirable futures. Transformation science as part of sustainability science focuses on the identification of disturbances and interventions that can divert the ELSS from its current trajectory out of the ‘safe operating space for humanity’ (Rockström et al. 2009) onto a trajectory towards desirable futures closer to the agreed-upon goals expressed in the SDGs.

However, the epistemological basis for the creation of transformation knowledge has been neglected to a large extent (Grunwald 2015). A major unsolved problem in the epistemology of sustainability science is therefore the understanding of how transformation knowledge can be generated, tested, and validated. This raises important epistemological questions: How is knowledge for transformation produced? What is the role of experimental interventions in producing transformation knowledge? What theories can support knowledge production for transformational sustainability (Miller 2013; Miller et al. 2014)? It would be important to include these questions in a thorough gap analysis of the knowledge and data needs related to the SDGs and sustainability in general.

Developing the interventions to change the system trajectory in a desirable way is an iterative process (Figure 2). Any intervention through policies, organizational changes, and technologies needs to be validated as far as possible prior to implementation, which poses epistemic challenges due to the fact that a priori validation is impossible: only during implementation can the impacts be monitored and there is no chance to go back in time and try another intervention. Model simulations can be used to explore possible futures under different scenarios for drivers, an approach used, e.g. for the Millennium Ecosystem Assessment (Carpenter et al. 2005) or the assessment of future climate change (e.g. Stocker et al. 2013).

Figure 2. The iterative nature of bending system trajectories towards desirable futures. Achieving the transformation from the current state and trend to a desired future requires an iterative process of disturbances exceeding the system's resilience and corrections to bring the system's trajectory closer to the desired future.

The iterative nature of implementing transformation (Figure 2) requires detailed monitoring of the complex system trajectory after interventions in order to ensure that the resulting trajectory brings the system closer to the desired future and accepted goals and to detect in a timely manner the need for further interventions.

This brings up the important question of what are the observations that can support the creation and validation of transformation knowledge? What is essential to be observed to inform the development of interventions and support the assessment of the potential impacts prior to the implementation, monitor the impacts, and identify timely corrections? The concept of Essential Variable (EVs) provides one avenue to answer this question.

In the next section, we discuss different aspects of science and Earth Observation (EO) support for the implementation of the 2030 Agenda for Sustainable Development. Section 3 introduces the EV concept and reviews recent applications of this concept to different domains. In Section 4, we introduce a goal-based approach (GBA) to the determination of EVs, which is complementary to the widely used expert-based approach (EBA). Section 5 summarizes different goal sets for which knowledge of EVs would be a valuable input for the development of EO systems. Applying the GBA to the creation of transformation knowledge for the SDGs results in several conclusions about the linkage between Goals, Targets and Indicators to the conceptual model with implications for the Essential Transformation Variables (ETVs) (Section 6).

2. Science and EO support for SDGs

The SDGs have been accepted by the United Nations in late 2015 as the global goal set for the period of 2016 to 2030 (United Nations 2015). These goals guide the continuing quest for sustainable development already expressed by the Millennium Development Goals (MDGs) for the period 2005 to 2015. The seventeen SDGs cover a wide range of aspects of the global society and our interaction with the Earth system (see Figure 6 in Section 6 below). For each SDG, there are a number of associated Targets, which define the milestones to be reach by a certain time. In total, 169 targets have been agreed upon.

Being able to measure progress towards the targets associated with the SDGs requires metrics defined by a set of indicators. Developing indicators that provide useful quantitative metrics is a long process involving the scientific community (e.g. Alcamo et al. 2013; Leadership Council of the Sustainable Development Solutions Network 2014). With the aim to develop a manageable indicator framework, the United Nations Statistical Commission (UNSC) created the Inter-Agency and Expert Group on SDG Indicators (IAEG-SDGs). Based on the proposal of the IAEG-SDGs, a global indicator framework with a total of 232 global indicators was adopted in 2017 by the United Nations General Assembly as a voluntary and country-led endeavor to monitor the 2030 Agenda for Sustainable Development (United Nations 2018), and these indicators are complemented by indicators at the regional and national levels. According to the level of data availability and methodological development, the SDG Indicators have been grouped in three different Tiers: from Tier I, for the ones having an established methodology and widely available data, to Tier II and Tier III, for those having not data available or no methodology established, respectively. As of 11 May 2018, the updated tier classification contains 93 Tier I indicators, 72 Tier II indicators and 62 Tier III indicators (IAEG-SDGs 2018). In addition to these, there are 5 indicators that have multiple tiers (different components of the indicator are classified into different tiers) (see https://unstats.un.org/sdgs/iaeg-sdgs/tier-classification). Reaching from the SDGs indicators to the ETVs that allow for quantification of these indicators is a process that needs to be informed by the scientific community.

There are several roles for science and EO support in the implementation and monitoring of the SDGs. Figure 3 identifies research and tools in support of policy development (A in the figure), data integration across all domains and societal sectors (B), developing the metrics for policy validation and monitoring of progress towards the targets (C), and the quantification of indicators (D). These areas depend on scientific expert knowledge and EOs. However, this is often not sufficient. Science and EOs are highly relevant for developing and validating transformation policies and for monitoring progress. Developing successful transformation policies requires knowledge on possible system trajectories in response to interventions (Figure 2). The implementation of the transformation policies requires careful monitoring of progress to ensure the system's response to interventions is as desired. A major challenge to this science and EO support is the linking of scientific results and knowledge derived from EOs to the societal decision space. This necessitates the development of suitable conceptual models that integrate system science with societal decision making and support scenario-based exploration of future system trajectories..

Figure 3. EOs in support of transformation knowledge for SDG implementation and monitoring. A. Science support for the planning of actions and the development of policies. B. Identifying the ETVs across all social, environmental, and economic domains and facilitating the integration of ETV data across these domains is crucial for the creation of transformation knowledge. C. Applying the GBA to the SDG Targets. D. Applying the GBA to the current set of SDG indicators. Modified from Jules-Plag and Plag (2016a).

A major challenge is the development of conceptual models that can inform the development of transformation policies. A key issue in this development is related to decision making within the context of governance. Stakeholder participation is recognized as a key element in successfully implementing foresight-based programs to develop different transformation and adaptation policies (e.g. van der Helm 2007; Sales 2009). Addressing the wicked problems associated with progress towards sustainability and posed by the SDGs often requires non-standard approaches to development issues such as zoning, planning, resource utilization, production and consumption, and location, that are often associated with conflict and controversy. The most important impacts of participation of societal agents in addressing wicked problems are reduced controversy and greater public support for, and acceptance of, action and implementation. By facilitating the generation and exchange of information and then developing understanding and agreement on problems and their solutions (e.g. Burby 2003; Creighton 2005; Walsh 1997), participatory processes can allow for decision making that incorporates public values into decision, increases stakeholder ownership of decisions, resolves conflict among competing interests, builds trust in institutions, and ensures the representation of the interests of the disadvantaged and powerless groups affected by the decisions (e.g. Beierle and Crayford 2002; Buck 1984; Burby 2003; Creighton 2005; Kweit and Kweit 1981; Thomas 1995; Walters, Aydelotte, and Miller 2000). As a result, stakeholders may be more supportive of the policies, which makes the implemented solution more likely to be effective and longer lasting.

Stakeholder participation is crucial for the co-development of transformation knowledge (e.g. Mauser et al. 2013). This participation would benefit from advanced tools, which can utilize different metrics and scenario-based exploration of potential futures to inform policy makers and stakeholders. The engagement of practitioners in a companion modeling effort (e.g.Guyot and Honiden 2006) aiming at the development of conceptual models facilitates an iterative process with the goal of improving the metrics and the use of these metrics in the validation of the policies and the monitoring of progress. Implementing the conceptual models in Stock-and-Flow Models (SFMs) and extending them through Agent-Based Models (ABMs) and subsystem models provides a basis for simulations as a tool to explore policy impacts before implementation of the policies. SFMs are based on the notion that many important processes can be modeled as ‘stocks’ that change through ‘flows’ between these stocks. Identifying the stocks and the variables that determine the flows between them is an important step informing the development of metrics. ‘What if’ questions and scenarios allow policy makers to assess the potential impacts of proposed policies on the system trajectory and the progress they can facilitate towards the agreed-upon goals and targets. ABMs can facilitate the linking of the SFMs to the the decision making of societal agents (Epstein 2008). Advanced ABMs support the exploring of the societal response to new policies and can play an important role in validating transformation policies prior to their implementation.

3. Essential variables

Since 1984, the Integrated Global Observing Strategy (IGOS) initiated by the G7 as a framework for EOs was developed with the goal to identify what was essential to be observed in order to document the changes that are happening on the planet (Dahl 1998). In 1998, the Integrated Global Observing Strategy Partnership (IGOS-P) was established, bringing together major organizations in the scientific and EO fields in an effort to first identify what needs to be monitored and then to implement the corresponding observing systems. IGOS-P used a well-defined theme approach to define the overall strategy, which ‘recognises that in reality it is impossible, in one step and for all eventualities, to complete the exercise of defining all the necessary Observational Requirements (ORs) and hence the observational systems, data handling, processing and analysis infrastructure for a comprehensive global system. The theme approach allows the coherent definition and development of an overall global strategy whilst recognising the different state and stage of development in different areas. Themes have not a priori been defined; rather it is anticipated that the user communities will identify areas that require action and bring forward themes for agreement and action’ (IGOS-P 2003). The resulting IGOS-P theme reports were excellent outcomes of the first step defining observational needs for societally relevant themes (e.g. IGOS-P Ocean Theme Team 2001; Lawford and The Water Theme Team 2004; Marsh and The Geohazards Theme Team 2004; Townshend and The IGOL Writing Team 2004; IGOS 2006), but the implementation of the findings by the space agencies and other providers did not follow suit.

Subsequently, the concept of EVs has been used in a number of EO communities to identify and prioritize variables and observations that are key to the missions of these groups. Examples are the Global Climate Observing System (GCOS) under the United Nations Framework Convention on Climate Change (UNFCCC), which developed a set of Essential Climate Variables (ECVs) (Bojinski et al. 2014). Several ocean communities are engaged in developing sets of Essential Ocean Variables (EOVs) for marine, chemical, and physical aspects of the oceans (e.g. Lindstrom et al. 2012; Constable et al. 2016). Likewise, in biodiversity communities a discussion is in progress with the goal to identify a set of Essential Biodiversity Variables (EBVs) (e.g. Kissling et al. 2018). Here, too, feasibility is an important criteria, and Peterson and Soberón (2018) even doubt that a global set of EBVs can be found because of regional data gaps. However, many of the spatial gaps in data could be addressed by remote sensing efforts.

The definition of EVs varies significantly from community to community. For example, an ECV is a ‘physical, chemical, or biological variable or a group of linked variables that critically contributes to the characterization of Earth's climate’ (Bojinski et al. 2014). Adopting a similar approach for EOVs, the ocean communities emphasized the ‘readiness level’ of observations in the identification of EOVs. In biodiversity communities, the EBVs are less system state variables and more ‘between primary observations and indicators’ (Pereira et al. 2013; Geijzendorffer et al. 2016; Kissling et al. 2018). Experts in other fields have also initiated discussions of field-specific set of EVs. For example, Ranchin et al. (2018) discuss effort to determine EVs for renewable energies. Recently, a discussion on EVs related to the increasing plastics pollution in the marine environment (e.g. Provencher et al. 2018) has been initiated in several related communities.

Common to these efforts of identifying EVs is the main motivation to document the state and trends of the system under consideration and to improve predictive capabilities. Fundamental to the concept of EVs is a system thinking, which also has been applied to other human and non-human domains of the Earth system with the goal to identify variables that relate to the system's state and trends (see Reyers et al. 2017, and the references therein for a summary). In all cases, the identification of these variables requires a conceptual system model that reflects the main processes and system variables (see the comment on the challenge of developing conceptual models in Section 2 above.

So far, the compilation of sets of domain-specific EVs are based on input from experts in the relevant fields, who identify what they deem necessary to observe and what is feasible to observe. However, experts tend to be experts in their own often narrow field and do not fully grasp the down-stream impact on other disciplines. The link to the creation of transformation knowledge is weak. A link to societal benefits often is constructed after the EVs have been identified. We denote this approach as EBA, and this approach gives the current feasibility of observing a variable a high weight in the selection of what is essential. As a result, the identified EVs mainly address the areas B and D representing high feasibility in Figure 4. However, there are potentially high societal benefits if research and technological development would increase the feasibility of observing those EVs that fall within area A currently representing low feasibility.

Figure 4. Societal Impacts Versus Feasibility of Observations.

Several efforts have been made to apply the concept of EVs to the challenge of sustainable development and the implementation and monitoring of the SDGs (e.g. Stafford-Smith et al. 2017; Reyers et al. 2017). It is acknowledged that extending the EV concept to the SDGs and, more general, sustainable development and sustainability of our global civilization requires additional criteria for what is essential. Reyers et al. (2017) introduce four criteria for the identification of Essential SDG Variables (ESDGVs), requiring that these ESDGVs (1) capture the system essence; (2) link to system transformation; (3) capture key areas where coordination is needed; and (4) are indispensable. However, their approach continues to be expert-based and gives feasibility a high value in the selection of ESDGVs.

We introduce here a complementary approach to the identification of ETVs denoted as GBA that starts with agreed-upon societal goals and determines those system variables that are essential for the development of transformation knowledge and the monitoring of progress towards these goals without considering current feasibility of observing these variables. The GBA favors the areas A and B in Figure 4 that represent high impact. The system considered includes all relevant human and non-human components of the Earth system in an integrated concept. This complementary approach allows for a gap analysis originating in societal goals and comprehensively covering both what is feasible today in terms of observations and what would have high societal impact if it could be made feasible.

To get a useful definition of ETVs it is necessary to clarify the meaning of the terms ‘essential’ and ‘variable’ in their relationship to the creation of transformation knowledge. The adjective ‘essential’ has a number of different meanings, ranging from absolutely necessary and indispensable to containing an essence of something. A variable that significantly improves the reliability and accuracy of desired results can be considered essential. A variable that provides important information related to a specific goal is essential for this goal, independent of the capability to actually observe this variable. The essence of a variable may also depend on the knowledge and information needs of different communities of societal agents (science, policy, etc.). Therefore, the definition of an ETV needs to be formalized and endorsed according to the collective interest and preferably in an internationally recognized process. Here we propose to consider the following definition:

ETVs are ‘a minimal set of variables that are required to develop, validate, and monitor transformation policies and interventions that aim at achieving societally agreed-upon goals’.

Thus, for a given goal set, the ETVs are those that characterize the state and trends of the underlying system, and knowledge of the ETVs is required to increase predictive and backcasting (Robinson 1990) capabilities. In the case of the SDGs, the underlying system corresponds to the ELSS (see Griggs et al. 2013; Plag and Jules-Plag 2017, for a discussion of the concept of the Earth's life-support system) and the embedded global human civilization as well as the processes in which humanity interacts with the ELSS.

The required spatial and temporal resolution of observations of any given EV depend on the use of the observations. In some cases, available observations of an EV may not be ready for use for specific applications. Limited knowledge of EVs implies limited forecasting and backcasting capabilities and limited means to measure where the system is heading. The generic nature of the above ETV definition allows us to identify sets of ETVs for whatever system is being considered and for whatever goals are established in relation to this system. It allows for a broad application across different communities, and can be complemented with the specific requirements that each community may have related to the system the community is focusing on. Moreover, in a system of systems approach, sets of ETVs can be aggregated into larger sets supporting prioritization of efforts across community and system boundaries. For example, different subsets of closely interrelated SDGs could be used to determine relevant subsets of ETVs, which in a second step could be aggregated into a joint ETV set.

The concept of ‘variable’ embedded in the definition given above has a certain level of abstraction. Identifying a variable does not imply that observation requirements in terms of spatial and temporal resolution, accuracy, latency, observation interval, etc., are also specified. Nor does it imply that measurement instruments are available to observe the variable. In some cases, a variable may not be observable directly and has to be derived from a combination of observations. In such cases, the essential variable may have to be composed of a set of sub-variables that together provide the required information.

4. The goal-based approach to EVs

The GBA to the identification of ETVs that correspond to societal goals is depicted in Fig. 5. In order to implement this generic approach, knowledge is needed about applications, activities and tools used in the development of transformation knowledge, as well as for the quantification of the indicators and the assessment of progress towards the targets. A conceptual model that reflects the relevant processes as well as the agents and system variables is required for simulations that can help to explore possible futures under different scenarios for drivers. The development of the conceptual model requires participatory modeling efforts that bring together societal agents and experts. The conceptual model is the starting point for the development of agent-based and equation-based models that provide a basis for simulations and exploration of ‘What if’ questions that are necessary for policy validation and for an assessment of the quality of the indicators as a metrics for progress.

Figure 5. Goal-Based Approach for ETV Identification. The goal-based approach to the identification of essential variables starts from agreed-upon societal goals and targets (goal knowledge) and utilizes system knowledge to identify those system variables that are essential for the development and validation of the transformation knowledge required to make progress toward the targets. The participatory modeling process that engages all relevant societal agents in the development of a conceptual model as well as the identification of the ETVs relevant for this conceptual model and the validation of the policies include those required for the quantification of the indicators. ORs for these ETVs are a crucial input for the matching of these requirements to existing products, as well as the detection of gaps where a match is not possible.

As pointed out by Biermann, Kanie, and Kim (2017), the ‘governance through goals’ that is constituted by the SDGs is a novel approach that is inclusive in the goal-setting process and based on a non-binding set of goals. It relies on weak institutional arrangements and leaves extensive leeway to the individual states. It will be important to maintain these characteristics in the creation of transformation knowledge and the monitoring of progress towards the targets and goals. The GBA makes an effort to maintain the inclusive nature of the process and provide flexibility to those engaged in creating transformation knowledge and monitoring process.

Ideally, the process of developing transformation knowledge takes place in a participatory modeling setting that can engage all relevant stakeholder groups. Major outcomes of the participatory modeling effort are a conceptual model for the system associated with the goal set, sets of scenarios for core drivers, modeling tools to explore possible futures associated with the scenarios, and sets of ETVs.

The specification of ORs for the ETVs is addressed by domain-specific experts. There are different approaches to this, including the assessment of signal contents (spatial and temporal variability), sensitivity studies based on models, and a desired reduction of noise. The latter can be assessed based on knowledge of signal and noise levels, data analyses to quantify the current noise level, and theoretical studies of signals and noise. Having well-defined ORs is crucial for the design and sampling systems (traditional EO systems, collection of statistical data, citizen science programs, and big data analyses) and to ensure that the information collected for the ETVs is meeting the needs.

In a market place, these ORs can be matched to existing products derived from observations. In cases where a match is not possible, gaps can be identified. In many cases, ETVs will result from combination of processed observations. For the market place, it is important to have clear rules and practices how data needs to be processed and combined to generate products that meet the ORs for the ETVs.

While expert-based EVs are generally domain or subject specific, goal-based ETVs will vary from goal set to goal set and also depend on the associated targets and the indicators defined for monitoring progress toward these targets. While expert-based EVs reflect the observational needs and capabilities in a given domain in order to improve understanding, monitoring and predictive capabilities, the goal-based ETVs are essential for the development of transformation knowledge based on simulations using both forecasting and backcasting and the monitoring of progress towards the societally agreed targets.

For expert-based EVs, the processes that lead to the acceptance in a community of the EVs is in general a consensus-building deliberation in the community. For goal-based ETVs, the consensus-building societal processes focuses initially on the conceptual model underlying the goals and targets as well as the identification of indicators providing a metrics for transformation progress, with the latter requiring considerable input from the scientific community. The step from indicators to the corresponding ETVs allowing the quantification of the indicators is a more scientific process, requiring input from experts and the availability of a conceptual system model.

5. Selecting priority societal goals: a brief review

Community goals can be very different depending on the societal area the community is embedded in. For example, in scientific disciplines goals may be derived from the desire to create new system knowledge, for example, by monitoring the environment, assessing the state and trends of the Earth system, and improving the capabilities to predict future states and trends. For social science disciplines, the focus is on developing goal knowledge, and the goals may or may not be be related to sustainable development.

Four example sets of societal goals are summarized in Table 1. The notion of the presents being a transition to a new geological epoch (e.g. Steffen et al. 2016; Williams et al. 2016; Gaffney and Steffen 2017) would provide a basis for a global change metrics based on global change indicators. The necessity of remaining within the global boundaries of a Safe Operating Space for Humanity (SOSH) (Rockström et al. 2009) can be translated into a set of goals related to these boundaries. In this case, global boundary indicators would provide a metrics and a basis to identify variables that are needed to quantify these indicators. As mentioned in the introduction, Griggs et al. (2013) proposed a new definition of sustainable development, which implies the safeguarding of the ELSS. With this mission in mind, a set of planetary health indicators could be part of a sustainable development metrics, and essential variables and observation requirements could be derived from these indicators. Finally, supporting progress towards the global agreement on the SDGs would benefit from the identification of a comprehensive set of ETVs.

Table 1. Societal goals and associated targets, indicators, and EVs.

Area	Goal(s)	Target(s)	Indicator(s)	EVs
Global Change	Comprehensive Archive of global change	Comprehensive Monitoring	Global Change indicators	EGCVs
Global Boundaries of SOSH	Staying within the SOSH	Reducing excess of boundaries	Global Boundary Indicators	EGBVs
Sustainable Development	Safeguarding the ELSS	Maintaining healthy subsystems of the ELSS, e.g. ocean, ecosystems	ELSS or Planetary Health Indicators	EPHVs
Sustainable Development	SDGs	SDG Targets	SDG Indicators	ESDGVs

In all these cases, the approach to identifying the ETVs is a GBAs starting with goals at the highest level (Figure 5). Progress towards these goals would be defined through a set of targets, and measured through indicators. Both the targets and indicators can be used to identify a set of ETVs needed to quantify the indicators.

It is obvious that there would be considerable overlap between the ETVs for each of these goal sets. Therefore, an integrated set of ETVs well linked to societal benefits could provide valuable guidance for the prioritizing of EO investments.

6. Relationship of SDGs and targets with the ELSS

Applying the GBA to the SDGs can take place at three levels: Goals, Targets, and Indicators. In each cases, the underlying conceptual model has to be comprehensive. Guidance for a high-level conceptual model for the SDGs is provided by the redefinition of sustainable development as ‘a development that meets the needs of the present while safeguarding the Earth's life-support system, on which the welfare of current and future generations depends’ (Griggs et al. 2013). Here we extend this definition in an attempt to reduce the anthropocentric nature of the definition by including all generations of human and non-human animals. The interactions of all animal communities (as well as plant communities) with the ELSS consist of flows of energy and matter. In modern human communities, these flows are all controlled by economic rules, social norms and the descriptive ethics (Figure 6 left). As a result of the growth-based economy developed over the last three centuries, most of these flows have been accelerated by two to four orders of magnitude (e.g. Gaffney and Steffen 2017; Plag and Jules-Plag 2017) and new flows have been introduced. This extreme acceleration of flows is at the base of the modern challenges to global sustainability. Thus, a high-level conceptual model against which the SDGs can be assessed and in which progress towards the SDGs can be measured is humanity embedded in the ELSS accounting for all flows. Accounting for the different layers of this conceptual model, the ETVs are grouped into five groups: (1) people and social/ethical aspects, (2) governance and policy development, (3) infrastructure and built environment, (4) economy, and (5) the ELSS and its physiology.

Figure 6. Earth's life-support system and SDGs.

Left: Human communities are embedded in the ELSS and interact with the ELSS through flows of energy and matter. As a basis of a growth-dependent economy, most of the flows have been increased by several orders of magnitude. Modified from Plag and Jules-Plag (2017). Right: However, only three SDGs focus on the ELSS but take an anthropocentric point of view and none of the goals focuses on reducing the flows between ELSS and humanity.

Looking at the SDGs in their relationship to the ELSS reveals that seven of the goals focus on the social fabric while another seven goals are at the interface between this fabric and the built environment and economy (Figure 6 right). Only three goals are focusing on the ELSS, and these goals take a rather anthropocentric view on the ELSS as a service provided to human communities. The goals do not acknowledge the importance of planetary health of the ELSS as a value independent of the service for humans. Very few of the targets are explicitly focusing on reducing some of the flows or even limiting further acceleration.

Applying the GBA directly to the SDGs may be helpful in identifying the ETVs supporting the high-level creation of transformation knowledge. Current attempts to identify the necessary transformations to make progress towards the goals focus mostly at the goal level. For example, UNRISD (2016) discusses the principle social and economic transformations required to reach the goals. TWI2050 - The World in 2050 (2018) also works at the goal level and identifies six main transformations in the areas of: i) Human capacity and demography; ii) Consumption and production; iii) Decarbonization and energy; iv) Food, biosphere and water; v) Smart cities; and vi) Digital revolution.

Most of the SDGs are strongly interconnected with each other and there are many positive and negative interferences between the goals (e.g. ICSU 2017; Singh et al. 2018). Looking at the ELSS and sustainability through the eyes of a single SDG can help to understand the interferences (Plag 2017b). Taking the perspective from one individual goal can reveal these challenging interferences. For example, Waage et al. (2015) considered the transformations required to achieve SDG 3 (Ensure healthy lives and promote wellbeing) and found that several other goals are not likely to be approached in ways supporting progress towards SDG 3.

Taking the perspective of SDG 2 reveals important gaps in the SDG framework (Figure 7). Producing food for a rapidly growing population requires an acceleration of flows in the Earth system, including a flow of energy, which potentially impacts progress towards SDG 7 and SDG 13, and also contributes further to the overload of the ocean with carbon, which is hampering progress towards SDG 14. More energy is used to increase the availability of fertilizers (mainly Nitrogen and Phosphorous), and to facilitate land use changes, both to compensate for conversion of arable land into urban areas and the demand for more arable land resulting from population growth. These processes interfere with SDG 6, SDG 15, and also SDG 12. The increased use of fertilizers and deforestation further amplifies the already occurring overload of the ocean with nutrients, leading to an increase in seaweed/sargassum and dead zones in the ocean, which is interfering with progress towards SDG 14. The ocean overload with nutrients stimulates harmful algal blooms leading to health issues impacting progress towards SDG 3. Finally, the combined processes triggered by efforts to reach SDG 2 also increase extinction that can impact food production and thus hamper progress towards SDG 2. What is missing to achieve SDG 2 is a goal addressing population growth, which in Figure 7 is denoted as SDG 0 ‘Responsible procreation’ (Plag 2017a).

Figure 7. Interferences of SDG 2 efforts with other SDGs. Efforts to meet the food needs of the rapidly growing population have the potential to hamper progress towards a number of other SDGs. Without a goal to achieve responsible procreation, it appears un tenable to achieve SDG 2 without compromising several other SDGs.

Applying the GBA to the SDG Targets provides further insight into the level at which the targets address the flows that challenge global sustainability. A key challenge is associated with the acceleration of the flow of carbon from fossil fuel deposits into society and from there into the atmosphere and ocean. The increase in this flow was caused by humanity's demand for energy, which is addressed by SDG 7 (Ensure access to affordable, reliable, sustainable and modern energy for all; Table 2). Among the many impacts of the accelerated carbon cycle are climate change and ocean acidification. SDG 13 aims to ‘take urgent action to combat climate change’ (Table 2).

Table 2. Targets for SDG 7 and SDG 13.

SDG-7	Ensure access to affordable, reliable, sustainable and modern energy for all
Target 7.1	By 2030, ensure universal access to affordable, reliable and modern energy services
Target 7.2	By 2030, increase substantially the share of renewable energy in the global energy mix
Target 7.3	By 2030, double the global rate of improvement in energy efficiency
Target 7.a	By 2030, enhance international cooperation to facilitate access to clean energy research and technology, including renewable energy, energy efficiency and advanced and cleaner fossil-fuel technology, and promote investment in energy infrastructure and clean energy technology
Target 7.b	By 2030, expand infrastructure and upgrade technology for supplying modern and sustainable energy services for all in developing countries, in particular least developed countries, small island developing States, and land-locked developing countries, in accordance with their respective programs of support
SDG-13	Take urgent action to combat climate change and its impacts ${}^{\ast }$
Target 13.1	Strengthen resilience and adaptive capacity to climate-related hazards and natural disasters in all countries
Target 13.2	Integrate climate change measures into national policies, strategies and planning
Target 13.3	Improve education, awareness-raising and human and institutional capacity on climate change mitigation, adaptation, impact reduction and early warning
Target 13.a	Implement the commitment undertaken by developed-country parties to the United Nations Framework Convention on Climate Change to a goal of mobilizing jointly 100 billion annually by 2020 from all sources to address the needs of developing countries in the context of meaningful mitigation actions and transparency on implementation and fully operationalize the Green Climate Fund through its capitalization as soon as possible
Target 13.b	Promote mechanisms for raising capacity for effective climate change-related planning and management in least developed countries and small island developing States, including focusing on women, youth and local and marginalized communities (${}^{\ast }$ Acknowledging that the United Nations Framework Convention on Climate Change is the primary international, intergovernmental forum for negotiating the global response to climate change)

The targets of SDG 7 focus on providing (unlimited?) energy to everybody, a transition to renewable energy, and an increase in energy efficiency. None of the targets for SDG 7 acknowledges that the sources of energy used by humanity is only one aspect of the challenge and that the question of what energy is used for is equally important. In an effort to meet the needs of a growth-based economy, most of the energy used by humans is used to accelerated other flows within the ELSS (including, e.g. nutrients, pollutants, plastics, bio resources, mineral resources) and this leads to a significant reduction in the planetary health of the ELSS (e.g. Diaz and Rosenberg 2008; Barnosky et al. 2012; Mora et al. 2013). Utilizing the conceptual ELSS model, a reduction of the total energy used to modify the ELSS is required to decelerate these flows, but none of the SDG 7 Targets addresses this requirement (although Target 14.1 aims to prevent and significantly reduce marine pollution of all kinds, in particular from land-based activities, including marine debris and nutrient pollution, and Target 14.3 aims to reduce ocean acidification). To account for the use of energy (not only the source), utilizing the ELSS conceptual model, ETVs that link energy usage to other flows in the ELSS can be identified.

Climate change poses a fundamental challenge to sustainability and SDG 13 acknowledges the urgency of addressing the challenge. Steffen et al. (2018) point out that there might be a threshold in the climate system and crossing this threshold would lead to serious disruptions to ecosystems, society, and economies. They conclude that

collective human action is required to steer the Earth System away from a potential threshold and stabilize it in a habitable interglacial-like state. Such action entails stewardship of the entire Earth System — biosphere, climate, and societies — and could include decarbonization of the global economy, enhancement of biosphere carbon sinks, behavioral changes, technological innovations, new governance arrangements, and transformed social values.

However, none of the SDG 13 Targets aims at the transformations in the economic processes linking society to the ELSS and in the anthroposphere enabling the actions identified by Steffen et al. (2018). Based on the conceptual ELSS model, ETVs that link the flows in the ELSS and between ELSS and humanity to climate change drivers can be identified.

The metrics defined by the current SDG Indicators provide information on progress towards the targets mainly for socio-economic variables but not necessarily the environmental sustainability (where ‘environmental sustainability means that the climate system is stable, biodiversity is conserved, ecosystems function well, freshwater is secured, rural and urban settlements are protected from pollution and are resilient to climate shocks’, see TWI2050 - The World in 2050 2018). Assessing the link between the SDGs Targets and ELSS sustainability, the GBA can be used for a review of the targets. A few examples illustrate this application of the GBA.

For SDG 3 ‘Good Health and Wellbeing’, Target 3.9 aims at ‘by 2030, substantially reduce the number of deaths and illnesses from hazardous chemicals and air, water and soil pollution and contamination’. However, the indicators focus only on consequences of environmental conditions: 3.9.1 ‘Mortality rate attributed to household and ambient air pollution’, (Tier I) and 3.9.2 ‘Mortality rate attributed to unsafe water, unsafe sanitation and lack of hygiene (exposure to unsafe WASH services)’ (Tier I). This focus on the consequences of air pollution does not account for a time lag between changes in environmental conditions and changes in mortality rates. Mortality is not a low-latency indicator because there is an accumulative effect that generates a large time delay between changes in air quality and changes in mortality. A decrease in mortality can happen even if pollution is increasing in the short term. The SDG Indicators for Target 3.9 do not account for the time-lagged relationship between pollution and mortality. Adding indicators focusing directly on the environmental conditions such as air pollution would be important to account for the time lag between cause and consequences.

Another example is SDG 6 ‘Clean water and Sanitation’ and the Target 6.6 ‘by 2020, protect and restore water-related ecosystems, including mountains, forests, wetlands, rivers, aquifers and lakes’. The sole indicator for this target is Indicator 6.6.1 ‘Percentage of change in the extent of water-related ecosystems over time’ (Tier II). The extent of water-related ecosystems could be extracted from EOs to some level. However, there are many other EOs relevant to the target that could be used to monitor progress to a higher degree of detail, such as change in wetland extent, and water resources monitoring. The global monitoring community is building an integrated water monitoring framework with data sources from EOs, completing those from surveys, regulatory frameworks and other administrative sources, as well as new and novel sources, including big data. Available networks include the Satellite-based Wetland Observation Service (SWOS), Global Groundwater Monitoring Network (GGMN), and Global Terrestrial Networks (GTNs) for glaciers, rivers, lakes and other water bodies.

For SDG 14 ‘Life Below Water’, Target 14.2 aims at ‘by 2020, sustainably manage and protect marine and coastal ecosystems to avoid significant adverse impacts, including by strengthening their resilience, and take action for their restoration in order to achieve healthy and productive oceans’. The Indicator 14.2.1 is defined as the ‘Proportion of economic zones managed using ecosystem-based approaches’ (Tier III). This indicator relies on socio-economic data. There are, however, many environmental variables that have information relevant to progress towards the target, which can be extracted from EOs. Example include water quality, pollution, algae blooms, sargassum blooms, and salinity. Moreover, EOs can contribute to better define integrated coastal management plans.

As a last example, SDG 15 ‘Life on Land’ has Target 15.2 ‘By 2020, promote the implementation of sustainable management of all types of forests, halt deforestation, restore degraded forests and substantially increase afforestation and reforestation globally’. The corresponding Indicator is 15.2.1 ‘Progress towards sustainable forest management’ (Tier I). In this case, EOs could play an important role and there is a need for continued coordination of data streams and the addition of new data streams and products to ensure multiple annual global coverages of the world's forested areas. The continued development of data services tools and pilots for data acquisition planning, data storage, data processing and delivery of tailored data packages is equally important.

These example illustrate that applying the GBA to all targets with the aim to analyze the relationship of these targets to the ELSS will result in a comprehensive set of ETVs in the ELSS group. The review also indicates a very limited connection between the current indicator framework and ELSS related ETVs. There is an urgent need to add indicators that would increase the connection of the indicator framework to these important ETVs.

7. ETVs for SDGs

Applying the GBA to the indicators can result directly in ETVs related to the framework used to monitor progress towards the SDG Targets. Several efforts have been made to link the indicators to EOs and geospatial information (e.g. Group on Earth Observations 2017). However, these efforts approach the linking based on current feasibility and thus miss those variables that are currently not extracted from EOs or not observed as traditional EOs. Reyers et al. (2017) take a broader approach and propose an integration of specific ESDGVs with the already available domain-specific sets of EVs and additional new domain-specific sets.

Here we apply the GBA to three examples illustrating the process used for the goal-based identification of ETVs (Figure 8). For SDG 1 and Target 1.4, the indicator 1.4.1 (Tier III) measures the proportion of the population living in households with access to basic services. All but one of the resulting ETVs are related to the built environment. In addition, the population per household is an ETV related to the people. For SDG 6 and Target 6.3, two indicators measure the percentage of wastewater safely treated (6.3.1, Tier II) and the percentage of bodies of water with good ambient water quality (6.3.2, Tier II). The first indicator relates to the ETVs of water usage, wastewater produced, and treated wastewater, which are also all in the infrastructure group. The second indicator relates to the one ETV of water quality, which is in the ELSS group. Finally, for SDG 12 and Target 12.5, the indicator 12.5.1 (Tier III) focuses on recycling rates, and associated ETVs are the amount of waste and the amount of recycled waste, which are variables in the economy group.

Figure 8. Link between SDGs and ETVs. The examples of SDG 1, SDG 6, and SDG 12 with the associated Targets 1.4, 6.3, and 12.5, respectively, illustrate the use of the GBA for identifying ETVs needed for the quantification of the corresponding indicators.

Carrying out this analysis for all groups of goals, targets and indicators reveals that most of the ETVs associated with the existing indicators are in the infrastructure and built environment group, and the second largest group is that of people-related ETVs. Only a small number of ETVs relate directly to the ELSS. The SDG Indicators are found to be focused strongly on human needs and biased toward social and economic information and the built environment. As a consequence, only very few indicators can currently be quantified based on information extracted from EOs (Plag et al. 2016). A comparison of the SDG Indicators to the European Environment Agency (EEA) indicators shows that the latter are far more focused on environmental characteristics (see EEA 2014, and for a general description, www.eea.europa.eu/data-and-maps).

Traditional EO techniques are designed for, and focusing on, observations of the ELSS or the non-human environment, while only a few variables related to the built environment and associated infrastructure are extracted from these EOs. The results of the GBA applied to the current SDG Indicators underline the need to focus EO on the built environment. New methods to extract information on the built environment from traditional EOs need to be developed and new approach and techniques to collect comprehensive information on the built environment and the embedded social fabric are needed. Geospatial data is of high relevance for the monitoring of progress towards the SDG targets (United Nations 2018). In many cases, the geospatial information could be derived from an integration of statistical, economic, and environmental data obtained with traditional EOs and novel approaches such as citizen scientists, crowd-sourcing, big data analysis and the emerging Internet of Things (IoT).

The GBA can also be used to aggregate the currently very large set of SDG Indicator. Sridhar (2016) discusses approaches how to making the SDGs useful by combining indicators into more complex indexes and concludes that at an integrated level, the metrics is better adjusted to communicate the merits of the SDGs to the public. Rosenstock et al. (2017) point out that the current indicator framework for monitoring progress is based on a ‘more is better’ philosophy and argue that a less burdensome and decision-relevant systems are needed. The ETVs identified using the GBA could provide a basis to develop decision-relevant monitoring systems.

A number of scientific communities have raised concerns about the environment not sufficiently represented in both the SDGs and the indicators proposed to monitor progress towards the targets. For example, Griggs et al. (2013) put a strong emphasis on safeguarding the ELSS, on which present and future generations depend. Lu et al. (2015) point out that a review of the proposed SDGs conducted by ICSU showed that addressing climate change, food and water security, and health requires coordinated global monitoring and modeling of many factors. The current indicator framework has not sufficiently integrated the environmental aspects of sustainable development.

8. Conclusions

The GBA to the identification of ETVs is complimentary to the EBA that is widely used in scientific and other communities engaged in EOs. The starting point for the GBA is a set of agreed-upon societal goals that can be associated with specific targets to be achieved in a given time interval, while progress towards the targets is measured based on metrics provided by indicators. The GBA is designed to identify those ETVs required for the development of transformation knowledge as well as the monitoring of progress towards the goals. Importantly, the GBA utilizes a conceptual model underlying and underpinning the goal set for an assessment of the goals, targets and indicators and uses this model in the processes of identifying the ETVs required both for the creation of the transformation knowledge and for the quantification of the indicators in such a way that they provide a report card for monitoring of progress.

There are a number of different societal goal sets that could benefit from the availability of EOs of the relevant ETVs. Although a few examples have been indicated, there are many more that could be introduced. For example, in the frame of risk reduction and governance, societal deliberations are focusing on societal goals that would significantly benefit from EOs of relevant ETVs.

A goal set of particular interest is the set of the seventeen SDGs. The conceptual model for a GBA to the identification of ETVs for the implementation of the SDGs is the ELSS with humanity embedded in the ELSS and interacting with it through flows of energy and matter. In modern society, these flows are almost exclusively controlled by economic rules, emphasizing the importance of the mainstream economic model for progress towards the SDGs. Applying the GBA to the SDGs at the goal level shows that very few of the SDGs are directly focusing on the ELSS and the flows between the ELSS and humanity.

At the target level, the GBA reveals that many of the SDG Targets are focusing on transformations in society and the built environment while only very few are related to the flows between the ELSS and humanity. Considering that these flows over the last two centuries have been increased by several orders of magnitude, having targets that explicitly focus on decelerating these flows would be important to ensure planetary health of the ELSS.

Applying the GBA to the current indicator framework shows that most of the indicators measure aspects of the infrastructure and built environment, while the second largest group of indicators relates to population related-variables. Only a small number of ETVs corresponding to the existing indicators are variables of the non-human components of the ELSS. The current indicator framework is biased toward social and economic information and the built environment. Because impacts of environmental conditions on human population often result from cumulative effects and are time-delayed, some of the indicators do not provide timely information on progress towards targets and goals, and introducing indicators that directly measure changes in environmental conditions relevant to human welfare would improve the support for validation of transformation policies.

Only a few of the ETVs corresponding to the current indicators can currently be quantified based on information extracted from traditional EOs. However, research on how information on the built environment could be extracted from EOs could significantly increase the contribution of traditional EOs to the quantification of the ETVs. Based on the notion that sustainable development requires the functioning ELSS, it would be of value to develop a set of complementary indicators that bring environmental aspects to the monitoring of SDG targets. The underlying concept for this could be the ‘safe operating space for humanity’.

Acronyms

ABM	=	Agent-Based Model
ELSS	=	Earth's life support system
EO	=	Earth Observation
EV	=	Essential Variable
EBA	=	expert-based approach
EBV	=	Essential Biodiversity Variable
ECV	=	Essential Climate Variable
EEA	=	European Environment Agency
EOV	=	Essential Ocean Variable
ESDGV	=	Essential SDG Variable
ETV	=	Essential Transformation Variable
GBA	=	goal-based approach
GCOS	=	Global Climate Observing System
GGMN	=	Global Groundwater Monitoring Network
GTN	=	Global Terrestrial Network
IGOS	=	Integrated Global Observing Strategy
IGOS-P	=	Integrated Global Observing Strategy Partnership
IAEG-SDGs	=	Inter-Agency and Expert Group on SDG Indicators
IoT	=	Internet of Things
MDG	=	Millennium Development Goal
OR	=	Observational Requirement
SDG	=	Sustainable Development Goal
SFM	=	Stock-and-Flow Model
SOSH	=	Safe Operating Space for Humanity
SWOS	=	Satellite-based Wetland Observation Service
UNFCCC	=	United Nations Framework Convention on Climate Change
UNSC	=	United Nations Statistical Commission

Acknowledgments

The authors would like to acknowledge the European Union ‘Horizon 2020 Program’ that funded the ConnectinGEO (Grant Agreement no. 641538) projects. Part of the work for one author (Plag) was conducted under NASA grant 80NSSC17K0241.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Hans-Peter Plag http://orcid.org/0000-0003-0909-4501

Shelley-Ann Jules-Plag http://orcid.org/0000-0003-4342-2126

Additional information

Funding

This work was supported by Horizon 2020 Framework Programme [ 641538] and National Aeronautics and Space Administration [ 80NSSC17K0241].