Climate extremes are usually driven by complex regional interplays among human influence, internal climate variability, land-atmosphere interactions, and other factors like Arctic sea ice loss and polar amplification.
The accurate detection of changes in regional climate extremes is sometimes challenging due to observational uncertainties, such as non-climatic series discontinuities or station scarcity in regions like Africa or high altitudes. Reliable attribution of regional climate extremes usually relies on model skills in simulating the extremes. Global models actually provide some useful evidence for the role of human influence, while regional climate models could boost confidence in attribution to regional forcings such as land use/cover or urbanization. The attribution uncertainties could be caused by different attribution methodologies used, e.g., optimal fingerprinting or Bayesian statistics, and different model strategies employed, e.g., multi-models or single-model large ensembles. Besides, how to address internal climate variability remains a key source of the attribution uncertainties. Emerging advanced techniques like artificial intelligence (AI), have the great potential to substantially reduce these uncertainties.
This session provides a venue to present the latest progress in reliable detection, modelling, and attribution of regional climate extremes, especially in quantifying or reducing their uncertainties for better risk management. We welcome abstracts focused on, but not limited to:
- address the quality issue of daily observation data relevant at the regional scale
- assess the fitness of global or regional modelling by designing tailored diagnostics for climate extremes and their drivers in a regional context
- improve climate models to realistically represent regional climate extremes, in particular to convection-permitting scale at a fine resolution or to mega-heatwaves by adding relevant land-atmosphere feedbacks such as through dynamic downscaling
- reveal and evaluate the strengths and weaknesses of attribution methodologies used for different regional climate extremes
- develop new detection and attribution techniques for regional climate extremes, including large ensemble and AI algorithms
- find key physical or causal processes to constrain the attribution uncertainties
Finally, abstracts associated with projection uncertainties of regional climate extremes are also appreciated.
The accurate detection of changes in regional climate extremes is sometimes challenging due to observational uncertainties, such as non-climatic series discontinuities or station scarcity in regions like Africa or high altitudes. Reliable attribution of regional climate extremes usually relies on model skills in simulating the extremes. Global models actually provide some useful evidence for the role of human influence, while regional climate models could boost confidence in attribution to regional forcings such as land use/cover or urbanization. The attribution uncertainties could be caused by different attribution methodologies used, e.g., optimal fingerprinting or Bayesian statistics, and different model strategies employed, e.g., multi-models or single-model large ensembles. Besides, how to address internal climate variability remains a key source of the attribution uncertainties. Emerging advanced techniques like artificial intelligence (AI), have the great potential to substantially reduce these uncertainties.
This session provides a venue to present the latest progress in reliable detection, modelling, and attribution of regional climate extremes, especially in quantifying or reducing their uncertainties for better risk management. We welcome abstracts focused on, but not limited to:
- address the quality issue of daily observation data relevant at the regional scale
- assess the fitness of global or regional modelling by designing tailored diagnostics for climate extremes and their drivers in a regional context
- improve climate models to realistically represent regional climate extremes, in particular to convection-permitting scale at a fine resolution or to mega-heatwaves by adding relevant land-atmosphere feedbacks such as through dynamic downscaling
- reveal and evaluate the strengths and weaknesses of attribution methodologies used for different regional climate extremes
- develop new detection and attribution techniques for regional climate extremes, including large ensemble and AI algorithms
- find key physical or causal processes to constrain the attribution uncertainties
Finally, abstracts associated with projection uncertainties of regional climate extremes are also appreciated.