NETTIP : Network Approaches to Climate Tipping Mechanisms

Topic description

Tipping elements in the climate system—such as ocean circulation, the Amazon rainforest, or ice sheets—can trigger abrupt shifts in temperature and precipitation patterns when the carbon cycle is sufficiently perturbed. We will use complex networks to represent interactions within the climate system and to study its behaviour during tipping phases. Simulation outputs will guide the application of network approaches to satellite data.

We expect this project to open new perspectives for understanding tipping behaviour in the climate system, leveraging the crucial interplay between topological network structures and dynamics, which we also plan to investigate using AI-based pattern recognition algorithms.

• Run the MIT General Circulation Model (MITgcm) together with additional models for continental ice and vegetation components under increasing atmospheric CO₂ concentrations.
• Perform climate simulations exhibiting tipping behaviour of one or multiple tipping elements on multi-millennial time scales.
• Test different types of climate-network constructions and indicators , using available packages such as py Unicorn and XGI, in order to identify the approach best suited as an Early Warning Signal (EWS) for a given tipping element.
• Develop operational EWS workflows by applying the network approach to satellite data.

06-01

Public funding alone (i.e. government, region, European, international organization research grant)

Funding further details

Université de Genève