Creating Synthetic Knowledge for Complex Systems (Model Integration)

To explore the dynamics of complex systems, integrated models describing those systems from different disciplinary perspectives are widely used. The use of integrated models has, however, serious limitations.

Raising the complexity of an integrated model does not necessarily result in a better representation of the underlying process; the accumulation of errors in the model’s multiple blocks of information can create misleading deviations in the outputs. 

Using a complex integrated model as the single research tool can be problematic because it restricts the researcher to working with simulations, leaving little room for the use of mathematical techniques. Another concern is that sample trajectories resulting from simulations may not deliver information on important generic cases and such gaps may lead to a misunderstanding of the system’s dynamics.

This project will provide methods for dealing with the limitations of integrated modeling by using alternative models, preferably of low and medium complexity, that describe a system from different perspectives and create synthetic knowledge of the underlying complex systems.

As a first step in the development of model integration techniques, researchers have undertaken a study on Conditional Integration Analysis. The CIA method was first used in the project on integrated assessment of uncertainty in net primary production of forests, launched jointly by the Advanced Systems Analysis Program and Ecosystems Services and Management Program in 2011.

Last edited: 19 January 2015