Abstracts

Self-Organization and Natural Disasters

Keith Donnelly
Risk Analysis Methods
keith.donnelly@sympatico.ca

A major challenge facing climate change impact scientists is understanding whether recent extreme events - e.g. European floods, hurricanes, heat deaths in France - represent the beginning of an upward trend or are the result of natural or climate change driven variability.

The concepts of self-organization and fractal geometry are finding application in the description of natural systems and phenomena, such as watersheds, landscapes, earthquakes and forest fires. In a self-organized critical state, events follow a power-law relating magnitude and frequency. Engineered and social systems self-organize to achieve a balance between acceptable risk levels and benefits, using the onset of small events to signal an increased likelihood of unacceptable events.

In a risk or natural disaster scenario, various categories of systems are involved, e.g. natural, engineered, economic, managed, and social. Generally, elements of several are present, each of which may be undergoing internal self-organization. This viewpoint has not been explicitly applied, to our knowledge, in cases where the frequency or severity of extreme natural events may be increasing due to climate change. The apparent return periods of severe events may decrease markedly due to self-imposed increase of vulnerability.

Several examples of possible applications may be given: Urban heat waves (Chicago 1995, Europe 2003) have caused large numbers of excess deaths, particularly in groups whose vulnerability was increased due to changing demographic and socio-economic factors. In hydrology, increased urbanization and growth can modify cities' capacity to deal with large precipitation events, while channelization and deforestation can change river basin properties.

2005-04-05