Estimating Wildfire Risk
Although the majority of wildfires in many areas of the world are caused by human activities (e.g., Spain) much wildfire risk research has focused on the biological and physical aspects of wildfire. Comparatively less attention has been given to the importance of socio-economic factors. With colleagues and me, Raul Romero-Calcerrada (Romero-Calcerrada et al. 2008) used methods from Bayesian statistics to examine the causal factors of wildfires in the area of central Spain in which my PhD research was located. Our results showed that spatial patterns of wildfire ignition are strongly associated with human access to the natural landscape. Our methods should be useful to optimize time and human resources in areas where urbanization is increasing the urban-forest interface and where human activity is an important cause of wildfire ignition.

Wildfire risk may also be examined statistically by examining the wildfire frequency-size distribution with concepts such as 'recurrence interval' and 'rotation period'. In the work that developed from my MSc research, Malamud et al. (2005) used normalized power-law exponents to compare the scaling of wildfire-burned areas between ecoregions and found a systematic change from east to west across the continent. We also produced maps of wildfire recurrence interval for the entire conterminous USA based on USFS data for the period 1970 - 2000. In the paper we suggested that these approaches allow us to classify of wildfire regimes for probabilistic hazard estimation in the same vein as is now used for earthquakes.

Example of a wildfire risk map produced using the GIS method