This Workshop brings together leading researchers in climate science, statistics, impact modelling, remote sensing, artificial intelligence, and social sciences to advance our understanding, observation, and anticipation of climate extremes. It builds on the outcomes of the 2025 ISSI Forum on “Climate Extremes in a Rapidly Warming World,” which identified key gaps in our ability to observe, attribute, and anticipate high-impact and unprecedented events.
Over five days of structured sessions, participants will address the representation of dynamical, compound, and high-impact extremes in models; new methods for anticipating unprecedented events, from large ensembles to AI-based approaches; and the transition from hazard attribution to impact attribution through closer collaboration between climate and social sciences. The workshop will close with a synthesis session defining concrete priorities and next steps for the community.
Despite substantial scientific progress, important gaps remain in our ability to observe, understand, attribute, and anticipate high-impact and unprecedented events, especially those shaped by complex circulation patterns, compounding hazards, limited observational coverage, and the interaction between physical hazards and societal vulnerability.
The scientific rationale rests on three closely connected themes. First, in a non-stationary climate, strengthening the observational and modelling basis for climate extremes is a key priority. This requires bringing together multiple sources of evidence, including in-situ measurements, ground-based observations, reanalyses, space-based Earth observations, impact datasets, and model simulations. Second, anticipating unprecedented and high-impact extremes requires the integration of statistical methods, large ensembles, storyline approaches, attribution frameworks, and emerging AI-based and hybrid modelling techniques. Third, improving preparedness requires moving beyond hazard-centred perspectives towards impact-oriented approaches.
The Workshop objectives include:
- improve the representation and evaluation of dynamical, compound, and high-impact extremes in models;
- identify observational needs and benchmark datasets for climate and impact extremes, including the complementary role of in-situ, ground-based, space-based, and impact data;
- advance methodologies for anticipating unprecedented and worst-case events;
- strengthen the transition from hazard attribution to impact attribution;
- enhance the translation of scientific advances into actionable risk information.
The scientific outcome will be a collection of peer-reviewed papers addressing key aspects of climate extremes and preparedness. Initial topics include advances in observing climate extremes (including satellite remote sensing, climate data, and multi-platform hazard and impact datasets), methodological approaches for anticipating unprecedented and compound events, integration of observations, modelling, and emerging AI techniques for extreme-event analysis, and frameworks linking physical hazards with societal impacts.
In addition, the Workshop will include a public evening event (November 26, 2026) hosted at the University of Bern, aimed at communicating recent scientific advances on climate extremes and their societal implications to a broader audience. The event will feature a public lecture by Professor Ed Hawkins, from the National Centre for Atmospheric Science and the University of Reading, followed by a moderated panel discussion and an interactive exchange with the audience. More details will follow soon.