The General Assembly of EGU will take place in Vienna, Austria & Online, 14–19 April 2024. The teams of XAIDA are conveners and/or co-conveners of 11 sessions in the frameworks of the project.
A lot of the XAIDA ECS (Early Career Scientists) are involved as conveners and co-converners:
Emanuele Bevacqua (UFZ), Miguel-Ángel Fernández-Torres (Univ. València), Vera Melinda Galfi (VU), Aglaé Jézéquel (IPSL), Marlene Kretschmer (Univ. Leipzig), Sebastian Sippel (Univ. Leipzig), Laura Suarez-Gutierrez (ETH).
WP leaders & Senior Scientists are also involved:
Ana Bastos (MPG), Gustau Camps-Valls (Univ. València), Davide Faranda (IPSL), Erich Fischer (ETH), Miguel Mahecha (Univ. Leipzig), Markus Reichstein (MPG), Jakob Zscheischler (UFZ)
The deadline to submit the abstracts is January 10th 2024 – 13:00 CET!
Breaking News: Davide FARANDA, CNRS (XAIDA, EDIPI), is now the NP (Nonlinear Processes in Geosciences) Division President of the EGU for the term 2025–2027.
Abstracts are solicited related to the understanding, prediction and impacts of weather, climate and geophysical extremes, from both an applied and theoretical viewpoint.
In this session we propose to group together the traditional geophysical sciences and more mathematical/statistical and impacts-oriented approaches to the study of extremes. We aim to highlight the complementary nature of these viewpoints, with the aim of gaining a deeper understanding of extreme events. This session is a contribution to the EDIPI ITN, XAIDA and CLINT H2020 projects and to the Swedish Centre for Impacts of Climate Extremes. We welcome submissions from both project participants and the broader scientific community.
Potential topics of interest include but are not limited to the following:
- Dynamical systems theory and other theoretical perspectives on extreme events;
- Data-driven approaches to study extreme events and their impacts, incl. machine learning;
- Representation of extreme events in climate models;
- Downscaling of weather and climate extremes;
- How extremes have varied or are likely to vary under climate change;
- Attribution of extreme events;
- Early warning systems and forecasts of extreme events;
- Methodological and interdisciplinary advances for diagnosing impacts of extreme events.
Attribution research in the context of climate change investigates whether and to what extent human influence is contributing to changes in the climate system as well as ensuing changes in natural, managed, and human systems. These analyses combine observations of those systems with model-based evidence or process understanding.
The aim of Detection and Attribution (D&A) studies is to identify historical changes over long timescales (typically multi-decadal), and quantify the contributions of various external forcings as their signal emerges from internal climate variability. Moreover, event attribution (EA) assesses how human-induced climate change is modifying the frequency or intensity of, e.g., extreme, weather events and/or their impacts (e.g., a drought, a heatwave, a crop failure). This rapidly evolving field has introduced a range of methodologies and different ways of framing attribution questions. The attribution of climate change impacts is particularly complex due to the influence of additional non-climatic human influences. This session invites recent studies from the broad spectrum of attribution research that address some or all steps of the climate-impact chain from emissions to climate variables, to impacts in natural, managed, and human systems and aims to explore the diversity of methods employed across disciplines and schools of thought.
This session covers research on common and new methodologies, including improved statistical methods and approaches based on statistical causality or Artificial Intelligence. It also covers a broad range of applications, case studies, current challenges of the field, and avenues for expanding the attribution community. We particularly welcome submissions that address trends, extremes, include compound/cascading events and/or assess implications of recent trends for constraining future changes – all of which test the limits of the present science. We equally welcome studies that push the limits by attributing impacts further downstream from climatic phenomena by attributing changes and events in ecosystems, agriculture, health, economics, and other natural, managed, and/or human systems. Contributions that compare approaches, develop or explore the influence of different counterfactual data for attribution studies, explore the use of new observationally-derived data sources, or account for changes in exposure and vulnerability are as welcome as applications of existing approaches to novel terrain.
ITS2.3/CL0.1.1
High-impact climate and weather events typically result from the interaction of multiple climate and weather drivers, as well as vulnerability and exposure, across various spatial and temporal scales. Such compound events often cause more severe socio-economic impacts than single-hazard events, rendering traditional univariate extreme event analyses and risk assessment techniques insufficient. It is, therefore, crucial to develop new methodologies that account for the possible interaction of multiple physical and societal drivers when analysing high-impact events under present and future conditions. Despite the considerable attention from the scientific community and stakeholders in recent years, several challenges and topics must still be addressed comprehensively.
These include:
(1) identifying the compounding drivers, including physical drivers (e.g., modes of variability) and/or drivers of vulnerability and exposure, of the most impactful events;
(2) Developing methods for defining compound event boundaries, i.e. legitimate the ‘cut-offs’ in the considered number of hazard types to ultimately disentangle enough information for decision-making;
(3) Understanding whether and how often novel compound events, including record-shattering events, will emerge in the future;
(4) Explicitly addressing and communicating uncertainties in present-day and future assessments (e.g., via climate storylines/scenarios);
(5) Disentangling the contribution of climate change in recently observed events and future projections;
(6) Employing novel Single Model Initial-condition Large Ensemble simulations from climate models, which provide hundreds to thousands of years of weather, to better study compound events.
(7) Developing novel statistical methods (e.g., machine learning, artificial intelligence, and climate model emulators) for compound events;
(8) Assessing the weather forecast skill for compound events at different temporal scales;
(9) Evaluating the performance of novel statistical methods, climate and impact models, in representing compound events and developing novel methods for reducing uncertainties (e.g., multivariate bias correction and emergent constraints);
and (10) engaging with stakeholders to ensure the relevance of the aforementioned analyses.
We invite presentations considering all aspects of compound events, including but not limited to the topics and research challenges described above.
ESSI1.2
As Artificial Intelligence (AI) applications in geosciences grow, the quest for understanding Machine Learning (ML) and Deep Learning (DL) models becomes pivotal. This session highlights the critical role of Explainable Artificial Intelligence (XAI) in strengthening our ability to trust, comprehend, and improve AI models. To achieve this, it brings together specialists in geoscience, data science, and AI.
We strongly encourage submissions that employ methods enabling AI systems to furnish lucid and understandable explanations for their decisions.
This multidisciplinary session encompasses contributions related to the following lines of research:
- Exploration of novel XAI techniques and methodologies that enhance the transparency and interpretability of ML/DL models used in geosciences.
- Real-world case studies where XAI has made substantial progress in understanding and managing specific geoscience tasks and physical processes.
- Process understanding via XAI and hybrid, physically-informed modeling.
- Quantitative evaluation and comparison of the effectiveness of XAI models.
- Strategies towards more scientifically valuable explanations (e.g., use of Large Language Models (LLMs) for XAI).
CL2.3
Recent extreme events with intensities unprecedented in the observational record are causing high impacts globally. The northern hemisphere summer of 2023 saw exceptional heat in North America, Europe and China. Sea surface temperatures across the North Atlantic and the Mediterranean reached record levels while the Antarctic sea ice was record low. Marine heatwaves affected almost the entire tropical north Atlantic. Some of these events would have been nearly impossible without human-made climate change and broke records by large margins. Further, compound behaviour and cascading effects and risks are becoming evident, such as the spike in food prices induced by the effects of the war in Ukraine on top of concurrent drought across regions with subsequent crop failure. Finally, continuing warming does not only increase the frequency and intensity of events like these, or other until yet unprecedented extremes, it also potentially increases the risk of crossing tipping points and triggering abrupt changes. In order to increase preparedness for high impact climate events, it is important to develop methods and models that are able to represent these events and their impacts, and to better understand how to reduce the risks.
This session aims to bring together the latest research on modelling, understanding, development of storylines and managing plausible past and future high impact climate events and their impacts. We are interested in rare and low-probability heavy precipitation events, droughts, floods, storms and temperature extremes from time scales of hours to decades, including compound, cascading, and connected extremes, high-impact event storylines, as well as the effect of tipping points and abrupt changes driven by climate change, societal response, or other mechanisms (e.g., volcanic eruption).
We welcome a variety of methods to quantify and understand high-impact climate events in present and future climates, such model experiments and intercomparisons; insights from paleo archives; climate projections (including large ensembles, and unseen events); attribution studies; and the development of storylines. We invite work on tipping elements/tipping points; abrupt changes; worst case scenarios; identification of adaptation limits; and the opportunities and solutions to manage the greatest risks.
The session is further informed by the World Climate Research Programme lighthouse activities on Safe Landing Pathways and Understanding High-Risk Events.
NH10.3
The complex, dynamic nature of the interactions between natural and human systems calls for a systemic perspective when assessing hazards and the multiple, often interconnected risks associated with them as well as when designing solutions to reduce adverse impacts. While major advancements have been made over the last years in developing methods for risk and impact assessments, gaps persist when it comes to grasping the complexity of systemic risks and impacts linked to hazards and shocks and translating that knowledge into practical action and policy.
This session invites innovative research on the relationships between hydrometeorological hazards (e.g., drought, floods, heatwaves, hurricanes), systems vulnerabilities and compounding, cascading or systemic socioeconomic and environmental impacts. We welcome conceptual, methodological and empirical contributions that (i) identify causal chains and feedback loops between hazards, risks and impacts, (ii) propose innovative methods (qualitative and quantitative) to identify cause-effect relationships between hazards, risks and impacts in time and/or space, (iii) analyze how climate variability and socioeconomic factors influence the co-occurrence of hazards/impacts, systems vulnerabilities and their cascading propagation within and across natural and human systems, (iv) propose methods to transfer this knowledge and incorporate it into decision-making processes and risk management policies, including strategies to communicate and better visualize causal relationships, and (v) make an effort to couple these cause-effect loops into climate and hydrological models.
ITS1.1/CL0.1.17
Machine Learning for Climate Science
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
ITS1.6/BG1.18
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.
ITS1.18/CL0.1.8
Learning causal relationships from Earth system data is of paramount importance for understanding its complex dynamics, predicting future changes, and informing effective mitigation and adaptation strategies. Causal inference provides a powerful framework for unraveling cause-effect relationships of different processes within Earth system sciences. This session welcomes contributions that highlight innovative approaches, methodologies, and case studies employing causal inference techniques across Earth sciences.
The session aims to foster interdisciplinary discussions, encourage collaborations, and promote the development of robust causal analysis frameworks tailored to the unique characteristics of the Earth system. We welcome presentations from researchers across different disciplines, highlighting theoretical advancements and practical applications of causal inference to improve our understanding of Earth system processes.
The topics of interest for this session include, but are not limited to:
– Causal discovery methods: algorithms and methodologies for uncovering causal networks among Earth system processes;
– Causal effect estimation: statistical techniques to estimate causal effects of interventions or natural forcings in the Earth system;
– Applications of causal inference: case studies investigating causal pathways and mechanisms driving natural and anthropogenic perturbations such as climate change, land-ocean interactions, extreme events, etc;
– Causal modeling and network analysis: development and application of causal models, network analysis, and graphical models to capture the intricate interconnections and feedbacks within dynamical systems;
– Causal model evaluation: application of causal dependencies to assess climate models performance;
– Challenges and limitations: associated with the application of causal inference, including issues related to violations of assumptions, or uncertainty quantification.
ITS4.10/NH13.1
Extreme climate and weather events, associated disasters and emergent risks are becoming increasingly critical in the context of global environmental change and interact with other stressors. They are a potential major threat to reaching the Sustainable Development Goals (SDGs) and are one of the most pressing challenges for future human well-being.
This session explores the linkages between extreme climate and weather events, associated disasters, societal dynamics and resilience. Emphasis is laid on 1) Which impacts on ecosystems and societies are caused by extreme events (including risks emerging from compound events)? 2) Which feedbacks and cascades exist across ecosystems, infrastructures and societies? 3) Where do these societal and environmental dynamics threaten to cross critical thresholds and tipping points? 4) Can we learn from past experiences? 5) What are key obstacles towards societal resilience and reaching the SDGs and Sendai Framework for Disaster Risk Reduction (SFDRR) targets, while facing climate extremes and compound events?
We welcome empirical, theoretical and modelling studies from local to global scale from the fields of natural sciences, social sciences, humanities and related disciplines.
