Cellular automata have been successfully applied to simulate the propagationof wildfires with the aim of assisting fire managers in defining firesuppression tactics and in planning fire risk management policies. We presenta cellular automaton designed to simulate a severe wildfire episode that tookplace in Algarve (southern Portugal) in July?2012. During the episode almost25000ha burned and there was an explosive stage between 25?and 33hafter the onset. Results obtained show that the explosive stage is adequatelymodeled when introducing a wind propagation rule in which fire is allowed tospread to nonadjacent cells depending on wind speed. When this rule isintroduced, deviations in modeled time of burning (from estimated time basedon hot spots detected from satellite) have a root-mean-square differenceof?7.1 for a simulation period of 46h (i.e.,?less than 20%). Thesimulated pattern of probabilities of burning as estimated from an ensembleof 100?simulations shows a marked decrease out of the limits of the observedscar, indicating that the model represents an added value to help decide locations ofwhere to allocate resources for fire fighting.
展开▼
