The climate system is a complex network with nonlinear interactions on multiple spatial and temporal scales among multiple variables. Due to its complexity, evaluating climate predictability, predicting climate changes, and forewarning major climate events have been grand challenges for a long time. Despite the rapid progress in dynamic models in recent years, it is still challenging for the current generation of models to fully capture many of the complex features of the climate system, thus inducing uncertainties in climate prediction and early-warning techniques. In recent years, novel approaches from complex-systems science, dynamical systems, and nonlinear dynamics as well as emerging machine learning and artificial intelligence approaches have been shown to be powerful with respect to estimating climate predictability and improving predictive/early-warning skill regarding climate changes/climate events; however, the extent to which these new approaches can compensate for the current dynamic models and further enhance our understanding of the climate system remains an open question.

In order to summarize the recent progress and promote the use of novel approaches in climate predictability, prediction, and early-warning studies, we would like to propose a special issue entitled Emerging predictability, prediction, and early-warning approaches in climate science. The special issue is intended to bring together researchers interested in complex-systems science, tipping points, and predictability. All submissions within the scope of this special issue are welcome.

Climate science, in particular climate prediction and projection, are heavily dependent on the use of Earth system models (ESMs), which are nonlinear, complex, and chaotic representations of the Earth’s spheres. As such, ESMs are susceptible to various sources of uncertainty. These include uncertainty in the initial state, parameter values, model formulation, structure, and external forcing. Ensembles have become a key tool to quantify these uncertainties and improve predictions. However, challenging questions remain regarding how to design and interpret such ensembles within the constraints of limited computational power and the lack of a rigorous framework for their design. Therefore, this special issue will be a valuable resource to climate scientists working on both theoretical and practical aspects of prediction ahead of Phase 7 of the Coupled Model Intercomparison Project (CMIP7) and future assessments.

This issue arises from the minisymposium Theoretical and Computational Aspects of Ensemble Design and Interpretation in Climate Science and Modelling hosted during the SIAM Conference on Mathematical & Computational Issues in Geosciences in Bergen, Norway (19–22 June 2023). It will feature works by participants as well as external contributions.

The climate system is a complex network with nonlinear interactions on multiple spatial and temporal scales among multiple variables. Due to its complexity, evaluating climate predictability, predicting climate changes, and forewarning major climate events have been grand challenges for a long time. Despite the rapid progress in dynamic models in recent years, it is still challenging for the current generation of models to fully capture many of the complex features of the climate system, thus inducing uncertainties in climate prediction and early-warning techniques. In recent years, novel approaches from complex-systems science, dynamical systems, and nonlinear dynamics as well as emerging machine learning and artificial intelligence approaches have been shown to be powerful with respect to estimating climate predictability and improving predictive/early-warning skill regarding climate changes/climate events; however, the extent to which these new approaches can compensate for the current dynamic models and further enhance our understanding of the climate system remains an open question.

In order to summarize the recent progress and promote the use of novel approaches in climate predictability, prediction, and early-warning studies, we would like to propose a special issue entitled Emerging predictability, prediction, and early-warning approaches in climate science. The special issue is intended to bring together researchers interested in complex-systems science, tipping points, and predictability. All submissions within the scope of this special issue are welcome.

Climate science, in particular climate prediction and projection, are heavily dependent on the use of Earth system models (ESMs), which are nonlinear, complex, and chaotic representations of the Earth’s spheres. As such, ESMs are susceptible to various sources of uncertainty. These include uncertainty in the initial state, parameter values, model formulation, structure, and external forcing. Ensembles have become a key tool to quantify these uncertainties and improve predictions. However, challenging questions remain regarding how to design and interpret such ensembles within the constraints of limited computational power and the lack of a rigorous framework for their design. Therefore, this special issue will be a valuable resource to climate scientists working on both theoretical and practical aspects of prediction ahead of Phase 7 of the Coupled Model Intercomparison Project (CMIP7) and future assessments.

This issue arises from the minisymposium Theoretical and Computational Aspects of Ensemble Design and Interpretation in Climate Science and Modelling hosted during the SIAM Conference on Mathematical & Computational Issues in Geosciences in Bergen, Norway (19–22 June 2023). It will feature works by participants as well as external contributions.