WP1 – Evaluation of existing decadal climate predictions to improve their societal impact

Background

Decadal climate predictions forecast the future evolution of climate for timescales ranging from a month to a decade into the future. The main characteristic of these predictions is that they require accurate information about the present (or ‘initial’) state of the climate system since it provides some predictability for the future climate. They are therefore often termed ‘initialised’ simulations. Decadal climate predictions come with their challenges: they require sets of hindcasts that are very expensive to run (these test the model by running it over a period in the past and comparing the simulations to observations), strategies to replicate observational uncertainty, and they strongly rely on a consistent observation system. However, improving decadal climate forecasting is crucial, since it covers vital timescales for decision makers, such as climate predictions for the next few years ahead.

Objectives

This Work Package works primarily towards EUCP objective 1, drawing on existing global climate prediction systems, collecting all the latest information and identifying the main factors that could improve their societal impact by:

• Outlining best practices in decadal climate prediction experimental design, real-time operations and analysis.
• Identifying the relative merits of decadal predictions from different sources, including the representation of regional climate trends and the assessment of the relevant uncertainties.
• Providing guidance on the use of existing climate predictions.

Exploring avenues to improve the next generation of climate predictions with experiments on the impact of resolution, initialisation, forecast drift reduction and short-lived climate forcings.

Read more about activities

• The team has collected pre-existing climate predictions from leading organisations around the world, with results of these predictions for up to 5 years ahead now available online. The forecast quality of these predictions has also been assessed, allowing end users to gain more informed insight into these relatively new predictions. This work will contribute to providing improved guidance on best practice when working with decadal predictions, as well as improving the next generation of climate models through a better understanding of the effects of various factors such as resolution and initialisation.
• Decadal prediction workshop: what are the biggest knowledge gaps?
• Smith et al, North Atlantic climate far more predictable than models imply, Nature
• Athanasiadis et al, Decadal predictability of North Atlantic blocking and the NAO, npj Climate and Atmospheric Science
• Smith et al, Robust skill of decadal climate predictions, npj Climate and Atmospheric Science
• Kushnir et al, Towards operational predictions of the near-term climate, Nature Climate Change

Objectives

This work package will draw on existing global climate prediction systems, collecting all the latest information and identifying the main factors that could improve their societal impact by:

• Outlining best practices in decadal climate prediction experimental design, real-time operations and analysis.
• Identifying the relative merits of climate decadal predictions, including the representation of regional climate trends.
• Providing guidance on the use of existing climate predictions.
• Exploring avenues to improve the next generation of climate predictions with experiments on the impact of resolution, initialisation, forecast drift reduction and short-term forcings.

The final goal is the integration across the European climate forecast systems developed for CMIP6 and those that will be built on them, making them a credible, reliable, authoritative, and action-oriented source of climate information to support more climate-resilient European economies and societies.