For instance, while IA researchers are trying to understand what drives land-use change, cooperation with ES researchers would allow them to assess the relevance of these changes in ways that go beyond their IA models. In this way, impacts via albedo changes or teleconnections could be included. Similar examples can be cited for modelling emissions of atmospheric pollution or future shipping routes. In our current paper, we provide several examples of areas that could benefit from further cooperation between the disciplines.

Cooperation can take different forms – from simple information exchange to full model integration. We have classified these forms of cooperation in our work and discuss criteria that may help determine the best way to cooperate. We applied these criteria to a list of potential collaborative projects and found that, at this stage, the simplest forms of cooperation often appear to be most useful. Indeed, it might be worthwhile developing the relevant mechanics involved in such partnerships ("meta-models") to find out how information could be better spread across the relevant research communities.

Finally, integrating complex models into a study appears to be particularly useful if there are strong interactions between the different research communities. These interactions could be important enough to change the way a system is analysed – one good example is the study of albedo changes and their impact on climate.