Running Earth system models, downscaling outputs and driving process-based impact models are computationally expensive tasks that hinder the exploration of climate change impacts. New generations of statistical, physical, and hybrid climate change emulators reduce these bottlenecks by mimicking global, regional and extreme-event responses to generate projections of climate change and its associated impacts more efficiently. This capability transforms the assessment of climate change impacts and risks by enabling rapid policy-relevant analyses for sectors such as natural systems, agriculture, water resources, economics, energy systems and human health.

This session invites contributions on the developments and applications in climate change emulation, highlighting opportunities and challenges for robust and computationally efficient climate risk assessments. This includes simple climate models, probabilistic emulation techniques and machine learning approaches, including traditional pattern scaling, as well as benchmarking efforts that aid in interpretability and consistency between emulation methods. To foster interdisciplinary exchange, we also encourage submissions regarding approaches for climate impact projections.