Key Points
- Technological Maturation: The translation of digital twin technology from industrial environments to living systems relies on the convergence of advanced sensor networks, exascale computing, and real-time model-data fusion.
- Beyond Simulation: Unlike static models, ecological digital twins operate via continuous feedback loops, adjusting predictive parameters based on dynamic observational data streams to create actionable "what-if" scenarios.
- Philosophical Tensions: The creation of planetary digital replicas surfaces an epistemological conflict between reductionist, physics-based modeling and the relational, unpredictable nature of holistic ecosystems.
- Societal and Ethical Risks: Unchecked development of environmental digital twins risks engendering an "Ecological Panopticon," where governance by numbers overshadows democratic accountability, indigenous data sovereignty, and localized ecological knowledge.
- Strategic Imperative: For design and strategic leaders, the ecological digital twin is not merely a monitoring tool but a socio-technical infrastructure that will redefine spatial planning, conservation policy, and human-nature interactions.
Context and Urgency
The global imperative to address the intersecting crises of biodiversity loss, climate change, and resource depletion has accelerated the demand for high-fidelity environmental monitoring. Traditional ecological models, characterized by static datasets and historical batch-processing, are increasingly inadequate for navigating the non-linear dynamics of modern ecological collapse. In response, transnational institutions and scientific consortiums are architecting dynamic, predictive computational replicas of natural systems.
Scope of the Analysis
This report synthesizes emerging signals across systems ecology, computer science, and science and technology studies (STS). It establishes the definitive framework of the ecological digital twin, catalogs cutting-edge pilot projects spanning marine, terrestrial, and biological domains, and rigorously interrogates the technological limitations and ethical frontiers defining this space. The analysis is structured to guide strategic leaders through the operational, philosophical, and ethical dimensions of engineering a predictive planetary mirror.
[1] Introduction: The Emergence of Ecological Digital Twins [source]
The concept of the digital twin (DT)—a high-fidelity, continuously updated virtual representation of a physical asset—has served as a foundational technology in aerospace, manufacturing, and urban infrastructure for decades 63. By ingesting real-time data from Internet of Things (IoT) sensors, industrial digital twins enable operators to monitor performance, predict mechanical failures, and optimize systemic efficiency within bounded, predictable environments. However, the transposition of this technology from engineered artifacts to complex, living biological systems represents one of the most critical and under-discussed frontiers in contemporary science and technology 1.
The transition from a manufactured asset to an Ecological Digital Twin (EDT) requires a fundamental leap in computational architecture and epistemological design. Ecosystems are not machines; they are characterized by non-linear dynamics, cascading interdependencies, and unpredictable biological behaviors 56. Designing a "predictive mirror" of the planet involves synthesizing vast, heterogeneous data streams—from satellite Earth observations and autonomous underwater vehicles to localized soil moisture sensors and environmental DNA (eDNA)—into dynamic models capable of simulating the complex metabolism of the Earth 39.
The urgency driving this technological leap is palpable. Driven by international policy mandates such as the European Green Deal and the UN Decade of Ocean Science, heavily funded consortiums are currently building localized and planetary-scale EDTs to simulate everything from single soil horizons 80 to the entire global climate system 10. Yet, as these nascent infrastructures come online, they instigate profound questions regarding planetary agency, the ethics of quantification, and the sociological implications of rendering nature as a manageable, computational object 26.
[2] Defining the Ecological Digital Twin [source]
[2] 1 Moving Beyond Mere Simulation [source]
To understand the EDT, it is necessary to rigorously distinguish it from traditional ecological modeling and simulation. Traditional ecological models are often static, relying on historical, batch-updated data to analyze past events or project generalized future trends 1. They suffer from a persistent "time lag" between the physical event, data collection, and model calibration 1.
Conversely, an Ecological Digital Twin is an actively evolving computational object that mirrors a living system in real time 2. It is defined by the following core characteristics:
- Continuous Data-Model Fusion: EDTs ingest real-time or near-real-time data streams, continually updating their internal state and refining their predictive algorithms based on observed reality 77.
- Bidirectional Feedback Loops: An advanced EDT does not merely observe; it informs direct interventions (actuators or policy controls) in the physical world, the results of which are subsequently fed back into the twin to measure efficacy 31.
- Inference of Unobservable Variables: Through complex solver chains, EDTs can infer high-order variables that are difficult or impossible to measure directly (e.g., a specific animal's physiological stress index or a forest's subterranean fungal carbon exchange) based on secondary sensor data 2.
[2] 2 The TwinEco Framework: A Structural Paradigm [source]
Because the adaptation of digital twins to ecology is a burgeoning field, it initially suffered from fragmented development and inconsistent terminologies. To rectify this, researchers have proposed unified frameworks, the most prominent being the TwinEco Framework 1.
The TwinEco framework emphasizes modularity and flexibility, acknowledging that the extreme diversity of ecological applications prevents a "one-size-fits-all" architectural approach 64. It is structured around specific conceptual layers 77:
| Layer / Component | Description within the TwinEco Framework |
| Physical State (S) | The real-world ecosystem, characterized by highly complex, dynamic variables, ranging from weather patterns to localized species interactions. |
| Digital State (D) | The computational replica. It mirrors the Physical State but provides researchers with reproducibility, scalability, and the ability to run "what-if" scenarios without risking the physical environment. |
| Observational Data (O) | The vital link connecting the physical and digital. Comprises heterogeneous inputs like remote sensing, acoustic monitors, and field surveys. |
| Control Inputs (U) | Mechanisms to influence the Physical State based on the Digital State's output. In ecology, these are often policy-based (e.g., dynamic zoning, altering hunting quotas) rather than mechanical actuators. |
| Rewards (R) | Evaluative metrics, such as biodiversity indices or carbon sequestration targets, that guide the system toward optimal ecological health. |
This modularity allows TwinEco to serve as a foundational paradigm, enabling different digital twins to share components, establish common vocabularies, and ultimately federate into larger "systems-of-systems" 3.
[2] 3 Key Technological Enablers [source]
The feasibility of the EDT is entirely dependent on recent, exponential advancements in several intersecting technological domains:
- Advanced Sensor Networks & IoT: The proliferation of low-cost, high-durability sensors (e.g., benthic observatories, acoustic biodiversity sensors, Lidar) allows for the continuous telemetry of previously unobservable natural phenomena 29.
- Dynamic Data-Driven Application Systems (DDDAS): This programming paradigm prioritizes the integration of real-time data streams within computational models, facilitating adaptive decision-making and allowing simulations to interact in real-time with evolving physical conditions 1.
- Exascale Computing: The sheer volume of data and the complexity of hybrid statistical/process-based models require massive computational power. Systems like the LUMI Supercomputer (capable of processing vast datasets across millions of cores) are critical for rendering high-resolution, global-scale simulations 49.
[3] Early Pilot Projects and Implementations [source]
The theoretical frameworks of EDTs are currently being stress-tested through a variety of high-profile, heavily funded pilot projects. These initiatives span marine, terrestrial, and physiological domains.
[3] 1 Destination Earth (DestinE) and the Digital Twin Engine [source]
The most ambitious digital twin project currently underway is the European Commission's Destination Earth (DestinE) 11. Launched as a flagship initiative under the EU Green Deal, DestinE aims to create a highly accurate, kilometer-scale digital replica of the entire Earth system to model natural phenomena and human activities 11.
DestinE operates as an ecosystem of interoperable twins, anchored by the Digital Twin Engine—a complex software infrastructure developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) 12. The engine manages extreme-scale simulations, data fusion, and machine learning across European supercomputers 12.
Currently, DestinE is deploying two high-priority twins:
- Weather-Induced Extremes Digital Twin: Designed to simulate extreme meteorological events a few days in advance with unprecedented local granularity, enabling rapid response from decision-makers 10.
- Climate Change Adaptation Digital Twin: The first-ever attempt to produce operational, multi-decadal climate projections, providing sector-specific information (e.g., hydrology, renewable energy) to guide long-term infrastructure and policy planning 10.
[3] 2 BioTwinRs and the DigitalSoma Framework [source]
While DestinE models the macro-climate, organizations like BioTwinRs are pioneering biological digital twins at the physiological and micro-ecological scale 71. Their flagship framework, DigitalSoma, is the first domain-agnostic digital twin substrate designed explicitly for living organisms across all taxa (livestock, wildlife, aquatic species) 2.
DigitalSoma treats the animal body as a continuously evolving computational object 80. It operates via a directed acyclic chain of physics- and process-based solvers 80. For example, in monitoring heat stress in livestock, the framework ingests data from wearable biosensors. The data passes sequentially through cardiovascular, metabolic, thermoregulatory, and respiratory equations 2. Because each solver consumes the outputs of the preceding one, the system can autonomously infer complex, high-order variables like physiological stress indices without redundant computation, allowing practitioners to project how an animal's state will evolve over the next six hours 2. BioTwinRs is also scaling this approach to agroecosystems through Digital Pedon, which twins multi-horizon soil profiles to bridge the gap between raw sensor signals and agronomically meaningful variables 2.
[3] 3 The Iliad Project and the Trondheim Fjord Water Quality Pilot [source]
In the marine sector, the Digital Twin of the Ocean (DTO) is gaining significant traction. The EU-funded Iliad Project is developing a federated system of interoperable ocean twins 9. A premier example is the Trondheim Fjord Water Quality Pilot in Norway 43.
This marine DTO integrates data from surface buoys, benthic observatories, and Autonomous Underwater Vehicles (AUVs) to monitor critical environmental threats, specifically algae blooms and microplastic pollution 79. When sensors detect early anomalies indicating an algae bloom, the system can autonomously deploy AUVs to gather higher-resolution spatial data 79.
To track microplastics, the pilot utilizes advanced particle sensors and feeds the data into OpenDrift, an advanced numerical modeling tool 7. By combining real-time ocean current data with particle characteristics, the twin simulates the transport and fate of pollutants, allowing environmental managers to forecast their long-term impact and visualize these processes using immersive 4D gaming engines like Unity and Cesium 9.
[3] 4 Terrestrial Twins: EcoTwin, BioDT, and Point Conception [source]
Terrestrial ecosystems are also undergoing extensive twinning:
- The Biodiversity Digital Twin (BioDT): Utilizing the LUMI supercomputer, this European consortium has developed an operational prototype to assess honey bee (Apis mellifera) performance 51. By bundling the established BEEHAVE model into Docker containers and running it in parallel across LUMI's architecture, BioDT integrates land cover data, weather, and nectar availability to generate massive, country-scale estimates of colony viability, providing actionable insights for agricultural policy 78.
- EcoTwin: Based in Germany, EcoTwin focuses on urban green spaces 8. It combines soil moisture sensors and AI to help city planners optimize land management and climate resilience, utilizing predictive models to forecast plant water requirements and direct targeted irrigation 8.
- Point Conception (Esri & The Nature Conservancy): In California, a digital twin of the Dangermond Preserve is being built to revitalize a unique coastal ecotone 15. The twin integrates flora, fauna, and food web data, specifically targeting the eradication of non-native species like ice plant 15.
[4] Scientific and Technical Hurdles [source]
Despite the rapid proliferation of pilot projects, the translation of dynamic biological systems into computational code faces profound scientific and technical hurdles.
[4] 1 Data Interoperability and the FAIR Principles [source]
A digital twin is only as robust as its underlying data. Historically, ecological data has been siloed, unstructured, and inconsistently formatted, rendering it largely incompatible with algorithmic ingestion 19. To function, EDTs require strict adherence to the FAIR data principles (Findable, Accessible, Interoperable, and Reusable) 19.
Building an EDT requires developing complex ontologies. For instance, the Anglian Water project in the UK successfully modeled the River Stiffkey by first creating a "River Health Ontology"—a common framework defining the categories, properties, and relationships within a river system 6. Without shared ontologies and unified architectural frameworks (like TwinEco), the ecosystem of digital twins risks devolving into isolated, incompatible simulations that cannot communicate across regional or disciplinary boundaries 64.
[4] 2 Modeling Complex, Unpredictable Systems [source]
The most significant scientific hurdle lies in the inherent nature of biology. While an aircraft engine operates according to known physics and bounded engineering tolerances, ecosystems are "systems of systems" driven by evolutionary biology, chaotic weather patterns, and deeply entangled multispecies interactions 16.
Researchers caution against the assumption that biological processes can be perfectly mathematized. Issues such as long-tail science events, undetected species networks, and microbial behaviors present massive modeling gaps 61. Furthermore, models rely heavily on historical data to train predictive algorithms; as climate change pushes ecosystems into unprecedented, no-analog states, the historical data powering these twins may become obsolete, severely degrading their predictive accuracy.
[4] 3 Computational Demands and Infrastructure Limits [source]
The computational burden of running real-time, high-resolution models of the Earth is staggering. Processing satellite telemetry, local sensor networks, and global circulation models simultaneously requires exascale computing facilities 12. Ensuring the continuous streaming of this data, minimizing latency, and powering the server farms introduces a paradox: the very digital infrastructure built to monitor and mitigate ecological degradation carries a massive ecological cost regarding energy consumption and hardware production 41.
[5] Integrating Indigenous Knowledge Systems [source]
As EDTs scale, there is a growing recognition that highly quantified, sensor-driven models represent a predominantly Western, reductionist epistemology. Ecosystems are not just biophysical networks; they are cultural landscapes. Consequently, cutting-edge projects are beginning to explore how to integrate Indigenous Knowledge Systems (IKS) into digital twin architectures.
[5] 1 Beyond Reductionism: Holistic Environmental Data [source]
Indigenous communities have managed complex ecosystems for millennia, relying on deep, localized, and multi-generational observations of biodiversity, seasonal shifts, and species interactions 37. Traditional Western digital twins often reduce a forest to its "carbon offsetting capacity" or "timber yield" 37.
Integrating IKS allows digital twins to encompass the socio-cultural values of territories and nuanced environmental indicators that remote sensors cannot detect 37. Using Knowledge Graphs and advanced graph databases, developers can create structured systems that link qualitative, holistic indigenous observations with quantitative sensor data 38. This semantic modeling allows for a digital twin that recognizes the relational and spiritual dimensions of land, moving beyond mere resource extraction metrics.
[5] 2 Ethical Preservation and Data Sovereignty [source]
Digital twins offer a powerful medium for the preservation of indigenous cultural heritage, allowing communities to create living virtual libraries and immersive learning environments for intergenerational knowledge transfer 36. Furthermore, digital twins can simulate traditional land-use practices (like controlled wildland burning) to scientifically validate their efficacy in modern climate adaptation planning 34.
However, this integration is fraught with ethical peril. The digital extraction of IKS poses severe risks of biopiracy and the weaponization of data 35. If a digital twin perfectly maps the resources of an indigenous territory, that data could be exploited by illegal loggers, poachers, or corporate mining interests 35. Therefore, applying the CARE Principles for Indigenous Data Governance (Collective Benefit, Authority to Control, Responsibility, Ethics) is paramount 38. Indigenous communities must maintain absolute data sovereignty over how their knowledge is ingested, modeled, and accessed within any ecological digital twin 34.
[6] Philosophical and Sociological Shifts [source]
Moving beyond the technical implementations, the advent of EDTs forces a radical reevaluation of our philosophical relationship with nature. When we possess a "predictive mirror" of the planet, our understanding of planetary agency, environmental stewardship, and human intervention is fundamentally altered.
[6] 1 The "Ecological Panopticon" and Planetary Agency [source]
The proliferation of ubiquitous sensors, satellite surveillance, and algorithmic oversight introduces the concept of the "Ecological Panopticon" 42. In this scenario, the natural world is rendered entirely visible, quantifiable, and subject to continuous optimization by algorithmic managers.
This total visibility challenges traditional notions of planetary agency. Philosophers like Bruno Latour have written extensively on the "Intrusion of Gaia"—the reality that the Earth is an active, unpredictable agent reacting to human systems 40. By attempting to encapsulate Gaia within a digital twin, we risk subjugating planetary agency to human technocratic control. The ecological digital twin could shift our conservation paradigm from "protecting wilderness" (allowing nature its autonomy) to "managing ecosystems" (treating nature as a highly optimized agricultural or industrial asset).
[6] 2 STS Critiques: Governance by Numbers vs. Democratic Accountability [source]
The field of Science and Technology Studies (STS) provides a critical lens for examining the sociological implications of EDTs. Prominent scholars, such as Andrea Saltelli, have rigorously critiqued projects like Destination Earth, arguing that they subscribe to a dangerously reductionist, Newtonian view of nature as a machine 54.
Saltelli and colleagues argue that the push for planetary digital twins reflects an epistemological arrogance—the belief that infinite computational scaling can resolve the deep, structural uncertainties of climate change 26. This over-reliance on complex mathematical models shifts political power away from citizens and democratic institutions and concentrates it in the hands of the technocrats, data scientists, and corporations that wield the models 55.
When policy decisions (e.g., land zoning, carbon taxes, water rights) are justified by the impenetrable outputs of a digital twin, it engenders a "governance by numbers" 56. If the models are treated as absolute truth, it marginalizes qualitative social sciences, local lived experiences, and democratic debate 56. STS scholars advocate for bringing these models "back to earth," demanding independent auditing, the use of diverse heuristic models, and a focus on societal well-being over the blind pursuit of computational sophistication 56.
[7] Ethical Concerns and the Risk of Technocracy [source]
The strategic implementation of EDTs carries profound ethical risks that design leaders and policymakers must navigate to prevent unintended, catastrophic consequences.
[7] 1 Data Ownership, Control, and Bias [source]
The development of EDTs requires massive capital and computational infrastructure, inherently favoring global technology monopolies and highly resourced state actors. This raises severe ethical concerns regarding data ownership and control. If a private corporation builds the definitive digital twin of the Amazon rainforest, who owns the insights generated? 35 There is a risk that environmental digital twins could become "evil twins," utilized by malign actors to identify optimal areas for illegal extraction or to surveil marginalized groups defending the forest 35.
Furthermore, algorithms are not neutral. The models driving EDTs encode the biases, assumptions, and priorities of their developers. If an EDT is optimized strictly for "economic yield" or "carbon capture efficiency," it may recommend interventions that destroy local biodiversity or displace human populations, presenting these deeply biased decisions as objective, mathematical necessities.
[7] 2 The Illusion of Control and Unintended Consequences [source]
Perhaps the most significant risk is the illusion of control. The high-fidelity visualizations and confident predictive outputs of an EDT can lull policymakers into a false sense of security, convincing them that they have perfectly mastered the environment.
Ecosystems are defined by "unknown unknowns." If a digital twin fails to account for a specific microbial interaction, a policy intervention based on that twin could trigger a catastrophic, real-world ecological collapse. Treating a predictive model as an exact replica, rather than an inherently flawed approximation, invites the risk of over-reliance, where human intuition, on-the-ground observation, and precautionary principles are discarded in favor of algorithmic dictates.
[8] Opportunities for Global-Scale Environmental Management [source]
Despite the valid philosophical and ethical critiques, the potential benefits of EDTs for global-scale environmental management are unprecedented, provided they are developed as socio-ecological tools rather than purely technocratic ones.
[8] 1 Socio-Ecological Digital Twins [source]
The next evolution of the EDT is the Socio-Ecological Digital Twin, which integrates economic, social, and human behavioral layers with biophysical data 30. Recognizing that cities and ecosystems are deeply entwined, these twins model the tensions between urban development and environmental sustainability 4.
By simulating how a proposed zoning law affects local poverty rates, commuter traffic, and estuarine water quality simultaneously, policymakers can move beyond zero-sum environmental planning 4. This approach democratizes decision-making by allowing communities to run their own "what-if" scenarios, visualizing the long-term consequences of interventions and shifting the focus from short-term outputs to holistic, long-term outcomes 4.
[8] 2 Proactive Conservation and Resilience [source]
At a planetary scale, EDTs provide the necessary infrastructure to transition from reactive disaster management to proactive resilience.
- Preventing Ecosystem Die-back: In regions like the Amazon, digital twins can identify specific areas at the highest risk of "savannization" (the tipping point where a rainforest devolves into a dry grassland), enabling precise, targeted conservation efforts before the tipping point is reached 35.
- Climate Adaptation: By running multi-decadal simulations, governments can accurately design coastal defenses, optimize renewable energy grids against future weather extremes, and restructure agricultural zones to ensure food security in a rapidly warming world 10.
Ultimately, the EDT operates as an advanced form of planetary situational awareness. If governed by ethical, open-source principles and guided by an ethos of stewardship rather than exploitation, these predictive mirrors could be the crucial tool required to synchronize human civilization with the Earth's carrying capacity 39.
[9] Conclusion [source]
The Ecological Digital Twin represents far more than an incremental advancement in environmental monitoring; it is a profound socio-technical paradigm shift. By converging exascale computing, real-time sensor networks, and complex systems modeling, humanity is attempting to construct a dynamic, predictive mirror of the biosphere.
As pilot projects like Destination Earth, BioTwinRs, and the Trondheim Fjord DTO demonstrate, the technical capacity to model nature is accelerating rapidly. However, the true challenges lying ahead are not strictly computational; they are deeply philosophical and ethical. The scientific community and strategic leaders must navigate the tension between the immense utility of predictive ecological management and the existential risks of reductionism, technocratic overreach, and the illusion of total control.
To ensure that the Ecological Digital Twin serves as a tool for planetary regeneration rather than a mechanism of the "Ecological Panopticon," its development must be democratized. It must integrate qualitative social sciences and Indigenous Knowledge Systems, adhere strictly to FAIR and CARE data principles, and remain subservient to democratic accountability. Only by recognizing the inherent unpredictability of the Earth and resisting the urge to treat nature as a machine can we utilize the predictive mirror to foster a resilient, sustainable coexistence with our planet.
References
[1] Khan, T., de Koning, K., Endresen, D., Chala, D., & Kusch, E. (2024). "TwinEco: A unified framework for dynamic data-driven digital twins in ecology." bioRxiv [source]
[2] BioTwinRs. (2024). "Biological & Ecological Digital Twins." BioTwinRs [source]
[3] Khan, T. et al. (2025). "TwinEco: A Unified Framework for Dynamic Data-Driven Digital Twins in Ecology." ResearchGate [source]
[4] No Author. (2025). "Going beyond net zero: digital twins for achieving socio-ecological sustainability." Smart and Sustainable Built Environment [source]
[5] UK Research and Innovation. (2024). "Digital twin projects to transform environmental science." UKRI [source]
[6] Anglian Water. (n.d.). "Ecological Digital Twin." Water Innovation Challenges [source]
[7] Iliad Project. (2024). "Digital Twin of the Ocean: Transforming Environmental Monitoring in Trondheim Fjord." Ocean-Twin [source]
[8] Hochschule Bonn-Rhein-Sieg. (2024). "EcoTwin." H-BRS [source]
[9] Iliad Project. (2024). "A Digital Twin of the Trondheim Fjord for Environmental Monitoring." MDPI [source]
[10] Destination Earth. (2026). "Digital Twins." ECMWF [source]
[11] European Commission. (n.d.). "Destination Earth." Destination Earth [source]
[12] Destination Earth. (2026). "Digital Twin Engine." ECMWF [source]
[13] BioDT. (2025). "Biodiversity Digital Twins and the Destination Earth Ecosystem." BioDT [source]
[14] European Space Agency. (n.d.). "Europe's digital Earth gets ready to grow user community." ESA [source]
[15] Esri. (2023). "Open Science and the Environmental Digital Twin." Esri Blog [source]
[16] BioDT Consortium. (2024). "BioDT Project Overview." RIO Journal [source]
[26] Saltelli, A. (2024). "Digital Twins of the Earth: Science or Pseudoscience?" Roger Pielke Jr. Substack [source]
[29] Haarsma, D. (2026). "How geospatial intelligence, AI and ecological digital twins are shaping the future of chimpanzee conservation." GIM International [source]
[30] No Author. (2025). "Socio-ecological approach: the imbalance between urban and natural systems." Smart and Sustainable Built Environment [source]
[31] Khan, T. et al. (2024). "TwinEco framework policy control." bioRxiv [source]
[34] Adefila, E. K. (2024). "Beginning of a new era: leveraging digital twin technology in the preservation of indigenous knowledge systems." Library Hi Tech News [source]
[35] Bioverse. (2023). "Can digital twins help save the Amazon?" Mongabay [source]
[36] No Author. (2025). "Preserving Indigenous Knowledge Systems by Employing Digital Twin Technologies in the Disruptive Era." ResearchGate [source]
[37] No Author. (n.d.). "Biodiversity technology and the limits of scaling." Cambridge Repository [source]
[38] No Author. (2025). "Integrating Indigenous Knowledge with Modern Technologies." ArXiv [source]
[39] No Author. (2026). "State Architecture: The Unification of Terrestrial and Space Engineering." ResearchGate [source]
[40] Helsinki Biennial. (2023). "The Helsinki Biennial Art Mediation Forum 2023." LJMU Research Online [source]
[41] No Author. (n.d.). "Metaverse and Digital Twin ecological costs." ArXiv [source]
[42] Sustainability Directory. (2026). "Future of Privacy-Preserving AI in Environmental Monitoring Systems." Prism [source]
[43] Iliad Project. (2024). "Digital Twin of the Ocean: Trondheim Fjord." Ocean-Twin [source]
[49] BioDT. (n.d.). "About BioDT." BioDT [source]
[51] LUMI Supercomputer. (2024). "First operational BioDT prototype created using LUMI: groundbreaking insights into honey bee performance." LUMI [source]
[54] Saltelli, A. (2026). "Ocean Seminar with Prof. Andrea Saltelli on the limits of digital twin modelling." TwinPolitics [source]
[55] Saltelli, A., Melsen, L. A., & Puy, A. (2025). "Digital twins of the Earth between vision and fiction." University of Birmingham Research Portal [source]
[56] Saltelli, A. et al. (2024). "Bring digital twins back to Earth." WIREs Climate Change [source]
[61] Mrosla, L. et al. (2025). "What grows, adapts, and lives in the digital sphere: Systematic Literature Review on the Dynamic Modelling of Flora and Fauna in Digital Twins." ResearchGate [source]
[63] Haarsma, D. (2021). "Digital Twins: A Comprehensive Solution or Hopeful Vision?" GIM International [source]
[64] Khan, T. et al. (2024). "TwinEco framework abstract." bioRxiv [source]
[71] BioTwinRs. (n.d.). "Digital twins for living systems." BioTwinRs [source]
[77] AI Overview / Source Extractor. (n.d.). "Detailed summary of the TwinEco framework." bioRxiv context [source]
[78] AI Overview / Source Extractor. (n.d.). "Details on the BioDT honey bee prototype." LUMI context [source]
[79] AI Overview / Source Extractor. (n.d.). "Trondheim Fjord Water Quality Pilot details." Ocean-Twin context [source]
[80] AI Overview / Source Extractor. (n.d.). "Details about BioTwinRs and DigitalSoma." BioTwinRs context [source]