A multi-level conceptual system dynamics architecture for examining sustainability and resilience in coupled earth–space systems

1 Introduction: foundations for earth–space system modeling

Human societies are entering a period of converging pressures and emerging capabilities. Environmental degradation, climate change, urbanization, and resource constraints (Bardi and Alvarez Pereira, 2022; Meadows et al., 1972; Meadows, 2008), together with global risks including technological threats, nuclear conflict, and asteroid impacts, are increasingly exposing the vulnerability of a civilization concentrated on a single planet (Bostrom, 2013; Moynihan, 2021; Ord, 2020; Steffen et al., 2005). Concurrently, advances in launch systems, robotics, and space infrastructure are expanding the feasibility of large-scale human activity beyond Earth, raising questions about how such developments may affect long-term Earth–space system dynamics (Marshall, 2023).

Earth is often conceptualized as a bounded planetary system, with concepts such as “Spaceship Earth” emphasizing its finite, interconnected, and materially constrained nature (Boulding, 1966; Rockström et al., 2009). This framing remains valuable for understanding planetary limits, but it may also reinforce the assumption that human systems must be analyzed only within a closed terrestrial domain. From a systems perspective, however, boundaries are analytical constructs rather than fixed entities. They may shift as technologies, interaction patterns, and forms of environmental modification change, including the capacity of organisms, human societies, and technological systems to extend beyond their immediate environments (Meadows, 2008; Odling-Smee et al., 2003; Trenberth et al., 2009; White, 1987; 2019). Despite this boundary-shifting potential, scholarly exploration of large-scale space migration (LSSM), here defined as a system-level process involving the redistribution of populations, resources, and infrastructure across terrestrial and off-world environments within the solar system, remains limited.

System dynamics modeling, originally developed by Forrester (1961), provides a useful methodology for examining such multi-level processes because it can represent feedback loops, delays, accumulations, and nonlinear behavior over time (Meadows, 2008; White, 2019). This approach was influentially applied in The Limits to Growth project (Meadows et al., 1972), which used the World3 model to examine interactions among population, resources, industrial activity, and pollution over extended time horizons. World3 operated primarily within terrestrial parameters, reflecting the premise of Earth as a bounded planetary system, and suggested that exponential growth within finite material constraints can produce overshoot and decline. Although controversial at the time, World3 remains an important reference point for modeling long-term feedbacks among population, resource consumption, industrial activity, and environmental pressure (Bardi and Alvarez Pereira, 2022; Costanza et al., 2007; Meadows et al., 2004; Meadows, 2008; Turner, 2008).

The Limits to Growth also stimulated early research into space settlement, notably through Gerard K. O’Neill’s proposals for off-world civilizations supported by extraterrestrial resources and solar energy (Johnson and Holbrow, 1977; O’Neill, 1977). More recent work includes the O’Neill–Glaser modeling framework, which examined interactions among energy production, infrastructure investment, labor allocation, and population growth within a techno-economic Earth–space context (Curreri and Detweiler, 2011), as well as interplanetary supply-chain and logistics models (Gralla et al., 2006; Ho, 2024), and studies of extraterrestrial resource utilization, off-world industrial activity, and equity-focused asteroid mining scenarios (Hein et al., 2020; Lewis, 1997; Zi and Cai, 2023).

Despite these advances, established modeling traditions remain only partially equipped to analyze large-scale space migration as a coupled Earth–space socioecological system. World3, integrated assessment models, O’Neill–Glaser-type models, space logistics frameworks, and spaceflight human factors research each address relevant components, but rarely integrate material, institutional, and psychosocial feedbacks within a single system dynamics architecture. The framework proposed here addresses this gap by bringing these domains together in a multi-level architecture designed to examine both adaptive and maladaptive Earth–space trajectories, including extractive, inequitable, or destabilizing forms of development. Table 1 summarises the conceptual gap addressed by the proposed framework in relation to selected modelling traditions.

TABLE 1. Conceptual gap addressed by the proposed framework in relation to selected modeling traditions.

2 A multi-level conceptual system dynamics architecture for large-scale space migration

2.1 Framework architecture and model status

Building on this gap, this paper proposes an exploratory system dynamics architecture for examining large-scale space migration as a coupled Earth–space socioecological system. The proposed architecture extends Earth–space modeling beyond predominantly techno-economic approaches by incorporating governance and psychological/social processes as interacting components of long-term system behavior, rather than treating them as external contextual factors.

The framework represents Earth and off-world human systems as interconnected rather than separate domains. It is organized across three analytical levels (Figure 1). Level I represents material and infrastructure dynamics, including population dynamics, resource systems, environmental processes, economic activity, space infrastructure, and Earth–space flows. Level II represents governance and institutional dynamics, including institutional structures and trade-offs among liberty, equality, and security. Level III represents psychological and social dynamics, including social adaptation and collective response through aggregate variables such as perceived system security, institutional legitimacy, and social cohesion.

These levels are linked through recursive feedbacks and shaped by cross-cutting scenario conditions, including complexity, constraint, and Earth–space interdependence. At this stage, the framework should be understood as a conceptual and methodological architecture for staged model development, rather than as a fully operational multi-level simulation model.

In this paper, migration refers to the movement of people, resources, and infrastructure between Earth and off-world environments. Settlement refers to the establishment of relatively stable communities beyond Earth. Habitation refers to the capacity to sustain human life within constructed or adapted environments. Colonization is avoided as a primary framing term because of its historical and political associations.

2.2 Level I: material and infrastructure dynamics

At present, model development has focused on the Level I material and infrastructure layer. An initial LSSM prototype has been developed by the Human Space Program in collaboration with Ventana Systems using Vensim system dynamics software (Human Space Program, 2024). The Level II governance and Level III psychological/social layers remain proposed extensions that require further operationalization, parameterization, and validation. Vensim was selected because it is a well-established system dynamics platform designed to represent stocks, flows, feedback loops, delays, nonlinear relationships, and long-term scenario analysis. It also allows model assumptions to be adjusted iteratively, making it suitable for exploring alternative Earth–space development scenarios.

Adapted from the World3 framework presented in The Limits to Growth (Meadows et al., 1972), the prototype extends the original terrestrial model to examine long-term dynamics across interconnected terrestrial and off-world environments. Scenario domains may include orbital and free-space habitats, Lagrange-point communities, lunar and Martian surface settlements, ISRU sites, and asteroid-resource systems. The prototype can also explore alternative settlement-capacity scenarios, including off-world population scales such as 1 million, 10 million, or 1 billion people, subject to assumptions about infrastructure capacity, resource availability, energy systems, transport costs, ISRU development, technological capability, and Earth–space interdependence.

Within the Level I prototype, core World3 subsystems are reconfigured within an Earth–space architecture. The material and infrastructure layer is organized around population dynamics, resource systems, environmental pressures, industrial activity, infrastructure development, transport capacity, and Earth–space flows. Population dynamics include population growth, migration between Earth and off-world communities, and changing settlement capacities. Resource dynamics include terrestrial resources and potential extraterrestrial materials from asteroid, lunar, and planetary sources. Environmental dynamics may include terrestrial environmental pressures, orbital debris, space-based contamination, and climate-energy interactions. Industrial and infrastructure dynamics include production, infrastructure development, energy use, transport capacity, ISRU development, and trade across Earth–space environments.

The prototype also allows Earth-system dynamics to be considered in relation to climate processes, the carbon cycle, greenhouse gases, solar radiation, population trajectories, energy technologies, and economic growth. In long-term high-energy scenarios, human energy use may also contribute directly to waste heat, including from energy sources that do not emit carbon. This makes off-world energy production, industrial activity, and resource use relevant scenario variables for examining whether Earth–space development reduces, displaces, or intensifies environmental pressures over the long term.

The Level I prototype also includes an economic subsystem relevant to space economy analysis. This subsystem captures how launch costs, transportation efficiency, infrastructure investment, industrial capacity, energy production, ISRU capacity, resource extraction viability, off-world manufacturing, trade, and value-added activity may shape Earth–space system development. These processes can be represented through stocks such as capital, industrial capacity, energy production capacity, and infrastructure capacity, and through flows such as investment, production, resource exports, technology transfer, trade, and transport-cost reduction. In future model iterations, this economic layer may be used to examine how launch economics, resource availability, industrial development, and trade relationships influence settlement capacity, systemic risk, inequality, resilience, and long-term Earth–space system stability.

Figure 2 summarizes the causal-loop structure of the Level I prototype, distinguishing dominant material, infrastructure, resource, demographic, and economic feedbacks from proposed Level II and Level III extension interfaces.

Figure 3 translates this causal logic into a high-level stock-flow architecture, identifying principal stocks, flows, and rate processes related to population, resources, infrastructure, economic activity, ISRU, transport, trade, and value-added capacity.

2.3 Level II: governance and institutional dynamics

Large-scale human activity beyond Earth raises fundamental governance questions. Existing international space law frameworks were developed when space activity was largely limited to scientific missions and satellite operations, and they provide limited guidance for large-scale permanent settlements, commercial resource extraction, or increasingly interconnected Earth–space systems. In the absence of stronger governance arrangements, future space activities could become a source of competition and conflict. Deudney (2020) argues that large-scale space expansionism may generate geopolitical and existential risks and should therefore be approached with restraint.

Within the proposed framework, Level II represents governance and institutional dynamics. This level captures how institutions may shape movement, coordination, regulation, resource allocation, and the management of Earth–space interdependence. It is defined by institutional arrangements and trade-offs among liberty, equality, and security (Rawls, 1971; Woods, 2025). Liberty refers to individual freedom, autonomy, rights, and mobility; equality refers to fairness, distributive justice, opportunity, and access to resources; and security refers to stability, protection, system reliability, and collective survival.

At this stage, these Level II governance dimensions are proposed as conceptual and scenario-based variables rather than fully operationalized parameters. In future model development, they may first be represented using ordinal scenario states, such as low, medium, and high, before being mapped onto measurable proxy variables. For example, liberty could be approximated through a regulatory burden index, mobility restrictions, or the proportion of decisions made at local rather than central governance levels; equality through a Gini coefficient, per capita resource allocation, or access rates to critical infrastructure; and security through an infrastructure redundancy index, life-support failure rates, conflict incidence, or emergency response capacity (Bruneau et al., 2003; OECD, 2026; World Bank, 2026). Trade-offs among these dimensions could later be represented through weighted indices, constraint functions, thresholds, or scenario-specific parameter settings, allowing changes in one dimension, such as security, to be tested against possible effects on liberty, equality, institutional legitimacy, and social cohesion.

International relations theory can inform how Level II institutions coordinate under conditions of complexity, constraint, and interdependence (Deudney, 2020; Miller, 2023). Within the model, this supports scenario comparisons among cooperative, competitive, and hybrid governance arrangements, ranging from shared resource management and collective risk reduction to decentralized autonomy and mixed systems combining local autonomy with coordinated resource governance, including issues relevant to commons governance and collective resource management (Ostrom, 1990).

Initially, off-world communities are likely to remain highly dependent on Earth-based systems for infrastructure, technology, governance, and resource support (Cockell, 2022). As local capabilities and populations grow, Level II governance choices may influence the degree of Earth–space interdependence and the distribution of access to infrastructure, resources, and productive capacity. Unequal access may create risks of governance asymmetry, structural inequality, and reduced institutional legitimacy.

These dynamics may be magnified by increasing technological interdependence. Earth-based societies already face systemic risks arising from highly interconnected and tightly coupled technological and cyber infrastructures (Brundage et al., 2018; Helbing, 2013; Perrow, 1984). Similar dynamics may emerge across linked Earth–space systems as critical infrastructure, communication networks, resource flows, and governance functions become increasingly interconnected. Stronger forms of planetary and interplanetary governance may therefore be needed to manage coordination demands, systemic risk, resource use, and institutional accountability (Crawford, 2015; 2022; Leinen and Bummel, 2018; López-Claros et al., 2020).

2.4 Level III: psychological and social dynamics

The long-term viability of large-scale human activity beyond Earth is likely to depend not only on technological feasibility, economic capacity, and governance arrangements but also on the capacity of individuals and groups to adapt to constrained and increasingly complex environments. Research in spaceflight human factors and extreme environments suggests that long-duration missions involve physiological, psychological, and environmental demands, with implications for adaptation, coordination, trust, and group functioning (Afshinnekoo et al., 2020; Anna et al., 2026; Kanas and Manzey, 2008; Leach, 2016; Patel et al., 2020).

As human activity beyond Earth increases in scale and complexity, psychological and social processes may become important in shaping collective adaptation, institutional functioning, and the stability of emerging Earth–space systems. Perceived system security, institutional legitimacy, and social cohesion are selected as Level III psychosocial variables because they operate at the interface between individual perception, group coordination, and system-level behavior. They are also closely linked to the Level II governance layer, as trade-offs among liberty, equality, and security are likely to shape how individuals and groups perceive safety, fairness, autonomy, legitimacy, and collective viability. These variables are not intended to capture the full complexity of human psychology, but are represented as aggregate system-level proxies for social adaptation and collective response. At this early stage of development, they may be represented using ordinal scenario states, such as low, medium, and high, and later mapped onto numerical weights, thresholds, or modifiers within recursive feedback processes linking the different layers of the model.

This selection is informed by space psychology and isolated, confined, and extreme-environment research, which identify stress, cohesion, leadership, interpersonal functioning, morale, trust, and adaptation as important factors in long-duration missions and constrained environments (Kanas and Manzey, 2008; Leach, 2016; Patel et al., 2020). It also draws on Social Identity Theory, which links group identification to cooperation and collective behavior (Tajfel and Turner, 1979); Self-Determination Theory, which emphasizes autonomy, competence, and relatedness as conditions for motivation and adaptation (Deci and Ryan, 2000); Conservation of Resources Theory, which frames stress and coping in relation to perceived resource threat, loss, and protection (Hobfoll, 1989); and legitimacy theory, which links perceived fairness and institutional legitimacy to compliance and cooperation (Tyler, 2006).

3 Cross-level integration and emergent outcomes

Although the framework is presented across three levels for analytical clarity, the relationships among them are recursive rather than unidirectional. Level I material and infrastructure conditions, including migration patterns, resource availability, environmental pressures, infrastructure performance, economic activity, and Earth–space interdependence, may shape Level II governance requirements and Level III psychological and social responses. Level II governance arrangements may influence resource allocation, coordination, institutional legitimacy, and perceptions of fairness and safety. Level III psychosocial variables, including perceived system security, institutional legitimacy, and social cohesion, may then affect cooperation, compliance, collective behavior, and support for alternative governance arrangements. These responses may feed back into Level I resource use, migration decisions, and infrastructure coordination, as well as Level II governance effectiveness and institutional adaptation. For example, cross-level coupling may show how inequality or perceived injustice weakens trust, legitimacy, and cooperation, while stronger legitimacy and social cohesion may support resilience, adaptive capacity, and system stability under conditions of constraint.

Complexity, constraint, and Earth–space interdependence are treated as cross-cutting scenario conditions that can be varied using ordinal states, such as low, medium, and high. These states are not precise empirical measurements. Rather, they provide structured modeling assumptions for examining how different operating conditions may shape relationships among the different levels. In future model development, these ordinal states would require translation into model effects through explicit assumptions, thresholds, weights, or sensitivity analyses.

Outcome dimensions such as sustainability, resilience, system stability, adaptive capacity, human wellbeing, and institutional resilience are conceptualized as possible emergent properties rather than predefined goals or fully specified model outputs at this stage. These dimensions may increase or decrease, stabilize or destabilize, and converge or diverge across alternative Earth–space trajectories. They therefore provide a basis for interpreting how interactions among Level I material and infrastructure conditions, Level II governance arrangements, and Level III psychological and social responses may shape broader system behavior over time. In this sense, they serve as orienting constructs for scenario analysis, hypothesis generation, and future model development.

4 Limitations, validation pathways, and future directions

Several limitations remain. LSSM is a hypothetical systems scenario rather than an empirically observable process, and the framework necessarily depends on uncertain assumptions about future technologies, launch economics, life-support closure, radiation mitigation, ISRU maturity, governance evolution, and off-world settlement capacity. These assumptions create substantial parameter uncertainty, particularly in relation to transport costs, infrastructure requirements, resource availability, technological learning rates, institutional development, and long-term population trajectories. The Level II governance and Level III psychological/social layers also remain conceptual extensions at this stage and require further operationalization, parameterization, and validation before integration into a fully specified simulation model. Long-term Earth–space system outcomes cannot be empirically verified in advance. Future model development should therefore be understood as supporting scenario exploration, hypothesis generation, and sensitivity testing rather than prediction.

Future validation would require a combination of structural, behavioral, and stress-test approaches. Structural validation would compare model boundaries, stocks, flows, and feedback assumptions with World3 (Meadows et al., 1972), National Aeronautics and Space Administration architecture studies (NASA, 2023), and relevant space logistics models (Ho, 2024). Behavioral calibration could draw on isolated, confined, and extreme-environment analogs, including Antarctica, the Hawai’i Space Exploration Analog and Simulation (HI-SEAS, 2026), Mars500 (European Space Agency, 2026), and NASA Extreme Environment Mission Operations (NASA, 2015). Historical stress testing could use the coronavirus pandemic, the Fukushima Daiichi nuclear accident, and supply-chain crises to examine infrastructure dependency, resource constraint, institutional response, and social coordination (Sasakawa Peace Foundation, 2012; Xu et al., 2020). These comparisons would not validate LSSM directly, but would test whether component relationships and feedback assumptions behave plausibly under conditions of complexity, constraint, and interdependence. Future development may also benefit from engagement with complex adaptive systems theory, particularly in relation to emergence, adaptation, nonlinear feedbacks, and multi-level interactions within socioecological systems (Levin et al., 2013).

Future work should proceed cautiously through staged and modular expansion, beginning with further refinement of the Level I prototype before incorporating selected Level II and Level III variables. Once operationalized, the model could support scenario comparison of how governance and economic arrangements shape Earth–space trajectories, including resource extraction, infrastructure development, trade, inequality, cooperation, resilience, and system stability. It should not be understood as predicting future civilizations or social forms. Instead, alternative pathways can be examined through scenario design, sensitivity testing, and model structures representing different assumptions about technological change, governance evolution, resource availability, off-world autonomy, and Earth–space interdependence. Future analyses may also examine whether changes in complexity, constraint, and Earth–space interdependence are associated with threshold effects, path dependence, or transitions between alternative system states.

The framework is relevant to sustainability science, resilience studies, systems engineering, governance, and social adaptation in extreme environments. Its relevance extends beyond large-scale settlement scenarios, as developments in cislunar space, low Earth orbit, and in-space economic activity already present governance, coordination, infrastructure, and sustainability challenges (NASA, 2023; National Space Society, 2018; United Nations, 2015). Modeling LSSM therefore provides a structured way to examine how linked Earth–space systems may adapt, fail to adapt, or reproduce unsustainable growth dynamics under increasing complexity, constraint, and interdependence, while identifying feedbacks that could shape future resilience or instability.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The author(s) verify and take full responsibility for the use of generative AI in the preparation of this manuscript. Generative AI was used to support language editing, grammatical refinement, and structural organization. All final editorial decisions were reviewed and approved by the authors.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: M. Reza Emami, University of Toronto, Canada

Reviewed by: Subhajit Hazra, LIFE-To and Beyond, India

Milad Zamanifar, Bauhaus-Universität Weimar, Germany