Machine learning (ML) is transforming data analysis and modelling of the Earth system. While statistical and data-driven models have been used for a long time, recent advances in ML and deep learning now allow for encoding non-linear, spatio-temporal relationships robustly without sacrificing interpretability. This has the potential to accelerate climate science through new approaches for modelling and understanding the climate system. For example, ML is now used in the detection and attribution of climate signals, to merge theory and Earth observations in innovative ways, and to directly learn predictive models from observations. The limitations of machine learning methods also need to be considered, such as requiring, in general, rather large training datasets, data leakage, and/or poor generalisation abilities so that methods are applied where they are fit for purpose and add value.

This session aims to provide a venue to present the latest progress in the use of ML applied to all aspects of climate science, and we welcome abstracts focussed on, but not limited to:

More accurate, robust and accountable ML models:
- Hybrid models (physically informed ML, parameterizations, emulation, data-model integration)
- Novel detection and attribution approaches
- Probabilistic modelling and uncertainty quantification
- Uncertainty quantification and propagation
- Distributional robustness, transfer learning and/or out-of-distribution generalisation tasks in climate science
- Green AI

Improved understanding through data-driven approaches:
- Causal discovery and inference: causal impact assessment, interventions, counterfactual analysis
- Learning (causal) process and feature representations in observations or across models and observations
- Explainable AI applications
- Discover governing equations from climate data with symbolic regression approaches

Enhanced interaction:
- The human in the loop - active learning & reinforcement learning for improved emulation and simulations
- Large language models and AI agents - exploration and decision making, modeling regional decision-making
- Human interaction within digital twins

Machine learning (ML) methods have emerged as powerful tools to tackle various challenges in ocean science, encompassing physical oceanography, biogeochemistry, and sea ice research.
This session aims to explore the application of ML methods in ocean science, with a focus on advancing our understanding and addressing key challenges in the field. Our objective is to foster discussions, share recent advancements, and explore future directions in the field of ML methods for ocean science.
A wide range of machine learning techniques can be considered including supervised learning, unsupervised learning, interpretable techniques, and physics-informed and generative models. The applications to be addressed span both observational and modeling approaches.

Observational approaches include for example:
- Identifying patterns and features in oceanic fields
- Filling observational gaps of in-situ or satellite observations
- Inferring unobserved variables or unobserved scales
- Automating quality control of data

Modeling approaches can address (but are not restricted to):
- Designing new parameterization schemes in ocean models
- Emulating partially or completely ocean models
- Parameter tuning and model uncertainty

The session welcomes also submissions at the interface between modeling and observations, such as data assimilation, data-model fusion, or bias correction.

Machine learning (ML) is being used throughout the geophysical sciences with a wide variety of applications.
Advances in big data, deep learning, and other areas of artificial intelligence (AI) have opened up a number of new approaches.

Many fields (climate, ocean, NWP, space weather etc.) make use of large numerical models and are now seeking to enhance these by combining them with scientific ML/AI.
Examples include ML emulation of computationally intensive processes, training on high resolution models or data-driven parameterisations for sub-grid processes, and Bayesian optimisation of model parameters and ensembles amongst several others.

Doing this brings a number of unique challenges, however, including but not limited to:
- enforcing physical compatibility and conservation laws, and incorporating physical intuition into ML models,
- ensuring numerical stability,
- coupling of numerical models to ML frameworks and language interoperation,
- handling computer architectures and data transfer,
- adaptation/generalisation to different models/resolutions/climatologies,
- explaining, understanding, and evaluating model performance and biases.

Addressing these requires knowledge of several areas and builds on advances already made in domain science, numerical simulation, machine learning, high performance computing, data assimilation etc.

We solicit talks that address any topics relating to the above.
Anyone working to combine machine learning techniques with numerical modelling is encouraged to participate in this session.

Cities are complex multi-scale systems, composed of multiple sub-components (e.g. for population, energy, transport, climate) that interact with each other on various time scales (e.g. hourly, seasonal, annual). Urban models and digital twins for urban planning applications and policies aimed at shaping healthier and more sustainable urban environments should account for such complex interactions as they regulate the growth and functioning of cities, often resulting in emergent large-scale phenomena. Yet our ability to quantitatively describe city behaviour is still limited due to the variety of processes, scales, and feedbacks involved.
In this session we welcome modelling and monitoring studies that focus on multi-sector dynamics and city-biosphere interactions. These include – but are not limited to – demography, urban transport networks, energy consumption, anthropogenic emissions, urban climate, pollution, urban hydrology and ecology.
The aim is to elucidate complex urban dynamics, identify strategies and methods for the development of models and digital twins of cities, and understand how the form and function of urban environments can improve liveability and well-being of their citizens.
This session welcomes concepts, methodologies and disruptive models to overcome current scientific bottlenecks, to better deal with non-linearities, multi-component systems and extremes over a wide range of scales in geophysical and urban systems.

Climate change and widespread biodiversity loss are urgent challenges facing humanity, whose effects threaten human wellbeing, economies and planetary stability. There is increasing evidence that these two crises are strongly interconnected and might even be mutually reinforcing. However, climate- and biodiversity change are typically investigated through siloed approaches. This limits our ability to assess the feedbacks between these two major trends and to ultimately/eventually design policy solutions that fully take into account the trade-offs and synergies between climate change mitigation, adaptation, and biodiversity conservation.

In this session, we invite scientists from all disciplines working at the interface of these fields, and in particular on the linked relationships and processes between climate (change, variability, extremes) and biodiversity (taxonomic, functional, structural). We are especially interested in studies that investigate feedbacks mechanisms between biodiversity and the climate system at different spatial and temporal scales, from experimental, observational, data-science, and/or modelling perspectives, as well as on how human activities, such as land cover conversion or nature conservation, might influence these interactions.

Digital twins of our planet, at present-day and over geological timescales, are becoming central to decision-making and de-risking for a broad range of applications from natural hazard risk assessments, climate modelling, and to resource analysis. Emerging modelling techniques are promising to value-add to complex, and sometimes obscure, geological and geophysical data through machine learning, artificial intelligence, and other advanced statistical and nonlinear optimisation techniques. In addition, these new techniques provide an avenue to increase the quantifiability of geological processes at a wide range of spatial and temporal scales. This includes the key requirement to incorporate better quantifications of uncertainty in both parameter values and model choice, as well as the fusion between geophysical sensing and geological constraints with numerical modelling of Earth Systems.

We invite submissions from all disciplines that aim to model or constrain one or more Earth Systems over modern and geological timeframes. We welcome submissions that are analytical or lab-focused, field-based, or involve numerical modelling. This session also aims to explore cutting-edge methods, tools, and approaches that push the boundaries of geophysical inference and uncertainty analysis, and geological data fusion. We ask the question `Where to next?’ in our collective quest to develop digital twins of our planet.

The session will also celebrate the contributions of early career researchers, open/community philosophy of research, and innovations that have adopted inter-disciplinary approaches.

The Coupled Model Intercomparison Project (CMIP) advances climate system understanding, but Earth System Models (ESM) exhibit disparities, particularly in responses to forcings and system coupling. As the IPCC relies on CMIP to provide information for policy decisions, a multidisciplinary approach is crucial to address uncertainties across the full CMIP production line. This session invites studies on climate forcings, climate responses, uncertainties in forcing agents, and model disparities in CMIP projections.

We welcome diverse climate-forcing research, including historical and future, anthropogenic and natural forcing development, idealized Earth System Model studies, observational evaluations, and works spanning all climate system components. Topics may include identifying disparities in CMIP ESMs, quantifying uncertainties, and addressing key scientific priorities for future model development. Contributions on opportunities, challenges, and constraints in using CMIP output for impact research, especially at regional scales, are encouraged.
This session ultimately aims at fostering collaboration among climate scientists, observationalists and modelers to address climate change challenges. Convened by WCRP CMIP Forcing Task Team and Fresh Eyes on CMIP, it aims to enhance understanding of CMIP uncertainties and prepare for CMIP6Plus and CMIP7 climate-forcing datasets.

Scientific disciplines strive to explain the causes of observed phenomena. In Earth sciences, in particular in climate research, the notion of causality is discussed and understood from several different points of view. Hannart et al. (BAMS, 2016), following Judea Pearl, state that “Causal counterfactual theory provides clear semantics and sound logic for causal reasoning and may help foster research on, and clarify dissemination of, weather and climate-related event attribution.” Changing focus from explanation of single events to understanding phenomena evolving in time, represented by time series, causality can be understood in terms of improved predictability, as proposed by Norbert Wiener and formulated for time series by C.W.J. Granger. Granger causality has been further generalized for nonlinear systems using methods rooted in information theory. Extensions from bivariate to multivariate time series can also point to indirect causations. X. S. Liang and R. Kleeman derive formulas for information flows based on dynamical equations. The Wiener-Granger concept of improved predictability has been translated into computer science as compressibility changes in effect data due to knowledge of cause data. The information-theoretic formulation of Granger causality and other methods have recently been adapted for complex systems with multiple time scales and/or heavy-tailed probability distributions and extreme events. Methods for turning multivariate data into causal graphs based on Bayesian reasoning and machine learning are also intensively applied in the Earth sciences.

The session welcomes contributions discussing these diverse approaches to causality analysis in Earth sciences, with an emphasis on comparative discussions. Learning causal relationships from Earth system data is vital for understanding complex dynamics, predicting changes, and informing strategies. This session invites innovative approaches and case studies employing causal inference techniques across Earth sciences, fostering interdisciplinary discussions and encouraging the development of robust causal analysis frameworks. Topics include causal discovery methods, causal effect estimation, applications of causal inference to climate change, causal modeling, network analysis, and addressing challenges and limitations in applying causal inference to Earth system science.

Downscaling aims to process and refine global climate model output to provide information at spatial and temporal scales suitable for impact studies. In response to the current challenges posed by climate change and variability, downscaling techniques continue to play an important role in the development of user-driven climate information and new climate services and products. In fact, the "user's dilemma" is no longer that there is a lack of downscaled data, but rather how to select amongst the available datasets and to assess their credibility. In this context, model evaluation and verification is growing in relevance and advances in the field will likely require close collaboration between various disciplines.

Furthermore, epistemologists have started to revisit current practices of climate model validation. This new thread of discussion encourages to clarify the issue of added value of downscaling, i.e. the value gained through adding another level of complexity to the uncertainty cascade. For example, the ‘adequacy-for-purpose view’ may offer a more holistic approach to the evaluation of downscaling models (and atmospheric models, in general) as it considers, for example, user perspectives next to a model’s representational accuracy.

In our session, we aim to bring together scientists from the various geoscientific disciplines interrelated through downscaling: atmospheric modeling, climate change impact modeling, machine learning and verification research. We also invite philosophers of climate science to enrich our discussion about novel challenges faced by the evaluation of increasingly complex simulation models.

Contributions to this session may address, but are not limited to:

- newly available downscaling products,
- applications relying on downscaled data,
- downscaling method development, including the potential for machine learning,
- bias correction and statistical postprocessing,
- challenges in the data management of kilometer-scale simulations,
- verification, uncertainty quantification and the added value of downscaling,
- downscaling approaches in light of computational epistemology.

Modelling and exploring forest ecosystems under future climate and management has never been more critical in the face of accelerated climate change and human-induced disturbances. Consequently, understanding the dynamics of forest ecosystems, which not only act as essential carbon sinks but also support biodiversity and a wide range of ecosystem services, and predicting their responses to changing environmental conditions and future management actions has become vital. To this end, this session aims to shed light on the innovations and advancements in forest modelling and assessment of ecosystem services, within the following focus areas:
1. Next-Generation Forest Models and Climate Dynamics: Overview of models designed to dynamically intertwine climate drivers with forest growth patterns, offering a thorough representation of carbon, nitrogen, and phosphorus cycles to predict forest responses to climate impact.
2. Model-Data Fusion in Forest Modelling: Discussion on how to integrate data from different sources (e.g., remote sensing, forest inventories, eddy-covariance) into forest modelling frameworks. Overview of computational techniques applied for model calibration, evaluation, and averaging and for data assimilation.
3. Large-Scale Forest Modelling for Feedback Mechanisms: Exploration of tools that assess the complex feedback loops among forest adaptive/mitigative strategies, localised climate changes, natural disturbances, and the permanence of forest carbon stocks.
4. Future Climate and Management Driven Forest Structural Modelling: In-depth look at how alternative management practices and climate drivers influence forest architecture, yielding structural indicators pivotal for the evaluation of biodiversity and ecosystem services.
5. Framework for Linking Forest Structural Indicators to Biodiversity and Ecosystem Services: Discussion on innovative methodologies establishing connections between forest structural variables from cutting-edge models and biodiversity and ecosystem service indicators, prioritizing the assessment and economic evaluation of potential synergies and trade-offs in forest management decisions.
This session invites contributions from researchers, practitioners, and policymakers. It seeks to become a vibrant forum for exchanging knowledge, insights, and best practices, furthering our collective goal of ensuring sustainable and resilient forest ecosystems in a rapidly changing world.

Space-based measurements of the Earth System, including its atmosphere, oceans, land surface, cryosphere, biosphere, and interior components, require extensive prelaunch and post-launch calibration and validation activities to evaluate scientific accuracy, characterise uncertainties and ensure the fitness for purpose of the geophysical information provided throughout lifetime of satellite missions. This stems from the need to demonstrate unambiguously that space-based measurements, which are typically based on engineering measurements by the detectors (e.g. photons), are sensitive to and can be used to reliably retrieve the geophysical and/or biogeochemical parameters of interest across the Earth.

Most geophysical parameters vary in time and space, and the retrieval algorithms used must be accurate and tested under the representative range of conditions encountered. Satellite missions also benefit from the availability of precursor data made available from other satellite missions, field campaigns, and/or surface-based measurement networks that are used in the definition of geophysical products and for the development and testing of the retrieval algorithms prior to launch during the satellite and ground segment development. Post-launch calibration and validation over the lifetime of missions assure that any long-term variation in observation can be unambiguously tied to the evolution of the Earth system. Such activities are also critical in ensuring that measurements from different satellites can be inter-compared and used seamlessly to create long-term multi-instrument/multi-platform data sets, which serve as the basis for large-scale international science investigations into topics with high societal or environmental importance.

This session seeks presentations on the use of surface-based, airborne, and/or space-based observations to develop precursor data sets and support both pre- and post- launch calibration/validation and retrieval algorithm development for space-based satellite missions measuring our Earth system. A particular but not exclusive focus will be on collaborative activities carried out jointly by NASA and ESA as part of their Joint Program Planning Group Subgroup on provision of precursor data sets for future ESA, NASA, and related partner missions, and the full range of pre- and post-launch calibration and validation and field activities for these satellite projects.

Modelling is fundamental for assessing various soil processes and interactions at different scales and resolution, while healthy soils are fundamentally important in sustaining a wide range of ecosystem services. Crossing interdisciplinary borders and integrating knowledge from various fields is essential in developing more accurate and comprehensive models to better capture the complexity of soil processes/mechanisms in natural and cultivated systems, address knowledge-gaps, and tackle the challenges related to data-integration, heterogeneity and uncertainty of modelling predictions across disciplines. An interdisciplinary approach is also needed in light of recent technological advances, such as computational approaches, model-coupling, geomatics, remote sensing/earth observation, machine learning, surveying and data collection sensors/sensor platforms, real-time data-streams, all of which provide opportunities for promoting new modelling generations integrating soil science across disciplines.

Integration of various disciplines and modelling is also essential for better understanding of the role of soil health, which includes concepts soil capacity and functionality towards a wide range of ecosystem services. Several measures to support soil health and tackle soil degradation have been proposed in the scientific literature, as well as several indicators to monitor expected benefits. The need for standardized data covering the broad concept of soil health and degradation is arising, along with the lack of information on relationships between soil quality and agriculture, forest and grassland resilience, and the socio-economic and environmental impacts of these measures. The scattered data availability and their complex integration for agronomic/environmental management and policy decisions may partly be covered by many European/international/national initiatives in the frameworks of the H2020, Horizon Europe, PRIMA, FAO programs, and other programs.

This session aims to promote and enhance communication and exchange of knowledge among scientists from modelling community, soil research and various related projects, linking different disciplines, and is open to contributions in a wide range of related topics, ranging from modelling soil systems to ecosystem and landscape modelling, soil health, degradation and living labs, while striving to contribute towards tackling current research challenges, addressing the knowledge-gaps, and informing policy.