My educational background is in Physics. I did a Physics degree at Warwick University and then went on to do a one-year course at Cambridge on Theoretical Physics ("Part III of the Mathematical Tripos").

My Ph.D. was in Plasma Physics at Imperial College London and Culham Laboratory (run by the United Kingdom Atomic Energy Authority). At that time (late 1980s) I was working on the theory of plasma instability in potential nuclear fusion reactors.

“The use of fossil fuels is so pervasive in our modern lifestyles that it is a great challenge for us to wean ourselves off these without compromising economic development.”

The main motivation for fusion research at that time was to find energy sources for when the fossil fuels ran out (i.e., from the perspective of continued global economic growth, it was considered there was not enough fossil fuel to burn). Now of course, a key reason for fusion research is to find a climate-friendly alternative to fossil fuels (i.e., from the perspective of climate change, there is too much fossil fuel to burn!).

Since joining the Met Office Hadley Centre for Climate Prediction and Research in 1990 I have worked on modelling future climate, with a focus on the role of the land biosphere (i.e., vegetation and soils) in climate change. This area of interest has been maintained throughout my subsequent periods at the UK Centre for Ecology and Hydrology (CEH), and now the University of Exeter.

I joined the Hadley Centre to work on global warming, but at that time the evidence of human-induced climate change was much less clear than it is today. So I started out on my research primarily because of an academic interest in predictive modeling of the Earth's climate, but as I became aware of the evidence of global warming I soon realized that this was also a very important problem to be working on.

Our 2000 Nature paper was the first published study to include a fully interactive carbon cycle in a state-of-the-art climate projection with a General Circulation Model (GCM). GCMs are at the heart of climate change research, and although these are very complex 3D numerical models of the physics of the atmosphere and ocean, they have tended to exclude interactions between the physical climate (i.e., air temperatures, humidities, and winds) and the biological and chemical components of the Earth System. This limitation seemed especially stark in the case of climate-carbon cycle interactions.

The ocean and land contain significantly more carbon than the atmosphere (about 50 times as much and about 3 times as much, respectively), and they exchange very large fluxes of carbon dioxide with the atmosphere. For example, the annual net land-atmosphere exchange of CO₂ is about 8 times as large as the annual CO₂ emissions from human activities. This means that slight imbalances between the "in" and "out" land-atmosphere and ocean-atmosphere CO₂ fluxes can yield significant changes in CO₂ concentration in the atmosphere, and could therefore significantly impact global warming. Furthermore, observations of atmospheric CO₂ tell us that the natural carbon cycle responds strongly to natural climate variations such as those associated with El Nino events or volcanic eruptions.

We therefore had plenty of reasons to believe that climate change would affect the fraction of human CO₂ emissions absorbed by the land and ocean (currently about half), and that anthropogenic climate change could therefore feed back on itself by influencing the rate of increase of atmospheric CO₂. But no one knew how important such climate-carbon cycle feedbacks might be, and we weren’t even sure of whether these feedbacks would be positive (which would accelerate climate change) or negative (which would slow down climate change).

In order to find out we took a state-of-the-art climate model (the Hadley Centre’s "HadCM3"), and in collaboration with others (especially at the National Oceanography Centre, Southampton), developed new components to model the ocean and land carbon cycle. Then we forced the model with a scenario of past and future CO₂ emissions from human activities, and allowed the interactive climate-carbon cycle model to determine the uptake of CO₂ by land and ocean, and therefore the rate at which atmospheric CO₂ increased. By disabling the new climate-carbon cycle interaction we were able to diagnose the impact of these previously neglected feedbacks on our projections of future climate change.

A large part of the reason for the paper being so highly cited relates to the dramatic effect we saw in this model. The climate-carbon cycle feedback was projected to increase atmospheric CO₂ by the year 2100 from around 730ppmv (already a cause for great concern) to more like 980ppmv. As a result the global mean warming by 2100 for this particular "middle of the road" emissions scenario was 5.5 K rather than 4K, with a mean land warming of 8K rather than 5.5K. The reason for this large positive climate-carbon cycle feedback was related to the failure of the land carbon sink, with a weak current day land sink for CO₂ turning into a strong source of CO₂ by around 2050, as global warming accelerated decomposition of the soil and caused "die-back" of the Amazon rainforest.

This dramatic result caused somewhat of a stir at the time of publication, and has encouraged a number of other GCM modeling groups to include an interactive carbon. One group, led by Pierre Friedlingstein at the Laboratoire des Sciences du Climat et de l’Environment in France, actually completed climate-carbon cycle simulations at a similar time to us, but got into print later. All of the existing coupled climate-carbon cycle GCMs also produce a positive climate-carbon cycle feedback (i.e., an acceleration of climate change). None as yet produce such a large effect as we simulated, but most are significant.

Much of my research these days is devoted to looking for observational constraints on the climate-carbon cycle feedback. Our Nature 2000 study raised more questions than it answered, which is often the case with highly cited papers—they tend to open up new research fields rather than close down old ones!

Whereas our 2000 paper was concerned primarily with climate feedbacks via the carbon cycle, the Betts et al. 1997 letter estimated vegetation feedbacks through energy and water fluxes. The global distribution and functioning of vegetation depends crucially on climate and is also influenced directly by CO₂. Furthermore, the distribution and functioning of vegetation determines the energy and water exchanges with the atmosphere, and thereby influences climate. So there is a potential for climate change to lead to changing patterns of vegetation, which further changes climate. Such "biophysical vegetation" feedbacks are also typically ignored in GCMs, and the Betts et al. 1997 study aimed to show how important this oversight might be.

We did this by coupling the Hadley Centre climate model (HadCM2, in this case) to the "DOLY" dynamic global vegetation model developed by our colleagues Ian Woodward and Susan Lee at the University of Sheffield. The latter describes how physiological and structural plant responses to increasing CO₂ act to offset one another. Although plant stomata close in response to high CO₂, which reduces water loss from each leaf, CO₂ also stimulates plant growth, which increases the number of leaves per unit area. The net effect in this model was a relatively small change in the land-atmosphere water flux. However, the results also highlighted the importance of vegetation in amplifying regional climate changes. Northward migration of the dark boreal forest was found to increase warming in the north due to masking of bright snow, and forest die-back in Amazonia was found to exacerbate local drying.

Everything. The use of fossil fuels is so pervasive in our modern lifestyles that it is a great challenge for us to wean ourselves off these without compromising economic development. As a result, I think we have to hit the problem with everything we have (carbon capture and storage, renewable energy, nuclear power, and perhaps even consider developing emergency climate engineering options).

I am optimistic that we can make progress, but it needs a global collective will that has so far been absent. But if we can keep making progress scientifically on the global warming problem I think we will help to define priorities for both governments and businesses. Once these are aligned we have a fighting chance.