Learning CA Wildfires from Probabilistic Risk Analysis of Colorado Wildfires

By Hira Dangol, Grant Glazer & Allen Wu

In recent years, Colorado, among other states, has seen rapid urban expansion. Termed wildland-urban interfaces, or WUIs, these areas are particularly vulnerable to hazards such as landslides, avalanches, and wildfires. This danger has motivated new risk analysis of natural disasters and investigations into preventative measures. The work in this paper is primarily intended to be a proof of concept for combining Markov models, decision analysis, and probabilistic risk analysis to simulate wildfire propagation, assess damages, and guide policy. We present a framework for assessing the effectiveness of pre-event mitigation actions in reducing wildfire risk in the WUI. We build and test a stochastic wildfire propagation and loss model on different scenarios, where each scenario is a different type of community-wide pre-event mitigation action as specified by the International Wildland Urban Interface Codes (IWUIC). The results give the expected damages from wildfire events in each scenario. We approximate the costs associated with each scenario and use decision analysis to find the optimal implementation of the IWUIC for the community in question. In this way, we are able to assess the value of pre-event wildfire mitigation in the wildland-urban interface.

https://www.researchgate.net/publication/326957309_Probabilistic_Risk_Analysis_of_Colorado_Wildfires