LEADER – Large Ensembles for Attribution of Dynamically-driven ExtRemes


Activity Leaders
Steering Committee

Alexey Karpechko, Finnish Meteorological Institute, Finland

Amy Butler, NOAA CSD, USA

Bernd Funke, CSIC Granada, Spain

Isla Simpson, NCAR, USA

Tiffany Shaw, University of Chicago, USA

Daniela Domeisen, ETH Zürich, Switzerland

Wenjuan Huo, GEOMAR, Germany

Jonathon Wright, Tsinghua University, China

Anja Schmidt, DLR, Germany

Andrea Steiner, University of Graz, Austria

Amanda Maycock, University of Leeds, UK

Activity description

The long-predicted climate change signal is emerging outside the noise in many regions.These changes in climate are accompanied by changes in extreme events that impact society. While early warnings of such changes are now potentially possible through, e.g., operational decadal predictions, there are several challenges: there is a lack of understanding of the dynamical mechanisms that enable such projections, there is evidence that global models underestimate some predictable signals, and these models suffer from biases. A better understanding of the causes of regional changes in climate is needed both to attribute recent events and to gain further confidence in forecasts. 

In order to meet these needs, a new Large Ensemble Single Forcing Model Intercomparison Project (LESFMIP) has begun. These coordinated model experiments will enable the impacts of different external drivers to be isolated. These experiments form a bedrock of the analysis plan of the World Climate Research Program Lighthouse Activity on Explaining and Predicting Earth System Change. The science that arises from the LESFMIP will allow for ongoing attribution statements to WMO State of Climate and Global Annual to Decadal Climate Update in 2025 and beyond. But in order to achieve this goal, the scientific community needs to better understand how to best use this model output, and also to communicate with operational centres as to which diagnostics are required for future analysis. Specifically, output from these experiments cannot be taken at face value; rather, we need to account for model errors, the under-representation of certain forced predictable signals in forecasts, while exploiting differences between models to diagnose the real-world situation. The large-ensemble will allow for isolating weak signals that otherwise would be buried under internal variability, while also offering a testbed for methods to extract predictable signals with correct amplitude. 

Many of the phenomena that can lead to seasonal to decadal predictability of near-surface extremes involve dynamical processes and/or composition changes in the atmosphere. These phenomena include variability of the stratospheric polar vortices in both hemispheres, changes in external forcing such as solar variability, volcanic eruptions, ozone depleting substances and short-term climate forcers; and internal variability such as the Quasi-Biennial Oscillation and changes in the storm track and jet stream. These phenomena bridge multiple existing APARC activities, and hence motivate the need for a limited term cross project focused activity to (1) coordinate the analysis and (2) ensure visibility of APARC science in WMO State of Climate and Global Annual to Decadal Climate Updates and IPCC reports. 

The Large Ensembles for Attribution of Dynamically-driven ExtRemes (LEADER) LTCF APARC activity is coordinating analysis of the LESFMIP output. Those interested in becoming involved should email Chaim Garfinkel () and Scott Osprey ().