As climate modeling increasingly leverages exascale computing and researchers warn of an impending computing gap in climate research, the World Climate Research Programme (WCRP) is developing its new Strategic Plan – and high-performance computing is slated to play a critical role.

The WCRP itself is an organization of scientists selected by the World Meteorological Organization, the International Science Council, and UNESCO’s Intergovernmental Oceanographic Commission. Through its work, the WCRP aims “to facilitate the analysis and prediction of Earth system variability and change for use in an increasing range of practical applications of direct relevance, benefit and value to society.”

Its new Strategic Plan outlines WCRP’s strategy and high-level scientific objectives for achieving these goals over the next decade of scientific research (2019-2029).

The four main objectives

The 2019-2029 Strategic Plan outlines four primary objectives: improving the fundamental science of climate change research, improving predictive skills in climate modeling, refining simulations and projections, and improving connections between the natural and social sciences. The excerpt below covers these objectives in more detail:

Fundamental understanding of the climate system

We will advance the science of both reservoirs and flows of energy, water, carbon and climate-relevant compounds within and between the components of the climate system.

The objective is to close, within the limits of uncertainties, the budgets obtained from observations and models. […] Important to this effort is improving the foundation of our understanding concerning: the transports, reactions and transformations of key chemical species; the influences of atmospheric, oceanic and cryospheric dynamics, their modes of variability, and their interactions with anthropogenic change; and the interactions between resolved and unresolved processes.

Advancing predictive skill on timescales up to a decade

We will advance partnerships to push the frontiers of initialized climate prediction and quantify the uncertainties for shorter time scales across all climate system components. Understanding predictability of the climate system, and the relative contribution of each of its components, focuses our attention on societally relevant outcomes, such as extreme events, as well as the uncertainties, limits and capacities that are inherent in these outcomes. […]

Constraining projections on decadal to centennial timescales

We will quantify the sensitivities, uncertainties and emergent constraints inherent in the changing climate system on timescales that exceed our predictive skill for past and future climates. Fundamental to our understanding of the climate system is that there are limits to predictability of the system, some of which are inherent (non-linear processes and internal variability) and some of which are functional (future emissions, land use change, volcanic eruptions and solar variability). […] Moreover, developing research on emergent constraints and uncertainty reduction promises useful information on longer time horizons.

Connecting climate science with policy and services

We will improve the generation of decision-relevant information and knowledge about the evolving climate system, building on partnerships with Future Earth and others. Scientific research across a range of spatial and temporal scales has the potential to refine risk management and disaster response, economic and infrastructure planning, and adaptation and mitigation strategies. Improving the usability and use of climate science — in urban areas, in the regions, and globally — provides fertile ground for collaboration with a wide range of international research communities. New opportunities can be explored for the co-production of project design and outcomes that are directly relevant to policy and decision makers. The time is right for a focus on these deeply collaborative efforts.

High-performance computing is imperative

The Strategic Plan also outlines eight “Imperatives” – “essential infrastructure to execute this Strategic Plan,” with each imperative designed to overcome anticipated challenges. The imperatives include a diverse climate research community; a range of simulation tools; well-coordinated observational field programs; sustained, quality observations; communication and outreach; societal engagement; institutional partnerships – and, of course, “high-end computing and data management.” The Strategic Plan reads:

We require the technology and infrastructure to take advantage of advances in exascale computing and cloud-based systems and software. Open access to simulation and assessment products is fundamental to this imperative, as are technologies for big data, machine learning, improved modeling capabilities, and other computational advances. Inter-operable and reliable data and meta-data management are also key requirements.

The path forward for the Strategic Plan

The WCRP hasn’t yet finalized the Strategic Plan – it has been undergoing extensive consultation and discussion, and just last month was made available for public comment (which is ongoing at the time of publication).

While they await public comments, the WCRP is preparing an Implementation Plan – an outline of how the Strategic Plan will be achieved that should be made available in the near future. Hopefully, the Implementation Plan will delve into further detail about how the WCRP plans to accelerate high-performance computing in the climate modeling space.