Predicting when a complex system will go through a sudden change is hard — either you need vast swathes of information, or you make rather bad predictions. But we want to spot critical transitions before they happen, regardless of whether we're talking about irreversible damage to coral reefs, stock market crashes, or fishery collapses. It can be done if you know what to look for, but getting a sufficient amount of data be tough.
Enter a new proposal by Steven Lade from the Max-Planck-Institute for the Physics of Complex Systems in Germany and Thilo Gross from the University of Bristol published yesterday in the journal PLoS Computational Biology (free online). In it, the scientists argue that there is a way to see these things in advance without needing quite as much data. Their "generalized model" allows the use of partial information, and allows for unknown and unmeasurable variables. Their idea is designed to be used in conjunction with existing models to provide extra warning before things go pear-shaped. Sure, it might not be quite as accurate, but it's easier and faster.
And it's possible that a model like this one could help us spot critical changes to any important system — financial, climate, or otherwise. Maybe one day we'll be able to prevent collapses before they start.