Coming Without Warning

Since the recent discovery of abrupt climate change -- that big changes can come quickly -- researchers have been looking for "warning signs" to help us avert "regime shifts" that could suddenly alter things we take for granted, such as storm tracks and weather patterns, sea levels and water supplies.

Recognizing a warning sign is tricky, though, because in a system that is subject to abrupt change, small variations can lead to impacts that are all out of proportion.  A widely recognized warning sign or "tipping point" is the recent unexpectedly high loss of Arctic sea ice, which could trigger major reorganizations of ocean and atmospheric circulation.  This temperature profile derived from ice cores in Greenland (taken from a book I wrote on the subject), shows numerous abrupt climate changes during the past 100,000 years.

But for all the "tipping points" that have been identified, a new mathematical study lends support to a darker, more discouraging train of thought -- that because the climate system is chaotic, with so many moving parts operating on different time scales, and feeding back on one another, its behavior is inherently unpredictable.   Abrupt change can come without warning.

Theoretical ecologist Alan Hastings, an expert on mathematical models of natural systems at the University of California, Davis, puts the implications of his analysis this way:

Many scientists are looking for the warning signs that herald sudden changes in natural systems, in hopes of forestalling those changes, or improving our preparations for them.  Our study found, unfortunately, that regime shifts with potentially large consequences can happen without warning -- systems can 'tip' precipitously.  This means that some effects of global climate change on ecosystems can be seen only once the effects are dramatic.  By that point returning the system to a desirable state will be difficult, if not impossible.

The research, co-authored by Derin Wysham, a research scientist at the John Innes Center, Norwich, UK, is published in the current issue of the journal Ecology Letters. The work applies to a large number of natural and man-made systems that share mathematical traits of chaotic behavior, such as wildlife and fish populations and securities markets.