Anticipating critical transitions in real time is a fascinating, albeit challenging, task. The application of early warning signals – both for past events and in real time – is already underway. In face of the rising pace of available data and techniques for identifying critical transitions, there is a need for a coherent methodological framework for the application of available early warnings for critical transitions.
The web-based Early Warning Signals Toolbox aims to host all relevant developments in the field of early recognition of approaching critical transitions. We present background theory, but the main focus is on practical methods that can be used by anyone who is interested in testing these ideas in hers/his particular case of interest. We achieve this by presenting methods together with case studies. We also offer computer code freely available.
At the moment, we include only methods that can be applied for time-series data. In the near future, methods for spatial data will be also available. Fortunately, the list of methods we present is not exhaustive. We hope that the list will grow and we invite anyone interested in joining our efforts!