Researchers at the Korea Institute of Science and Technology (KIST) have developed KIST-Ocean, a data-driven global ocean prediction model designed to reproduce three-dimensional ocean conditions and forecast how they will evolve. The system learns from decades of atmospheric and oceanic observations, as well as simulated data, rather than calculating every physical process from first principles each time a forecast is produced. According to the research team, the model can generate an ocean forecast extending approximately 200 days into the future in only a few seconds using a single graphics processing unit, or GPU.

KIST-Ocean is trained to work with multiple physical variables that describe the state of the global ocean, including sea-surface temperature, salinity, currents and subsurface heat distribution. It predicts how these variables will change over five-day intervals and resolves ocean conditions down to a depth of 600 meters. The model receives a three-dimensional ocean state and atmospheric boundary conditions as its initial input. It then predicts the ocean’s condition five days later, feeds that prediction back into the system, and repeats the process up to 40 times. This iterative approach produces a global forecast covering nearly seven months at regular five-day intervals.

The researchers say the model’s speed could transform how scientists investigate climate risk. Traditional numerical ocean models use detailed equations governing fluid motion, heat transfer and other physical processes. Although these systems are scientifically powerful, their calculations are computationally demanding, particularly when researchers need to run hundreds or thousands of simulations to examine possible climate scenarios. KIST-Ocean replaces much of that repeated calculation with a trained neural model that has learned statistical relationships embedded in historical and simulated ocean data. The result is a system capable of rapidly generating forecasts and large ensembles at a fraction of the usual computational cost.

Speed alone, however, does not guarantee scientific value. To test whether the artificial intelligence system had learned meaningful ocean dynamics rather than merely reproducing familiar patterns, the research team conducted experiments involving atmospheric forcing. In a virtual wind-generation experiment, changes in wind produced ocean responses including waves, upwelling and downwelling. These processes are central to ocean physics: upwelling carries colder, nutrient-rich water toward the surface, while downwelling transports surface water and heat into deeper layers. The behavior generated by KIST-Ocean was consistent with established physical theories, suggesting that the model captured important links between the atmosphere and the ocean.

The team also tested KIST-Ocean against the development of the 2015 Super El Niño, one of the most powerful El Niño events recorded. During El Niño, unusually warm surface waters spread across the equatorial Pacific, altering atmospheric circulation and influencing weather patterns across much of the world. The model reproduced key features of the event, including the warming of the equatorial Pacific and changes in the distribution of heat beneath the surface. These results provided evidence that the system can represent both visible surface changes and the hidden subsurface processes that help drive long-lasting climate variability.

The significance of the technology extends beyond faster ocean maps. Seasonal and annual forecasts depend heavily on the ocean because seawater changes more slowly than the atmosphere and can preserve climatic information for months. A model that can rapidly update three-dimensional ocean conditions could help researchers explore the likelihood of prolonged heatwaves, droughts, heavy rainfall or shifts in typhoon behavior. It could also support early-warning systems by allowing scientists to test many possible atmospheric and oceanic developments rather than relying on a small number of expensive simulations.

KIST-Ocean may also become a building block for broader artificial intelligence-based Earth system models. Such systems would combine the atmosphere, ocean, land surface, ice and carbon cycle in a unified framework. Integrating these components is technically challenging because each operates on different timescales and interacts through complex feedbacks. A fast ocean component could make it easier to conduct the repeated experiments needed to study those connections, while reducing the computing resources required for climate research. The researchers believe this could lower barriers for institutions that do not have access to the largest supercomputing facilities.

The team cautions that artificial intelligence does not eliminate the need for observations, physical understanding or continued model evaluation. AI forecasts depend on the quality and range of the data used during training, and unusual conditions outside that historical experience can test the limits of any data-driven system. For that reason, the researchers evaluated whether KIST-Ocean reproduced recognized physical mechanisms, not just whether its numerical predictions matched past datasets. Dr. Kang Daehyun, who led the work at KIST’s Center for Climate and Carbon Cycle Research, said the results show that AI can achieve both computational efficiency and a realistic representation of atmosphere–ocean relationships. The team now plans to refine the model as a practical forecasting tool aimed at improving preparedness for climate-related disasters and reducing their social and economic costs.

Subject of Research: AI-based global ocean forecasting and atmosphere–ocean dynamics

Article Title: Data-driven global ocean model resolving atmospherically forced ocean dynamics

News Publication Date: 12-Jun-2026

Web References: https://doi.org/10.1126/sciadv.aed1225

References: Science Advances, DOI: 10.1126/sciadv.aed1225

Image Credits: Korea Institute of Science and Technology

Keywords

KIST-Ocean, artificial intelligence, ocean forecasting, climate prediction, El Niño, ocean dynamics, climate change, machine learning, Earth system models, seasonal forecasting

Deep learning, a subset of machine learning, utilizes neural networks to analyze vast amounts of data and glean insights that were previously unattainable. This article highlights remarkable case studies where deep learning has already made strides in enhancing sustainability efforts. For instance, climate modeling has benefitted significantly from deep learning algorithms capable of processing and interpreting intricate data sets such as satellite imagery and atmospheric measurements. By improving the accuracy of climate predictions, societies can better prepare for the adverse effects of climate change and devise strategies that mitigate its impact.

In the realm of energy management, Sharma and Kaur’s research underscores the role of deep learning in optimizing resource allocation and consumption patterns. Predictive analytics powered by machine learning models help to forecast energy demands more accurately, enabling utility companies to stabilize grids and enhance efficiency. Furthermore, the integration of these technologies with renewable energy sources can facilitate smarter decision-making, ensuring that solar panels and wind turbines are deployed optimally to harness the maximum potential of natural resources.

The agricultural sector, which is often on the frontline of climate change’s effects, stands to gain tremendously from deep learning applications. Advanced image recognition systems can identify crop diseases earlier than traditional methods, allowing farmers to take timely action that not only preserves yield but also reduces the environmental impact of excessive pesticide use. Moreover, deep learning can help in precision farming, where tools equipped with AI analyze factors like soil moisture and nutrient levels to recommend targeted interventions. This data-driven approach promotes more sustainable agricultural practices while simultaneously increasing productivity.

Urban systems are also being transformed through the application of deep learning methodologies. The study articulates how smart city initiatives leverage AI to enhance public transportation systems, reduce energy waste in buildings, and monitor air quality in real-time. By utilizing sensors and machine learning models, urban planners can make informed decisions that promote sustainability and improve the quality of life in metropolitan areas. These advancements lead to the development of greener, more livable cities, effectively addressing growing urbanization challenges.

Additionally, the research delves into the ethical considerations surrounding the deployment of AI technologies in sustainability efforts. With the increasing reliance on data, concerns regarding privacy, surveillance, and the digital divide arise. As AI systems become more prevalent, the need for transparent algorithms that prioritize equitable access to resources and opportunities is imperative. Sharma and Kaur emphasize the necessity of inclusive policies that ensure marginalized communities are not left behind in the transition to AI-infused sustainable systems.

The findings presented in this study are not merely theoretical; they offer a blueprint for actionable strategies that policymakers, businesses, and researchers can utilize. The integration of deep learning into sustainability frameworks has the potential to create significant ripples across industries. This paradigm shift requires investment in research and infrastructure to cultivate an ecosystem in which AI-driven solutions can flourish.

The authors advocate for collaborative efforts among governments, tech companies, and civil society to create a robust regulatory framework that fosters innovation while addressing ethical challenges. Public-private partnerships can accelerate the development of sustainable technologies and provide real-world applications that can be scaled effectively. Beyond technological advancements, fostering a culture of sustainability and environmental awareness is equally crucial. Engaging communities in the dialogue surrounding these innovations promotes broader societal support and understanding of the importance of sustainability.

Moreover, it is vital to educate the next generation about the capabilities and responsibilities associated with AI technologies. By embedding sustainability principles into educational curricula, future leaders will be better equipped to manage the complexities inherent in balancing technological advancement with ecological conservation. In this context, deep learning becomes not only a tool for current challenges but also a cornerstone for shaping a sustainable future.

In conclusion, the study by Sharma and Kaur serves as a beacon of hope, illustrating how deep learning can be an ally in the fight against the existential threats posed by climate change, energy crises, agricultural inefficiencies, and urban challenges. This research is a compelling call-to-action for all stakeholders to embrace the transformative potential of AI in fostering sustainable development. As we stand at this critical juncture, the marriage of technology and sustainability must be a priority, resonating with collective responsibility for the well-being of our planet and future generations.

By embracing deep learning and its vast capabilities, we have the opportunity to pave a path towards a more sustainable, equitable, and resilient world. The insights offered by Sharma and Kaur will undoubtedly play a central role in guiding research, policy formulation, and real-world applications as we continually seek to harmonize human progress with ecological balance.

Subject of Research: Deep learning applications for sustainable development.

Article Title: Deep learning for sustainable development across climate, energy, agriculture and urban systems.

Article References:
Sharma, H., Kaur, S. Deep learning for sustainable development across climate, energy, agriculture and urban systems.
Discov Sustain 6, 1408 (2025). https://doi.org/10.1007/s43621-025-02186-6

Image Credits: AI Generated

DOI: https://doi.org/10.1007/s43621-025-02186-6

Keywords: Deep learning, sustainability, climate change, energy management, agriculture, urban systems, artificial intelligence.

The intense and comprehensive analysis carried out by the team highlights the unique capabilities of satellite-derived data in monitoring various critical sectors affected by climate change. Unlike conventional methods that rely primarily on ground-based measurements, which are often sparse or infeasible in remote areas, Earth observation satellites provide a consistent and holistic view of the planet. The data collected spans up to six decades, delivering repeatable and objective measurements that no other monitoring system can offer. This capacity for extensive data acquisition enables researchers and policymakers alike to gain insights into the ongoing transformations occurring within different climate-sensitive sectors.

The study focuses particularly on four pivotal areas: agriculture, biodiversity, extreme events, and health. In agriculture, satellite technology has proven instrumental in monitoring water productivity, irrigation efficiency, and shifts in crop migration patterns. These insights not only facilitate improved agricultural practices but also empower farmers to adapt to changing climatic conditions more effectively than ever before. By leveraging Earth observation data, agricultural stakeholders can optimize resource use and increase crop yields, which is crucial for ensuring food security in an increasingly uncertain climate landscape.

Biodiversity conservation efforts, too, are benefiting from satellite data. Platforms such as Global Mangrove Watch and Global Forest Watch are equipped with geospatial information that tracks changes in ecosystem extent and health. This critical data enables conservationists to monitor critical habitats and implement measures for protecting endangered species and ecosystems facing the brunt of climate change. Such information is invaluable for crafting effective management strategies that prioritize biodiversity preservation in the face of mounting environmental pressures.

The study further underscores the importance of monitoring extreme events, where satellites play a vital role in characterizing floods, droughts, and heatwaves. These extreme phenomena can have devastating impacts on human society, infrastructure, and natural ecosystems. Satellite-derived data allows for improved real-time assessments of these events, providing communities with crucial information that supports disaster preparedness and response. The ability to understand the extent and implications of extreme events can save lives and mitigate infrastructure damage, underscoring the life-saving potential of space-based observations.

Addressing health concerns, the research team emphasizes how Earth observation data on land surface temperature and air quality can inform assessments of heat exposure and disease outbreaks. With the increasing frequency of heatwaves and the spread of vector-borne diseases, such information is essential for public health planning and response strategies. Policymakers can utilize these insights to develop targeted interventions, ensuring that vulnerable populations receive the support and resources they need to cope with climate-induced health risks.

Leading the research, Professor Aaron Golden articulated the unique role of satellite technology in supporting global climate agreements such as the Paris Agreement. He underscored that the insights derived from long-term observations empower decision-makers to assess progress toward adaptation goals and identify regions most at risk from climate impacts. The ability to quantify and track adaptation efforts is vital for developing tailored strategies that enhance resilience and reduce vulnerability to climate change.

Dr. Sarah Connors, the lead author of the study from the European Space Agency, further emphasized the necessity of integrating Earth observation data into the frameworks of global climate indicators. By ensuring that satellite data is considered from the outset of adaptation tracking, researchers can avoid the pitfalls experienced with the Sustainable Development Goals, where retrofitting data sources proved to be a considerable challenge. Such foresight will undoubtedly facilitate more effective tracking of adaptation progress, leading to improved outcomes across sectors.

In light of these findings, the research team advocates for a concerted effort to incorporate satellite-derived information into adaptation frameworks globally. By harnessing the transformative potential of Earth observation data, policymakers, scientists, and communities can collaborate more effectively to respond to climate threats. The synergy between satellite technology and artificial intelligence not only enhances our understanding of climate adaptation but also equips stakeholders with the tools necessary to drive meaningful change in a time of urgency.

Professor Frances Fahy, Director of the University of Galway’s Ryan Institute, echoed the sentiment that this research exemplifies the university’s commitment to world-class, impact-driven research. By utilizing satellite Earth observation data, researchers are addressing pressing climate challenges and shaping international climate policy with acumen. This multidimensional approach emphasizes the importance of interdisciplinary research in tackling the complexities of climate adaptation.

As the world grapples with the multifaceted implications of climate change, the insights provided by this study offer a beacon of hope. By bridging the gap between satellite technology and real-world applicability, researchers are paving the way for a future where evidence-based strategies empower societies to adapt and thrive amidst the challenges posed by a changing climate. The full study, published in the esteemed journal npj Climate and Atmospheric Science, presents the pioneering findings and innovative methodologies that promise to redefine our understanding of adaptation in an era marked by environmental uncertainty.

The urgency to act on climate adaptation cannot be overstated. As the impacts of climate change continue to evolve, the role of Earth observation in monitoring progress and guiding decision-making becomes increasingly critical. The innovative methodologies borne from the collaboration between the University of Galway researchers and the European Space Agency present a significant leap forward in understanding how satellite-derived indicators can serve as essential tools in tracking and enhancing climate resilience globally.

The combination of satellite technology and data-driven insights represents a transformative shift in how we perceive and address climate adaptation. As this field continues to evolve, it holds the potential to empower communities and policymakers with the knowledge and tools necessary to navigate an uncertain future while fostering resilience in the face of unprecedented climate challenges.

Subject of Research: Climate adaptation using satellite-based Earth observation
Article Title: Earth observations for climate adaptation: tracking progress towards the Global Goal on Adaptation through satellite-derived indicators
News Publication Date: 11-Nov-2025
Web References: Published Study
References: DOI 10.1038/s41612-025-01251-1
Image Credits: Credit – European Space Agency

Keywords

Earth observation, climate adaptation, satellite data, agriculture, biodiversity, extreme events, health, global climate policy, Paris Agreement, resilience.

Carbon dioxide removal has emerged as a critical component in the global efforts to mitigate climate change, complementing emission reductions by aiming to actively extract CO₂ from the atmosphere or prevent its emission at source. However, until now, gaps in the accessibility and synthesis of the burgeoning scientific output have hindered policymakers and researchers from fully appreciating the scale and depth of knowledge in this field. The innovative application of AI algorithms to systematically categorize and analyze thousands of publications represents a paradigm shift, revealing hidden connections and underexplored avenues that traditional review methods could not capture at this scale.

The research team employed advanced natural language processing models to sift through the entirety of indexed research, spanning diverse disciplines from engineering and environmental sciences to economics and policy analysis. This approach allowed them to overcome the limitations imposed by human bias and manual screening, which often restrict the scope or lead to incomplete assessments due to the sheer volume and heterogeneity of research. The AI system’s ability to rapidly process and classify articles by methodology, regional focus, and maturity stage resulted in the construction of a dynamic, high-resolution map of the CDR research landscape.

One of the most striking revelations from the systematic mapping exercise is the identification of an unexpectedly high volume of literature focusing on various CDR technologies, including direct air capture, bioenergy with carbon capture and storage (BECCS), afforestation, and soil carbon sequestration. Contrary to earlier assumptions that research concentrated on a handful of predominant methods, the AI-driven analysis demonstrates that the scientific community has investigated a much broader array of techniques, each bearing unique challenges and potentials. This comprehensive cataloging opens new pathways for comparative assessments crucial for prioritizing resources and guiding innovation.

Moreover, the study highlights significant geographic disparities in CDR research attention, with a concentration of publications emanating from North America, Europe, and parts of East Asia, while voices from developing regions remain underrepresented. Such findings underscore the need for greater inclusivity and support for research initiatives in areas disproportionately vulnerable to climate impacts but currently underserved in scientific inquiry. The AI mapping tool equips stakeholders with data to develop more balanced and equitable research agendas, fostering international cooperation essential for global climate mitigation.

Technical scrutiny of the mapped literature also exposed varying degrees of technological readiness and scalability among different CDR approaches. Some techniques, like enhanced weathering and mineral carbonation, have received less empirical validation despite theoretical promise, revealing gaps that could hamper their practical deployment. The interconnection of these findings with policy frameworks is particularly timely, as governments worldwide debate the integration of CDR into national climate strategies and the mechanisms for incentivizing innovation.

Beyond cataloging, the AI-enhanced mapping brings a meta-analytical perspective by illuminating trends over time, revealing accelerating research output and evolving thematic emphases aligned with global policy developments such as the Paris Agreement. This dynamic understanding provides a real-time dashboard for funders, scientists, and decision-makers to monitor the research ecosystem’s responsiveness and pivot based on emerging needs or technological breakthroughs. The visualization tools accompanying the study translate complex bibliometric data into intuitive formats, helping non-specialists engage with the scientific progress effectively.

Importantly, the researchers discuss the methodological rigor of the AI approach, detailing the training and validation processes ensuring the systematic map’s reliability and reproducibility. They emphasize transparency by making their dataset available to the wider community, encouraging collaborative refinement and the integration of complementary data sources. This openness addresses common criticisms related to black-box AI systems and builds confidence in deploying such techniques for large-scale knowledge synthesis in environmental research fields.

The implications of this study extend beyond academic boundaries. By demonstrating the feasibility and advantages of using AI to enhance systematic reviews in rapidly growing fields, it sets a precedent for environmental science disciplines grappling with information overload. The approach enables continuous updating and refinement of knowledge maps, essential in contexts where timely insights can influence urgent policy or investment decisions. This agility contrasts with traditional static literature reviews that often become outdated before influencing practice.

Critically, this expanded understanding of the scientific landscape surrounding CDR can inform risk assessments, as diversifying technology portfolios reduce reliance on single solutions vulnerable to unforeseen challenges. By bringing clarity to the distribution and maturity of knowledge clusters, the study aids in identifying research synergies and knowledge gaps, facilitating strategic collaborations across disciplines and sectors. The holistic view supported by AI could accelerate technology transfer and hybrid approaches combining multiple CDR strategies.

Furthermore, the AI methodology’s scalability offers potential applications in monitoring the scientific discourse on other pressing global issues such as biodiversity loss, water security, and renewable energy transitions. The capacity to synthesize multidisciplinary knowledge in near real-time empowers the global research community to engage adaptively with complex environmental challenges. By harnessing AI as an analytical partner rather than merely a data processing tool, scientists augment their ability to discern patterns and emerging paradigms hidden within voluminous academic outputs.

While celebrating the technological advances embodied in their work, the authors caution against overreliance on automated methods without critical human oversight. They advocate for integrating expert judgment to contextualize findings appropriately and navigate nuanced interpretations beyond algorithmic outputs. Their interdisciplinary team, combining climate scientists, computer scientists, and knowledge management experts, exemplifies the collaborative spirit necessary to maximize AI’s benefit in environmental scholarship.

In a broader perspective, this research encapsulates a significant stride toward democratizing scientific knowledge on climate solutions. By revealing the true scale and complexity of CDR literature, it encourages informed dialogue among scientists, policymakers, industry stakeholders, and the public. Effective communication of such comprehensive evidence bases bolsters societal trust in emerging technologies and facilitates consensus-building essential for coordinated climate action.

As the global community intensifies its commitment to achieving net-zero emissions and tackling the climate crisis, tools like the AI-enhanced systematic mapping unveiled by Lück and colleagues will become indispensable. They provide a robust foundation to streamline research efforts, allocate funding strategically, and design policy interventions grounded in a rich, nuanced understanding of existing knowledge. This fusion of artificial intelligence with climate science research heralds a new era of evidence-based environmental innovation and governance.

Subject of Research:
Systematic mapping of carbon dioxide removal scientific literature using artificial intelligence to reveal the extensiveness and diversity of research in the field.

Article Title:
Scientific literature on carbon dioxide removal revealed as much larger through AI-enhanced systematic mapping.

Article References:
Lück, S., Callaghan, M., Borchers, M. et al. Scientific literature on carbon dioxide removal revealed as much larger through AI-enhanced systematic mapping. Nat Commun 16, 6632 (2025). https://doi.org/10.1038/s41467-025-61485-8

Image Credits: AI Generated