The Dresden University authors of this paper discuss reasons disaster statistics depart from a normal distribution, and network explanations of cascade effects. They look for applications of network theory to assess alternative emergency response methods.
Among the authors' comments:
"The tendency of globalization of economic and other systems is likely to increase the frequency of large-scale disasters, as it reduces the diversity required to stop certain chain reactions and to adapt to changing economic and enviromental conditions. Another danger is the ever-growing population and the trend to push social, economic, technological, and biological systems to their limits."
They analyze causality networks in earthquakes and resulting fires, power blackouts, storms and floods, terrorist attacks such as that on September 11, and epidemics such as SARS.
They propose some methods of mathematically modelling causality networks of disaster spreading and means of assessing disaster management models. They observe that disasters can often be characterized by power laws, which they attribute to the tendency to drive systems to their critical thresholds for increased efficiency.
They conclude that network theory may contribute further to disaster response management, including analysis of interaction of networks of networks. They suggest looking to the way biological systems have evolved to optimize network interactions and developing new principles of managing disaster respones based on self-organization.
Trafficforum: "Disasters as Extreme Events and the Importance of Networks for Disaster Response Management"
(originally posted February 14, 2005)