Probabilistic risk modeling at the wildland urban interface: the 2003 Cedar Fire

The October 2003 Cedar Fire in San Diego County was a tragedy involving 15 deaths, the burning of some 280 000 acres of land, the destruction of approximately 2227 homes, and costs of suppression near $30 million. It was the largest fire in California history. The data associated with the fire, however, do provide an opportunity to carry out probabilistic risk modeling of a wildland-urban interface (WUI) event. WUI’s exist where humans and their development interface with wildland fuel. As home building expands from urban areas to nearby forest areas, these homes become more likely to burn. Wildfires are an exceedingly complex phenomenon with uncertainty and unpredictability abounding, hence a statistical approach to gaining insight appears useful. In this research, spatial stochastic models are developed. These relate risk probabilities and losses measures to a variety of available explanatory quantities. There is a consideration of economic aspects and a discussion of the difficulties that arose in developing the data and of carrying out the analyses. Purposes of the work include highlighting a statistical method, developing variates associated with a destruction probability and employing the fitted risk probability to estimate future and possible losses. Copyright # 2008 John Wiley & Sons, Ltd.