Action 7: Improving global climate modelling

Where do we stand?

Global climate models that depict climate-relevant processes for the entire Earth have existed since the late 1950s. Initially, the focus was on describing the basic processes in the atmosphere. Only in the last three decades has it been possible to take other important influences on climate into account by coupling atmospheric models with ocean, land surface and sea ice models. Ocean models, for example, simulate ocean currents, which transport gigantic amounts of heat and are a key factor in understanding climate change. Reproducing the climate system and its components will remain a key task for research in the coming years.

What are the research needs?

In order for research to better predict future climate change, the spatial resolution of global climate models needs to be increased by about two orders of magnitude, from approximately 100 kilometres to around one kilometre. This requires the use of new artificial intelligence (AI) tools, for example in the areas of parametrisation, downscaling and the complete screening of huge data fields. In order to increase efficiency and reduce computing time, it is also important that research focuses more strongly on modularised model systems. These systems allow a wider range of uses, more effective adaptation to new computer systems and specific issues, and uniform quality management.

Implementation steps and milestones

• The development of a high-resolution global climate model will be a central component of the National Climate-Modelling Strategy (NMS), currently in planning. For the NMS, we are working jointly with research institutions to professionalise modelling as a consulting tool. Based on the NMS concept, from 2021 we will primarily introduce a greater division of labour in model development. This is intended to lead to more programmatic standardisations, more modularised model systems, and more efficient model maintenance.
• From 2021 onwards, we will focus our funding on expanding the development of high-resolution climate modelling using innovative AI elements.
• From 2020, we will continue our research of climate dynamics on geological time scales with the help of model studies of the palaeoclimate. This will make it easier to assess the resilience of climate projections.