The CMIP 2026 Community Workshop was held in Kyoto, Japan from March 9 to March 13, 2026. As an Early Career Researcher (ECR) representative and the co-lead of the recent CMIP7 data request in land & land ice, I attended the whole workshop and became particularly interested in the in-person poster session 3.3: Extreme Events: Observations, Modeling and Single Model Large Ensembles (SMLE).
What is SMLE and what can we use it for?
SMLE by its name refers to a single Earth system model to be run multiple times during a period, differing only in initial conditions (with small perturbations in atmospheric fields). For example, the Community Earth System Model version 2 (CESM2) Large Ensemble Community Project (LENS2) belongs to this category with many ensemble members of up to 100. Each member of this large ensemble includes a full coupled simulations using CESM2 from 1850 to 2100 under CMIP6 historical and SSP370 future radiative forcing scenarios.
Atmospheric processes can be chaotic with tiny differences in starting conditions (e.g., temperature at one grid point at a certain vertical pressure height) causing nonlinear changes in atmospheric status. This creates uncertainty related to internal variability that will influence our understanding of the forced drivers (e.g., greenhouse gases) in determining the atmospheric responses. In this context, SMLE provides a useful modelling tool to create many possible weather trajectories with the same model and the same forcing. Therefore, SMLE is expected to be useful for reducing the uncertainty of internal variability and improved understanding of mechanisms related to atmospheric patterns that favor specific extreme events (e.g., extreme rainfall).
SMLE application in research of extreme events
As summarized by the IPCC Sixth Assessment Report, extreme events include temperature extremes, heavy precipitation and pluvial floods, river floods, droughts, extreme storms including tropical cyclones, and compound events that include dry/hot events, fire weather, compound flooding, and concurrent extremes. Motivated by high impact of these extreme events on societies, human lives, and ecosystems, this session aims to enhance our understanding of the mechanisms and predictability of extreme events using SMLE. According to this session’s description, SMLE has unique benefits for detecting the likelihood of multiple similar extreme events in the modeling dataset, offering an accurate estimate of the distribution of extreme events, and providing the possibility to find analogues to observed extreme events and exploring the related mechanisms.
Topics and reflections from the in-person posters
There are about 25 posters in this session with half of presenters being identified as the ECR. Focusing on the extreme events mentioned above, the topics in this session include extreme value analysis, prediction, projection, probability and distribution of extreme events, atmospheric teleconnection, SMLE design and evaluation using a variety of SMLE projects’ simulations in combined with observations. Given all possibilities simulated by the large ensemble simulations, key roles of, such as anthropogenic forcing in East Asian hot-wet compound extremes and in Antarctic sea ice variability, freshwater anomalies in the subpolar North Atlantic, ENSO phases in major fires in Indonesian Borneo, sea surface temperature in atmospheric structure over the Western North Pacific, and atmospheric variability in wind power generation, could be explored and potentially identified.
A key reflection from these topics comes from the powerful datasets that SMLE could provide for the research community. For example, extreme precipitation and flooding events can usually be hard to predict or forecast by a single model with limited ensemble members. But SMLE provides ample simulations that could capture the observed extreme precipitation and flooding in some of the ensemble members. In this case, researchers could identify the associated weather patterns given that SMLE provides comprehensive output of atmospheric fields at different pressure levels. This provides possibility for improving future forecasts of extreme precipitation and flooding events by recognizing early signals of similar weather patterns, which are derived from the SMLE analyses.
Outlook and key takeaways
From this in-person poster session, I saw the possibility of applying SMLE datasets to identify roles such as anthropogenic forcing in driving the changes in extreme events. SMLE is particularly helpful to identify atmospheric patterns and mechanisms associated with extreme events. Besides, given that atmospheric fields (e.g., temperature, precipitation, wind) are almost related to many aspects of our societies, SMLE looks like to have a continued potential to be applied to a broader field (e.g., wind power generation) to reduce uncertainty and improve our understanding of atmospheric mechanisms that could influence our societies, human wellbeing, and ecosystems.
