Modelling interactions between ocean ecosystems, biogeochemical cycles and climate change

The weight of evidence is that the climate is warming on a global scale and that human release of carbon dioxide and other gases is responsible. Projections of the evolution of climate change over the next century are critical to policy development and are based on supercomputer simulations of the global ocean-atmosphere system, including the terrestrial and marine biospheres. The most uncertain parts of these simulations by far are the models that describe how ecosystems respond to a changing climate and in turn influence the amount of greenhouse gases in the atmosphere. This research project will deal with two issues that have received relatively little attention in the modelling community. The first is how do models cope with the fact that species and ecosystems are not passive but respond actively to changes in their environment and adapt to change. The second is how to accurately represent the impact of complex food web interactions on ecosystem dynamics without having to explicitly simulate the myriad of species that make up an ecosystem. The approach I will try to tackle both issues is to use patterns that apply at the level of whole ecosystems to define how the constituent parts must assemble to fit those patterns. I will focus specifically on ecosystem-level rules based on stability theory, that is on attempts to understand how ecosystems adapt and persist through perturbations and change. I will work with students, a research associate and international collaborators in developing and testing adaptive rules that can be incorporated in simulations of ecosystem dynamics. This work will generally involve 3 steps: (1) review and design ecosystem-level adaptive rules, using syntheses of ecological data to extract useful patterns, (2) evaluate the rules in a 'numerical laboratory' to ensure that they lead to realistic model behaviour, and (3) evaluate the rules in simulations of marine ecosystems compared to data sets covering a wide range of ocean environments. The long-term goal is to get ocean ecosystem models to correctly represent the present ecological and biogeochemical state of the ocean. If that can be achieved, there will be greater confidence in using these models to make projections about the future.