Terrestrial Carbon Cycle Dynamics in CMIP5 Last Glacial Maximum and mid-Holocene climate simulations

1. Context
New state-of-the-art global climate models are currently running the next generation of future climate change predictions, which will be incorporated into the fifth release of the Intergovernmental Panel on Climate Change (IPCC) assessment report. These models are more sophisticated than before and most include a wide range of biogeochemical processes for the first time (e.g. interactive vegetation, carbon cycle dynamics, atmospheric dust). These model simulations provide great opportunities to advance scientific understanding of Earth system feedbacks and provide more robust projections for future impacts of anthropogenic greenhouse gas emissions.

In order to develop this understanding inter-model comparison is essential to elucidate the processes that cause model discrepancies and how they influence predictions of climate change. Climate models which show broad agreement when modelling present day climate can differ considerably in their estimates of future changes. Modelling past climates such as the Last Glacial Maximum (LGM; 21,000 years ago) provide the opportunity to test the ability of those same models to recreate climates very different to today, and thereby offer greater potential to evaluate important Earth system feedback processes and how they contribute to future projections than by using modern-day validation alone.

In particular, the effect of climate feedbacks through biogeochemical processes, such as the carbon cycle, has been highlighted by CMIP5 (Coupled Model Intercomparison Project) and the IPCC AR4 as a key scientific issue. For understanding feedbacks through the carbon cycle, past climates again are vital to enable to study the Earth system in a state that is unperturbed by human emissions.

2. Aims and objectives
The main aim of this study is to use the simulations of two particular past time periods to test the ability of state-of-the-art climate models that being used for future climate projections to reproduce very different climates. The first time period is the mid-Holocene (6,000 years ago), when summer insolation in the northern hemisphere was larger than today, causing intensification of monsoon systems. Feedbacks between the hydrological cycle and the terrestrial biosphere resulted in large-scale vegetation in the Sahara, known as the 'Green Sahara'. There is good evidence about the extent of Sahara vegetation during the mid-Holocene, and it provides a strong feature to test in the climate model simulations. The second time period is the LGM, when large ice-sheets covered much of the northern hemisphere, carbon dioxide levels in the atmosphere were lower (180 ppm) and global temperatures were reduced by several degrees. One of the major questions from this time period is how changes in terrestrial carbon reservoirs and climate-carbon feedbacks contributed to the low atmospheric carbon dioxide levels. We aim to quantify changes in global terrestrial carbon storage and by quantifying and evaluating these natural processes this will help to constrain climate-carbon cycle feedbacks in future climate change projections.

3. Potential applications and benefits
This study will directly address evaluate realistic current climate models are and address the question of how natural carbon-climate feedbacks influence changing climates. They will thereby improve our understanding of the factors which are important for future projections of climate. The results of our analyses will contribute to the fifth assessment report of the IPCC. Ultimately, they will inform global and UK policy on allowable carbon emissions. In addition, scientists from a wide spectrum of research fields are likely to be interested in the results of this project. These disciplines include palaeoclimate and palaeoenvironmental research, Earth system modelling, palaeontology, biogeochemistry, and future climate change research.

In addition to those involved in research relevant to the proposal here, the results of this project will be used to engage at least three non-academic communities.

(i) The research and research methods have potential to motivate and inspire the next generation of scientists. There is currently a lack of scientists engaged in palaeoclimate research in the UK, and fewer with expertise with both modelling and palaeo-data. Our research will analyse state-of-the-art climate model simulations with direct relevance and input into the next IPCC report (AR5). It will also be multi-disciplinary, involving climate, carbon cycle, vegetation, and tracer dynamics, in terms of both model output and palaeo proxy-records for evaluation. We want to let students in higher education gain experience of analysing real cutting edge climate data to encourage them to obtain the confidence to consider careers in scientific research and engage their interest in climate science in general. This will be achieved through the development of a practical for undergraduate or postgraduate students which will use webpage based visualisation of the PMIP3 model data to perform a series of analyses. The practical will be employed in Bristol geographical sciences in the 2011-2012 academic year, and once successful, will be used for a wider base of courses and summer schools.

(ii) There is an enormous public interest in climate change, both natural and anthropogenic. We will seek to engage with the public and use our results to discuss how climate models are tested, how robust they are, and how modelling the past can inform future projections. To do this we plan to run an evening Science Café in Bristol in summer 2012. The subject matter will specifically concern the 'Green Sahara' during the mid-Holocene as a basis on which to address the aforementioned questions, whilst also examining the impacts of climate and environmental change on human societies in terms of past occupation of the Sahara.

(iii) UK policy makers, such as DECC and DEFRA, will benefit from the evaluation of the new state-of-the-art GCMs under past climate scenarios, which provide a greater test of the models to reproduce climates very different to today. The research addresses carbon cycle and vegetation feedbacks directly, and will help to constrain carbon-climate feedbacks, which will influence our understanding of current anthropogenic change, and therefore inform emissions targets. To disseminate our work we will hold a one-day workshop in Bristol in Nov 2011 with the purpose of bringing together key UK scientists working on future climate change with palaeoclimate scientists and policy makers to discuss the implications of the results for policy and future climate change research.