Scientists have found a way to predict big changes in complex systems, like ecosystems, by using signals from different parts of the system. These big changes, called tipping points, can drastically alter how a system works. For example, tipping points can explain events like mass extinctions or the start of an epidemic. Early warning signals can help predict these shifts before they happen.

It used to be unclear how to best combine early warning signals from different parts of a system. The new method, based on complex math, helps pick which parts of the system to watch for better predictions. This method looks at both the strength and the uncertainty of the signals. Using signals from several parts of the system can be helpful, but not always.

The researchers tested their method using different networks and models. They found that their approach works well, especially when different parts of the system experience different levels of noise and stress. For example, in an ecosystem, some species might show early warning signals sooner than others. Choosing the right set of species to watch can improve the quality of predictions and save time.

This new method can be used in many fields, like ecology, climate science, and disease control. It does not need detailed information about the system’s network, making it easier to use with real-world data. By focusing on the right parts of a system, scientists can better predict and respond to potential crises, making our understanding of complex systems better and more useful.