As I stated at one of the Berlin workshops in December, my favored methodological approach when confronted with a set of interacting complex systems is to try to break them apart, and create SCENARIOS from plausible ways in which the systems might interact.  To somewhat oversimplify the approach to Global Systems Science described in this agenda, I don’t think one can have any realistic hope of creating one big model to represent the likely interactions among several complex systems such as the economy, the energy system, the ecological system, the food and agriculture system, the global patterns of land-use, etc.  It is even hard for me to conceive of developing a successful model to simulate the future of even one such system.  So the reason I would propose to focus Global Systems Science around the creation of interesting future scenarios (like the four the Tellus Institute has recently updated) is that the components of these scenarios are not overly “hard-wired” together in ways that we can not really justify.  Also, the scenario methodology will inherently come with the types of narratives which the agenda correctly describes as quite useful.

Another reason for supporting a scenario creation approach to global systems science is because the global data by region for almost all the key parameters needed to support creation of one big simulation model (agent-based or otherwise) is terrible.  The poor quality of most data does not make it amenable, in my view, to the kind of intensive analysis described in the agenda, so I would also drop that approach.  When one creates complex scenarios for interacting complex systems, our experience is that a lot of data for various regions of the world has to be made up, based on comparisons to other regions.

To me, the key requirement for creating interesting scenarios for the future of our planet is intelligent and creative analysts who can actively consider futures that others dismiss.  Anyway, the key goal of Global Systems Science in my view, is to lay out a plausible set of scenarios that can be used to guide the world to a better future.  So to me, trying to simulate the future based on a set of equations in a model is somewhat besides the point.  The key set of questions we need to ask and answer is how we can change human behavior and both the individual and collective level to change the relevant set of complex sub-systems in the ways we want to change them.

Comments on my perspective above will be much appreciated.

— Rich Rosen, Senior Fellow, Tellus Institute, Boston