Humanity faces ‘twin crises’ in biodiversity loss and climate change. Although we know the risks from these crises are large, we still have much to learn about their exact impacts, and how we can successfully avoid those impacts. Computer models play a crucial role in this respect.
Models have been developed to simulate changes in natural and human systems and can reveal a great deal about the ways in which crises play out. But in the process, models simplify key aspects of human societies and natural ecosystems. This can make it hard to develop crisis-avoidance strategies that we can rely on in the real world.
We suggest that different scientific fields are developing new modelling approaches that could help us to better understand – and avoid – crises. The most promising of these approaches involve simplifying real-world complexity in very efficient but accurate ways, producing models that experts can apply to big problems without losing their sharp focus. In particular, there have been recent successes with models that include just a few key processes that are known to be important across human and natural systems. We compare a number of suggestions about what these ‘key processes’ could be, and propose a list intended to stimulate discussion and model-based experiments. We also identify a range of existing models and datasets that could be especially useful, and the range of contributions they could make.
While models will never be able to predict exactly how and when a specific crisis will occur, they can help to forewarn and forearm us as we face mounting global challenges.
Relational Thinking 