Writing in Nature, Martin Scheffer of Wageningen University and colleagues detail how the early-warning signs of a shift could include a slowing down in the rate of the system’s recovery from small disturbancess, increased autocorrelation (so that the system is more similar to its past self), increased variance, more asymmetrical fluctuations, flickering between states and particular spatial patterns.
Most studies of these phenomena have to date been of simple models or carried out in the lab using lasers or neurons. The situation in the real world is much more complex but initial results suggest that these early-warning signals have been seen in climate and ecosystem changes in the past.
For example, an earlier analysis of eight abrupt climate transitions by Scheffer and others found that autocorrelation increased before the shifts. The researchers analysed the end of the greenhouse Earth period around 34 million years ago, the end of the Younger Dryas, the desertification of North Africa roughly 5000 years ago, the Bølling-Alleröd transition and the end of four glacial periods.
Similarly Bakke et al. reported in Nature Geoscience that flickering occurred before the abrupt end of the Younger Dryas cold period. And the increasing climate variability in the Pleistocene may have been a forewarning of a transition from a period of glacial–interglacial osciallations to a stable state of permanent mid-latitude northern-hemisphere glaciation, according to Crowley and Hyde in Nature.
Similar early-warning signals have been seen before shifts in the state of lakes, marine ecosystems, rangelands and financial markets, as well as in phenomena such as desertification, and destabilization of fish stocks. Human brains, meanwhile, have shown warning signals hours before epileptic seizures, as have lungs before asthma attacks.
“More work is needed to find out how robust these signals are in situations in which spatial complexity, chaos and stochastic perturbations govern the dynamics,” wrote the researchers. “Also, detection of the patterns in real data is challenging and may lead to false positive results as well as false negatives.”
But they also believe that the generic character of the early-warning signals is reason for optimism, because they occur largely independently of the precise mechanism involved. “Thus, if we have reasons to suspect the possibility of a critical transition, early-warning signals may be a significant step forwards when it comes to judging whether the probability of such an event is increasing,” said the scientists.
A follow-up project will see Marten Scheffer search for warning signals for migraine attacks.