Carbon Cycle: land, atmosphere & oceans

Carbon is a critical part of the Earth system, as well as providing food and fuel for life. Improving our understanding of regional carbon budgets and the feedbacks between physical and biological processes involved in the carbon cycle is essential for addressing climate change. Earth Observation is a key tool for measuring and understanding
the carbon cycle across scales.

Carbon is continually cycling between the biosphere, atmosphere and the oceans. Atmospheric carbon dioxide is fixed by vegetation as it grows, it is absorbed by the oceans, and is released by combustion of fossil fuels or biomass fires. We want to understand the feedbacks between the physical and biological processes involved in the carbon cycle, as well as how this impacts people and nature more generally to enable us to better mitigate and manage climate and biodiversity impacts.

NCEO’s work in this area is focused on quantifying stocks and fluxes of carbon across scales, from specific locations such as vulnerable tropical forests and dynamic ocean basins, to continental and global scales. We are combining these observations with state-of-the-art models to improve our understanding of processes, reduce uncertainty in predictions of the future and provide information to help inform decision-making.

We are making use of the latest developments in sensors like lidar, UAVs and aircraft-based measurements to link from fine-scale up to global satellite observations. These links are crucial to help develop new satellite retrieval algorithms and validate observations of the carbon cycle from new space missions such as the UK-led European Space Agency BIOMASS mission. Along with colleagues in the Data Assimilation, biosphere-atmosphere and fire themes, we are exploiting new AI and high-performance computing tools to get the most out of these new observations.

A goal of this work is to help improve the COP global stocktake in 2028, feed into IPCC assessments and provide evidence as the UK aims for net zero and nature recovery. We are working with our stakeholders and the wider scientific community to help integrate EO data with C-cycle models to help in detection and attribution of change as well as identify knowledge gaps and new directions for observation systems.

Understanding the carbon cycle is essential for addressing climate change and protecting ecosystem health. This means making the most of all the tools available, from large-scale satellite missions to working with field measurement teams so that we know what the satellite measurements are really showing us.

Professor Mat Disney
NCEO Divisional Director of Carbon Cycle, based at UCL.

Key research groups

Carbon work in NCEO runs across a number of institutions, for example at:

• UCL: using lidar to measure and map forests;
• University of Leicester: using machine learning to develop new satellite maps of terrestrial carbon at global scales developing new models and datasets of ocean carbon cycle;
• University of Edinburgh: coupling observations and models across scales for predictions and diagnostics, using the UK JULES model;
• Plymouth Marine Laboratory: developing new models and datasets of ocean carbon cycle.

All our work is underpinned by expertise in data assimilation and large-scale computation, particularly with colleagues at the University of Reading and CEDA. We also work closely with colleagues at the UK Space Agency, the European Space Agency, NASA and others in developing new missions and methods.

Research priorities

• Identifying locations where we can quantify carbon stocks and fluxes top-down and bottom-up to reduce uncertainty and increase understanding.
• Exploiting new observations across scales.
• Locating hot-spots of change in the global carbon cycle.
• Improving models to help us attribute observed changes to their drivers.
• Working with stakeholders and the wider community to enable best use of EO information for policy action.

Our current work includes

• Measuring forest biomass using ground-based observations to help calibrate and validate satellite observations.
• Improving the process representation of biogenic ocean carbon cycle models and developing new long time-series of ocean carbon based on this.
• Developing machine learning-based approaches to upscaling satellite observations of terrestrial carbon from regional to global scales.
• Building new global diagnostics of terrestrial carbon cycling using ecological and atmospheric constraints and novel machine learning-based emulators of JULES.